Business Models for Mobile Broadband Media Services

– Case Study Indonesia Telecom Market

LAILI AIDI

Degree project in Communication Systems Second level, 30.0 HEC

Stockholm, Sweden

Business Models for Mobile Broadband Media Services – Case Study Indonesia Telecom Market

Master of Science Thesis

Student: Laili Aidi (aidi@kth.se)

Supervisors:

Jan I. Markendahl (janmar@kth.se), KTH Royal Institute of Technology Greger Blennerud (greger.blennerud@ericsson.com), Ericsson Sweden

Examiner:

Konrad Tollmar (konrad@kth.se), KTH Royal Institute of Technology

Communication Systems

School of Information and Communication Technology KTH Royal Institute of Technology

Stockholm, Sweden 2012

I!

Abstract

The increase mobile data traffic from the emerging Internet services, especially multimedia, has posed considerable challenges for the telecom industry. Their initial mobile data services business models are generally not compatible with these emerging Internet services. Thus, there is a substantial need to investigate the suitable options to make media as a profitable telecom business sector. However, there are different challenges and opportunity factors in developing sustainable mobile media business in each market, due to the unique circumstances applied as the result of customer characteristics, mobile market situation and regulatory/law enforcement.

The first purpose of this thesis is to explore the business model options to deliver media

services on top of mobile broadband. Although, we limit our focus to Indonesia, we first analyzed the worldwide patterns toward the media services in order to get a broader view of the current trend. We mapped multitudes of actor involved in digital online / on the top (OTT) media service, which together they form different types of constellation in the value network, as well as service, delivery and revenue model. We also put our focus to get the lessons learned from Spotify’s business model, by framing it using Chesbrough and Rosenbloom’s model.

The second purpose is to understand the Indonesian mobile user's characteristic toward the

mobile media services. We conducted survey to 119 Indonesians, analyzed and validated the result with the correlation tests (Cronbach Alpha and Pearson correlation), within the Unified theory of acceptance and use of technology (UTAUT) framework. Our findings confirm the low willingness to pay, but an open attitude for the services. The mobile device and network quality are not the barriers for them to adopt the services, and there is a tight connection between the decisions to adopt the services with the perception that the service is popular.

Through those findings, we assessed the feasibility of the identified options and

formulated the recommendations. We used our understanding about Indonesian market structure (telecom and media), regulation, and mobile user, as well as the lesson that we got from media services provisioning in Sweden and worldwide trend. We found that the pricing tiers, adjustable pricing, and differentiated features are some of the key success factors. Meanwhile, being part in the point-to-multipoint partnership with the well-known OTT player is the potential position that the Mobile network operators (MNOs) in Indonesia should take in provisioning OTT media services, rather than deliver the services by their own. Key words: business model, mobile Internet, multimedia services, mobile broadband

II!

Acknowledgements

I would like to acknowledge my academic supervisor, Associate Professor Jan I. Markendahl at KTH. Jan’s consistent insightful comments on numerous occasions are invaluable for me to improve this work. In addition, the extraordinary dedication and encouragements that he has shown to his students through the whole process of the research are ones that inspire me.

I thank my industrial supervisor, Greger Blennerud at Ericsson. Greger opened

possibilities to discuss with important people in the industry and gave access to the valuable market-related reading materials. He also exposed me to a valuable experience in real industrial environment and provided help in many aspects of this research.

I am also thankful to Associate Professor Konrad Tollmar at KTH, for accepting role as an

examiner for this thesis research. Konrad has also given helpful feedbacks and directions, even during the process of this research. I am also grateful to Jan, Greger, and Konrad for the chances that they have given to authoring together with them in conferences, which later encouraged me to explore fascinating things within this field.

I also have to address my appreciation to researchers and staffs at Wireless@KTH and

Ericsson Sweden and Indonesia, especially the team of Marketing and Communication department, Kajsa Arvidsson, Mattias Engvall, and Dharma Simorangkir. I thank them for helping in conducting interview with several contacts in the industry, providing supporting environment for me while doing this research, as well as sharing valuable ideas in numerous discussions and presentations.

Last but not least, a special grateful to my family for their support that makes me possible

reach this stage. I specially dedicate this work for my beloved one, in loving memory. Stockholm, October 5, 2012 Laili Aidi

III!

Contents Abstract ....................................................................................................................................... i Acknowledgements ....................................................................................................................ii Contents.....................................................................................................................................iii List of Figures ............................................................................................................................ v List of Tables.............................................................................................................................vi List of Abbreviations................................................................................................................vii Chapter 1 – Introduction ............................................................................................................ 1

1.1. Background.................................................................................................................. 1 1.2. Problem Definition ...................................................................................................... 2 1.3. Contribution................................................................................................................. 2 1.4. Thesis Outline.............................................................................................................. 3

Chapter 2 – Literature Study ...................................................................................................... 4 2.1. Business Model............................................................................................................ 4 2.2. Technology Acceptance .............................................................................................. 6 2.3. Mobile Multimedia Services ....................................................................................... 7

Chapter 3 – Methodology........................................................................................................... 8 3.1. Research Approach...................................................................................................... 8

3.1.1. Literature Study .................................................................................................... 8 3.1.2. Data Collection ..................................................................................................... 8 3.1.3. Data Analysis........................................................................................................ 9

3.2. Quality of the Study................................................................................................... 10 3.2.1 Reliability............................................................................................................. 10 3.2.2 Validity ................................................................................................................ 11

Chapter 4 – The Digital Music Industry .................................................................................. 12 4.1. Industrial landscape ................................................................................................... 12 4.2. Actors, Relations, and Networks ............................................................................... 12

4.2.1 Actors................................................................................................................... 13 4.2.2 Value Network of the Core Actors ...................................................................... 14 4.2.3 Value Network in OTT Distribution.................................................................... 15

4.3. Service, Distribution, and Revenue Models .............................................................. 18 Chapter 5 – The Digital TV/Video Industry ............................................................................ 21

5.1. Industrial landscape ................................................................................................... 21 5.2. Actors, Relations, and Networks ............................................................................... 21

5.2.1 Actors................................................................................................................... 21 5.2.2 Value Network of the Core Actors ...................................................................... 22 5.2.3 Value Network in OTT Distribution.................................................................... 23

5.3. Service, Distribution, and Revenue Models .............................................................. 27 Chapter 6 – Spotify Business Model........................................................................................ 29

6.1. Value Proposition ...................................................................................................... 29 6.2. Revenue Generation and Market Segment ................................................................ 30 6.3. Competitive Strategy ................................................................................................. 31 6.4. Value Network........................................................................................................... 32

6.4.1 Relation with the core actors of the music industry............................................. 32 6.4.2 Relation with the Telecom actors ........................................................................ 33

Chapter 7 – The Indonesian Market......................................................................................... 36 7.1. Regulatory Framework .............................................................................................. 36

7.1.1 QoS and Technology Convergence ..................................................................... 36 7.2. Telecom Market......................................................................................................... 37

7.2.1 Structure and Characteristic................................................................................. 37 7.2.2 Mobile broadband ................................................................................................ 38

7.3. Media Services Market .............................................................................................. 40 7.4. Mobile user ................................................................................................................ 41

IV!

7.4.1 Sample Characteristics......................................................................................... 41 7.4.2 Technology Acceptance....................................................................................... 42

Chapter 8 – Implications to Indonesia ..................................................................................... 47 8.1. Actors, Relations and Business models ..................................................................... 47 8.2. Indonesian Mobile Media user’s characteristic ......................................................... 49 8.3. Potentials and Challenges for MNOs in Indonesia.................................................... 50 8.4. Positions and Roles for MNOs in Indonesia ............................................................. 51

Chapter 9 – End of Discussion................................................................................................. 54 9.1. Criticism of the Study................................................................................................ 54 9.2. Future Work............................................................................................................... 54

Bibliography............................................................................................................................. 55 Appendix A – Qualitative Interviews .....................................................................................viii

A.1. Interviewee List .......................................................................................................viii A.2. Interview Transcripts and Notes................................................................................ ix

Appendix B – Mobile Network Operator in Indonesia .......................................................xxviii Appendix C - Quantitative Survey......................................................................................... xxx

C.1. Summary of the Hypotheses and Questions ........................................................... xxx C.2. Summary of Results................................................................................................ xxx

V!

List of Figures Figure 1. Data tsunami trend by various researches [53] ................................................................... 1!Figure 2. Business model concept hierarchy [1] .................................................................................. 4!Figure 3. The Technology Acceptance Model (TAM) [30]............................................................... 6!Figure 4. The Unified theory of acceptance and use of technology (UTAUT) [33] .................... 6!Figure 5. Digital music revenue and its percentage share over total revenue worldwide .........12!Figure 6. Conceptual actors and relations of the music industry’s core network........................14!Figure 7. Actors and relations of “Direct-To-Fan” in the digital music ........................................15!Figure 8. Typical actors and relations of the “Point-to-Point” partnership in the digital music

.............................................................................................................................................................16!Figure 9. Typical actors and relations of the “Point-to-Multipoint” partnership in the digital

music ..................................................................................................................................................17!Figure 10. Revenue stream share from the digital music [68].........................................................20!Figure 11. Actors and relations in the TV/Video industry’s core network...................................22!Figure 12. Typical actors and relations of the “Direct-to-Fan” partnership in the digital

TV/Video...........................................................................................................................................23!Figure 13. Typical actors and relations of the “Point-to-Point” partnership in the digital

TV/Video...........................................................................................................................................25!Figure 14. Typical actors and relations of the “Point-to-Multipoint” partnership in the digital

TV/Video...........................................................................................................................................26!Figure 15. Revenue share from the digital TV/Video [79] ..............................................................28!Figure 16. Spotify’s user based by service type in Europe, Sept 2008 – March 2011 [73] ......31!Figure 17. Spotify monthly revenue by service type in Europe, Sept 2008 – March 2011 [73]

.............................................................................................................................................................31!Figure 18. Actors and relations in Spotify point-to-multipoint partnership .................................32!Figure 19. Fixed broadband penetration in the BRICI countries [81] ...........................................37!Figure 20. Mobile phone penetration in the BRICI countries [81].................................................39!Figure 21. Question 1 result ..................................................................................................................xxx!Figure 22. Question 2 result ..................................................................................................................xxx!Figure 23. Question 3 result ..................................................................................................................xxx!Figure 24. Question 4 result ..................................................................................................................xxx!Figure 25. Question 5 result ..................................................................................................................xxx!Figure 26. Question 6 result ..................................................................................................................xxx!Figure 27. Question 7 result ................................................................................................................xxxii!Figure 28. Question 8 result ................................................................................................................xxxii!Figure 29. Question 9 result ................................................................................................................xxxii!Figure 30. Question 10 result..............................................................................................................xxxii!Figure 31. Question 11 result..............................................................................................................xxxii!Figure 32. Question 12 result..............................................................................................................xxxii!Figure 33. Question 13 result.............................................................................................................xxxiii!Figure 34. Question 15 result.............................................................................................................xxxiii!Figure 35. Question 16 result.............................................................................................................xxxiii!Figure 36. Question 17 result.............................................................................................................xxxiii!Figure 37. Question 18 result.............................................................................................................xxxiii!Figure 38. Question 19 result.............................................................................................................xxxiii!Figure 39. Question 20 result.............................................................................................................xxxiv!Figure 40. Question 21 result.............................................................................................................xxxiv!Figure 41. Question 22 result.............................................................................................................xxxiv!Figure 42. Question 23 result.............................................................................................................xxxiv!

VI!

List of Tables Table 1. Interpretation of Internal consistency from Cronbach alpha ...........................................11!Table 2. Interpretation of Correlation coefficient from Pearson Correlation ...............................11!Table 3. Revenue models in the digital music ....................................................................................18!Table 4. Service and Distribution models in the digital music........................................................19!Table 5. Revenue models in the digital TV/Video.............................................................................27!Table 6. Service and Distribution models in the digital TV/Video ................................................27!Table 7. Formation changes in Indonesia's Telecom Regulation....................................................36!Table 8. Hypothesis 1 result....................................................................................................................41!Table 9. Hypothesis 2 result....................................................................................................................42!Table 10. Average and Internal consistency of respondent’s willingness to use ........................42!Table 11. Internal consistency of respondents’ expectancies ..........................................................42!Table 12. Hypothesis 3 result .................................................................................................................43!Table 13. Hypothesis 4 result .................................................................................................................43!Table 14. Average of respondents’ expectancies ...............................................................................43!Table 15. Internal consistency of respondents’ expectancies ..........................................................44!Table 16. Hypothesis 5 result .................................................................................................................44!Table 17. Average and Internal consistency of respondents’ perceive to MNO network &

billing .................................................................................................................................................44!Table 18. Hypothesis 6 result .................................................................................................................45!Table 19. Average and Internal consistency of the respondents’ perceive in mobile device

feature and ownership.....................................................................................................................45!Table 20. Hypothesis 7 result .................................................................................................................46!Table 21. Average and Internal consistency of the respondents’ believe about social

environment ......................................................................................................................................46!Table 22. Hypothesis 8 result .................................................................................................................46!Table 23. Interviewee List ......................................................................................................................viii!Table 24. Interview with Lars Roth........................................................................................................ ix!Table 25. Interview with Andreas Liffgarden .....................................................................................xv!Table 26. Interview with Märta Rydbeck ...........................................................................................xvi!Table 27. Interview with Jockie Heruseon .........................................................................................xxi!Table 28. Interview with Yose Tireza Arizal ...................................................................................xxiii!Table 29. Interview with Didik Akhmadi Usman ............................................................................xxv!Table 30. Interview with Hengky Philip Ginting ............................................................................xxvi!Table 31. MNOs in Indonesia............................................................................................................xxviii!Table 32. Hypothesis and variable.......................................................................................................xxx!

VII!

List of Abbreviations Telecom operator Provider of telecommunication services that delivers telephony and

data communication access, and mostly functions as internet service provider (ISP) as well.

Mobile Network Operator (MNO)

Telecom operator that provides the communications services over wireless access that owns or controls necessary elements to deliver cellular services to the end-user, especially radio spectrum license and network infrastructure

Communications service provider (CSP)

Provider that delivers communication services to the end-user (telecommunications, entertainment and media, and Internet/Web services) encompasses telecom operator (landline and mobile), Internet service provider, and cable and satellite provider.

Mobile broadband Wireless Internet access through mobile device, using one of the wireless network technologies: CDMA2000 1X EV-DO (EV-DO Rev. A and Rev. B), WCDMA, WCDMA HSPA (HSDPA, HSUPA and HSPA+), TDSCDMA, WiMAX and LTE (FDD-LTE and TDD-LTE).

Third-generation (3G)

Wireless network technology defined by the ITU (International Telecommunication Union), include all of those listed under ‘Mobile broadband’ as well as CDMA2000 1X, some also includes the LTE and WiMAX which are not classed as 4G (LTE Advanced and WiMAX 2).

Average revenue per user (ARPU)

Revenue divided by weighted average number of customers at the same period (e.g. monthly).

Blended ARPU Revenue divided by average number of total subscribers at the same period (e.g. monthly).

Voice ARPU ARPU from voice services Data ARPU ARPU from data services Value added services ARPU (VAS ARPU)

ARPU from value added services (e.g. ring tones, ring-back tones (RBT), wallpaper downloads, screensaver downloads, music, TV/Video)

Non-voice ARPU ARPU from all non-voice services (Data, Messaging, Value added services)

Over-The-Top (OTT)

Services that are utilized over the network, delivered directly to the end-user via any connected devices, but the network access operator itself does not offer that. It rides on top of the services already provided and do not require any business or technology affiliations with the underlying network access operator

OTT TV/Video OTT service that range to TV broadcasting to video services, e.g. video conferencing, video chat, video on demand, Live TV broadcasting, etc

OTT Music OTT Music services, e.g. music streaming, music download, etc

1!

Chapter 1 – Introduction This chapter presents a brief introduction along with the description of the addressed problem areas of this study. At the end, the expected contributions to the knowledge and solution are described. 1.1. Background The evolution of Internet and growth of smart phone penetration have resulted a paradigm shift in society. It triggers higher expectation to the user and increases competition level across the markets. Cisco reported that the typical users using this device consume data traffic ~79 MB/month, compared to ~3.3 MB/month from basic feature mobile phone users. This is predicted to keep increase in the future, reaching ~1.3 GB/month [50]. Mobile data usage was also reported to grow to 83 % in 2011, in line with the data traffic tsunami predictions [51], as shown in Figure 1. This poses a considerable challenge for the telecom industry, such as weakening user relationships and diminishing revenue from existing voice and messaging services [52].

Figure 1. Data tsunami trend by various researches [53]

However, this also offers a potential moment for telecom operator to evolve, remain

competitive, and copes with those challenging situations. The data shows that mobile broadband subscriptions have increased to almost half of 2.4 million Internet users by 2011 [62]. It grew by 45% annually over the last 4-years and reached twice of the fixed broadband subscriptions. There are five emerging developing markets (Brazil, Russia, India, China, Indonesia), called as the BRICI countries that represent 45 % of the world population and predicted to add 1.2 billion more Internet users by 2015 [54]. Mobile broadband likely becomes a preferred access mechanism there, due to cheap smart phones, low PC penetration, limited fixed broadband, etc [55].

The combination between mobile broadband and media services offers a huge potential to

be a source of revenue and a key to growth for telecom operator in those emerging markets. The digital music revenue, for example, has increased to 8 % for the first time since 2004 [56]. It was followed by the growth in the subscription to streaming services, purchasing download, and expansion of digital media services across the globe, especially in the mobile devices. This is in contrast to what have happened in the fixed broadband, where the file sharing services dominated. The mobile broadband’s role in these emerging markets is important, as it is likely not just as a substitution for fixed broadband, as we have seen in developed markets [55].

2!

Indonesia is interesting to be a case study due to its market size and potential growth. The mobile penetration has passed 113 %, but it is still expected to grow [61]. In addition, despite the 3G has been rolled out for more than five years, its penetration is still less than 10%. The GDP is growing fast compared with neighbors in the region, followed by booming of the middle-class and young population [98][60]. There are tendencies to watch OTT video through mobile devices rather than the Internet-TV [57]. The personalization service is still dominating the digital music landscape and MNO has a strong position towards the billing relationship, as bank account penetration is extremely low, covering less than 20% of the population [59].

1.2. Problem Definition There are different challenges and opportunity factors in developing sustainable business, both when viewing it in the global and local market perspective. The combination of the aspects above creates unique circumstances to the Indonesian market. In addition, our findings show the complexity that the telecom actors should face to offer media services. First, it requires them to serve users with a higher demand and to understand the differentiation aspects to suit types of user segment [42]. Second, there is a trend in the telecom industry, where new actors with different economic interest enter the ecosystem and introduce changes [37]. Third, the digital piracy remains a barrier for business growth [56].

In addition, the initial mobile data services business models that the telecom actors usually use are not compatible with these emerging Internet services "#$%. Therefore, there is a substantial need to investigate and/or analyze suitable options in the existing mobile broadband business models, in order to make it as a profitable telecom business sector. However, the business model itself is a vast scope of a subject, as it covers broad range of aspects and is seen in different points of view [1][2]. The telecom industry itself also runs in a dynamic environment, e.g. in term of demographical, social, economical, and political situation. Despite the similarities among particular user segments across these emerging markets, there are differences in usage habit [54].

Therefore, the strategies used to monetize the potentials and to overcome the challenges

mentioned above have to be targeted to a specific service and market. Hence, the universal solutions would not be suitable. Therefore, the main intention of this thesis is to explore the business model options for mobile broadband media services in the Indonesian telecom market. In order to do that, the following four questions below are answered in this thesis: 1. Who are the actors involved in serving the media services, and what kind of relations and

business models do they use? 2. What are the mobile user’s usage characteristics, especially in Indonesia, toward the media

services through mobile broadband? 3. What kind of potentials and challenges would the MNOs in Indonesia face in providing

media services as a mobile broadband business? 4. What kind of positions and roles can the MNOs in Indonesia take to make media services

as a profitable business?

1.3. Contribution In order to analyze and evaluate the feasibility of solutions described in the problem definition above, we will need to see it from a multitude of perspectives. Therefore, this tele-economic discussion was done from a range of aspects and presents outcomes as below:

1. Business models and market structure We do classification and assessment of different types of existing business model that are currently used to deliver mobile broadband media services worldwide, and then put focus to Sweden, as our reference market. We then provide insights of the telecom and media market structure in Indonesia, which is our target market. In other words, this result mainly concerns about the supply side aspect. The discussion contains analysis of eco-systems, which are the

3!

identification of actors, their relation model, business roles, and responsibility distributed among them, as well as drivers and challenges of their strategies. This industrial background is structured as market maps that consist of telecom operators and other actors in the media and Internet industry. 2. Mobile media user The second outcome is the understanding of mobile media user, which means the discussion relates to the demand side aspect. The majority of users are not interested with the technical details, and the successful loyalty programs demand the provider’s ability to capture the actual or further composition structure of their preferences. In this context, we analyze the customers’ preferences’ patterns worldwide toward the media services, then we are in deep to Indonesia customer characteristic. It ranges to satisfaction aspect, willingness to pay and continue using the service, and expectation. This is done because, before we are able to identify the suitable business model options for a market, we need to first identify and predict these aspects. 3. Recommendation Through those two studies above, we analyzed the feasibility of the identified options. We used the market understanding data that we had about Indonesia as an affirmation tool. In other words, we highlight the available business model options to deliver media services on top of mobile broadband in Indonesia. We do this according to our understanding of its market structure (telecom and media), regulation, and end-user, as well as the lesson that we learn from the trend that has happened / is currently happening in the global and reference market (Sweden). We expect that, by using our analysis and recommendation, the MNOs can take a strong position and build profitable business in delivering mobile broadband using media services. 1.4. Thesis Outline The structure of remaining chapters in this thesis is done as follows. Chapter 2 lays theoretical foundation and frameworks required to understand the research areas. It ranges in the business model, technology acceptance field, as well as related study of the mobile media services.

Chapter 3 describes the activities undertaken to accomplish the research. This chapter also argues the frameworks and theory used as the basis of the work. Chapter 4 and chapter 5 present the introduction of current landscape of digital music and TV/Video industry. It is followed by discussion of the actors involved, and relations and networks that they form, as well as evaluation of the distribution models, service models, and revenue models used.

Chapter 6 contains the analysis of Spotify business model, and discusses its collaboration

with other actors in different industries. Chapter 7 presents the Indonesian mobile telecom and media market, as well as regulatory framework. After that, the discussion of the mobile users is presented, along with the empirical results of the quantitative study, where the hypotheses about these users are tested and analyzed.

Chapter 8 consists of conclusions of final analysis and implications for Indonesian market,

related to the research questions. The discussion is formulated based on the findings derived from the global market trend, lesson learned from mobile media service pre-study in Swedish market, as well as understanding about the Indonesian market (regulatory, telecom and media market, and mobile user). Finally, some critics of the study along with suggestions for future work are presented.

4!

Chapter 2 – Literature Study This chapter presents the review of theoretical foundations, frameworks, and related studies. The discussion ranges from the business model concepts, to the technology acceptances and the mobile multimedia services. 2.1. Business Model Despite its popularity, there is no common consensus on how to define the business model, and this term is often used to express different things [1]. Thus, it is important to stand on the specific definition, before we review various concepts. According to [2], business model is “a conceptual tool containing a set of objects, concepts and their relationships with the objective to express the business logic of a specific firm. Therefore we must consider which concepts and relationships allow a simplified description and representation of what value is provided to end-users, how this is done and with which financial consequences.” Following this, various business models written in range of sources are classified into three categories, which might be hierarchically linked to each other, as shown in Figure 2.

Figure 2. Business model concept hierarchy [1]

First, business model is stated as a concept that can be used to describe various businesses

(Concept point of view). This is the abstract layer, which is designed to answer typical basic definitions and/or the meta-model questions. The works in [2], [8], and [9] for example, provide both concept and meta-model framework to design a business model. The work in [2] defines the Business Model Ontology (BMO) that constructs business model using four main pillars (product, customer, interface, infrastructure management, and financial aspects), and then split it into nine interrelated business model building blocks as core of the ontology (value proposition, target customer, distribution, channel, relationship, value configuration, capability, partnership, cost structure and revenue model). In [9], this framework consists of six components, which are the value proposition, market segment, value chain structure, revenue generation and margins, value network and competitive strategy.

Second, business model is explained as types of model that each is used to describe a set

of business with common characteristics (Taxonomy point of view). The typical questions answered here are either the taxonomy of types and/or the sub-meta question. One example is

5!

[6] that classify the different business models according to the types of value proposition and revenue stream. It classify nine generic forms of e-business models based on value proposition and revenue stream are proposed, which are brokerage, advertising, infomediary, merchant, manufacturer, affiliate, community, subscription, and utility model. Meanwhile, the work in [5] defines specifically business model in wireless business based on revenue stream and target user segments. This was modified from previous models, which consist of nine models as well: advertising, application provisioning, brokerage, content provisioning, infrastructure provisioning, network operating, service provisioning, application service provisioning, internet service provisioning, and terminal manufacture.

Third, business model is discussed as a “bottom-line” analysis that represents an instance

of the generic business model (Instance level point of view). It describes the real world business, such as the particular operating case or firms’ business model. The works in [3] and [4] cover the tele-economic calculation, which are the network rollouts assessment for a mobile broadband access. The objective is to evaluate, present the viewpoint from radio-access network investment, and make an evaluation of providing amount of user with forecasted traffic demand. The profitability and the viability of business model are then evaluated using cost and revenue matrix. The work in [9] also matches into this category, as it assesses the implemented and existing business models. In this case, the Xerox’s business model is analyzed using the framework designed previously. The Network-centric approach There has been a concern raised that business model should also be seen as cross-firm collaboration as the value network analysis. Focusing to only a single company’s perspective is insufficiently suited to address the services that require multi-actors collaboration [10]. The mobile telecom business itself is also getting complex because there is multitudes of actor outside the industry, with different economic interests, enter the market [12]. Therefore, it is fundamental to identify all of these actors and their roles, as well as to understand their interaction with each other. Numbers of works specialized to ICT field have broad business model analysis of traditional manufacture industry into variety of actors, interaction and strategy discussion, either in the concept, taxonomy and/or taxonomy point of view.

In [5], there has been an identification of nine basic types of actors and roles in the wireless business, which are application provider, application service provider, content aggregator, content integrator, content provider, infrastructure provider, network provider, service provider, and terminal manufacturer. In [13], there are assessments of two existing models for multi-actors analysis, one is based on the system theory (MACTOR) and other is game theory (Allas model). The new model, MASAM, is then proposed to capitalize and overcome previous’ strengths and weaknesses. The cooperation models used by telecom operator to interact with some of these actors are discussed in [7], using the Actor, Relation, and Activity (ARA) point of view. It focuses to three services (public mobile broadband access services, services and solutions for indoor wireless access, and mobile payment and ticketing services), by identifying main drivers of cooperation and how to organize it.

The primary lines of reasoning of the partnership between actors are grouped in [12]. The

first motive argued is “Cospesialization”, which is to make the product in optimal and economically scale through specialized resource or knowledge of partner. Such common examples are outsourcing and sharing infrastructure. The second motive is “Cooption”, which is to reduce or share the risk and uncertainty in the environment. Such example is the engagement in a temporary alliance with the actor/s that was/were previously the competitor (“co-opetition”). The last is in order to do “Learning and internationalization”, which is the resources acquisition. The targets are mostly knowledge, data, or end-user access. This is done in order to leverage the business model or competencies, e.g. to conquer the new market. This is also mentioned in [18] as “accessing complementary asset”. This is done to leverage the existing business model and/or competencies.

6!

2.2. Technology Acceptance Mobile media streaming service would bring shift in the habit of the early user. The diffusion of a new technology is a result of decision series and comparison factors from end-user’s perception regarding the benefits and costs of adoption. Therefore, the ability to predict this diffusion plays an important aspect in assessing the suitable business models of new services, because the potential user might not adopt it, in spite of the availability of that technology in the market. This factor includes user’s readiness to use and willingness to continue using the technology, especially in Indonesia where mobile broadband media services do not take off yet, compared with the developed markets.

Researchers have performed various theories and designed frameworks to analyze factors

that affect end-user intention to adopt new technologies and services. However, substantial theoretical and empirical works accumulate to some of it. Such works in mobile Internet field are [34], [35] and [36] those adopt, extend, and combine several frameworks. They assess range of mobile value-added services, due to its flexibility to fit in various cases. The prominent one is Technology Acceptance Model (TAM), shown in Figure 3, designed for two purposes during early stage of computer [30]. First is to understand a successful acceptance process from implementation of new information system. Second is to provide a basis of methodology to predict further usage of a technology after a brief interaction.

Figure 3. The Technology Acceptance Model (TAM) [30]

However, as TAM gives more attention to examine the factors that drive user’s adoption,

it is inadequate to assess how the external interventions would affect individual perceptions and their attitude intention. These are important to evaluate successful new implementation of information system [31]. Moreover, TAM is also designed to examine user’s adoption and intention to continue using the technology after user had exposed to it. This means, this framework is rather to explain post-adoption behavior. It is not originally designed to be used during pre-implementation stage, which is also critical to assess the sustainability of the new technology [32].

Figure 4. The Unified theory of acceptance and use of technology (UTAUT) [33]

Design feature Cognitive response Affective response Behavioural response

Perceived usefulness

Perceived Ease Of Use

Attitude Toward Using

Actual System Use

X1!

X1!

Xn!

Performance Expectancy

Effort Expectancy

Social Influence

Facilitating Conditions

Behavioural Intention

Use Behaviour

Gender Age Experience Voluntariness of Use

7!

The Unified theory of acceptance and use of technology (UTAUT) should be taken into account [33]. This framework is visualized in Figure 4 above. It was designed as an extension of eight previous prominent models1, in order to overcome their imitations, especially on considering social context and personal expectancy. This model is able to explain 70% of the observed variances in the study; compared to 17 - 53 % of TAM’s cite success. It has been tested to the technology usage cases for both individual and organizational/enterprise oriented, as well as in the voluntary and mandatory context. It includes the demographic factors ignored previously and it was designed to examine the technology during their initial introduction state.

The UTAUT framework consists of four user acceptance and behavior factors. The

performance expectancy means a degree that user thinks that this new technology will help them in gaining in a job performance. The effort expectancy is a degree of ease associated with usage of the new system. Social influence is a degree to which user perceives the importance of others to think he/she should use this new technology. The facilitating conditions a degree that user thinks the infrastructure is available to support them to use this new technology. These determinants lead to behavioral intention, with is influenced by four key moderators: gender, age, experience, and degree of voluntariness to use the technology.

2.3. Mobile Multimedia Services Several works have been done to research the emerging media services in mobile environment, but most of it focuses to either the technical underlying details or technology acceptance factors. As far as our knowledge, the study about mobile media services that analyses both the business aspect (in multi point of views) and user acceptance factors are not covered yet, especially one that is intended for the Indonesian market and considers the unique circumstances there (regulatory, telecom/media market and mobile user characteristics)

The work in [38] assesses the factors that will affect the mobile video services usage and

implications to the service providers. It limits the focus to user-generated on-demand video, where the assessment was done based on theories of the Long tail, user-generated content (UGC), technology acceptance, as well as mobility and self-expression. The findings covers the influence factors connected to the service offering and value for the users, as well as indicate the importance to the service provider to cooperate with other actors in offering and marketing the service.

Moreover, the study in [39] also explores user’s preference in the mobile multimedia

service and put the focus to the young people. It was done to investigate the possible value added services from the multimedia that can be offered to this user segment. The findings, which were investigated using the Innovation Diffusion Theory, reveal that the cost and technical constraint to use the service are the main barriers for this user segment, and it is followed by the knowledge level about the offered services. This knowledge factor is also confirmed in [38] as one of the important factors that should be considered by service provider.

Furthermore, the impact of the convergence that happens in the multimedia business is

analyzed in [17]. This work explores how this convergence drives change in the competitive strategies of the media and communication actors. It also mentions about multitude of actors form different origins that may attempt to assume the most favorable position, and assess the strategic options that can be taken by these actors in form of migration and integration. However, this work only point out the actors’ positioning in value chain point of view, and actors’ constellation in the value network were not covered.

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!1 Technology Acceptance Model (TAM), Theory Reasoned Action (TRA), Motivational Model (MM), Theory of Planed Behavior (TPB), Combined Theory Acceptance Model and Theory of Planned Behavior (C TAM-TPB), Model of PC utilization (MPCU), Innovations Diffusion Theory (IDT), Social Cognitive Theory (SCT)

8!

Chapter 3 – Methodology This chapter presents the motivation of the research approach, theories, and frameworks used. In addition to that, we also present the detail methods for the data collection and data analysis, as well as strategies to assure the quality of this study. 3.1. Research Approach In many cases, a research is done either using a qualitative or quantitative approach. The intention of this thesis is more into understanding the phenomenon and answering the problem that the important factors are still unclear. Thus, the main research approach that fits to be used is a qualitative.

However, there is also a need to use the quantitative approach, especially to get an understanding of the media user by using a statistical validity. Therefore, the mixed approach is suitable, where the qualitative approach is complemented with the quantitative approach [19]. We applied this approach to four research phases: literature study, data collection, data analysis, and then forming the conclusion and recommendation for the future research. 3.1.1. Literature Study

As the research needs to dive into a range of perspectives, we first establish clarity of theoretical building blocks. We have to get an understanding of the business model concept. This includes the mechanism of how frameworks and theories work, as well as their antecedent and consequence. In this discussion, we also get deep into a value network discussion, which is to model the relevant actors in the digital media ecosystem and relations among them.

The industrial background and user’s acceptance factors also play an important aspect, which also as a based of the analysis. Thus, we reviewed the technology acceptance frameworks, especially ones that can be used to study the mobile services. This was used to understand the current and future pattern of media services diffusion that is delivered on top of mobile broadband in the Indonesian market.

3.1.2. Data Collection

Beside the literature study, there were also data collection and analysis stages. These were done to understand the business models used to deliver the media services on top of broadband in worldwide markets today. However, the final intention is not to design such new solutions, but to evaluate the existing ones and to identify suitable options for the case study market. Therefore, analyzing the pattern of existing business models becomes a requisite, before we be able to evaluate and propose the relevant options. In order to do that, we used the data collection techniques, as mentioned below: • First, we conducted the unobtrusive technique through secondary resources, as the basis

of our research problem definition and pre-study cases analysis [20]. Discussions with the advisors and examiner also helped us to formulate the problems. Different types of source from both the relevant and adjacent areas were studied. It consists from academic and market report, to the newspaper and magazine articles. The KTH library catalogue was used to gather the scholarly journals2, while the market reports and analysis documents were mainly searched through Ericsson’s business intelligent center (BIC)3.

• Second, we have undertaken the qualitative interviews with the contacts in the companies

that operate in the reference market (Sweden) and target market (Indonesia). These persons are relevant due to their expertise and professional activities in the telecom and

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!2 “KTH Bibliothek”, available http://www.kth.se/kthb/ 3 “Ericsson Business Intelligent Center”, available http://internal.ericsson.com

9!

media industry that relate to our study. These interview sessions took around 1 hour and conducted as semi-structured and open questions4. This was done to get an understanding of the context of these actors’ decision, intention, and opinion about the case we were studying. It means we performed the expert sampling [20] that covered the contacts from MNOs, telecom regulator, OTT music service provider, TV broadcaster and mobile content provider. The list of the interviewees is available in Table 23 at Appendix A – Qualitative Interviews.

• Third, we gathered the quantitative user data using a simple random sampling survey

[20]. This empirical data was used to perceive user patterns (behavior, expectation, perception, and acceptance) toward the mobile media services. Therefore, we could get a better understanding of this market and snapshot of end-user affirmation. The questionnaire was a combination of types of question: multi-item statement (“Likert response scale”), multi-choices, dichotomous, contingency, and open questions. We first conducted the pre-test survey to a small sample of respondents in Sweden, to improve the questionnaire accordingly. The survey then was opened for 14 days from July 30 to August 12, 2012 and published in the survey monkey, an Internet-based survey tool5. The link to this survey was privately mailed to an extensive list of 1488 Indonesians, who were randomly selected from social media websites and mailing lists. During that period, we got 195 responds, which consist of 119 valid data. Then, this valid data was used to test and validate our hypotheses about the Indonesian mobile users.

3.1.3. Data Analysis

It is important to do the analysis for the broadband business in its complete ecosystem (network, aspect, services, and application aspects), in order to get a complete understanding and picture of what we are dealing with [40]. In the network aspect, we focused on the analysis of the business model specifically for mobile broadband. The service and application aspect was concentrated on the multimedia in entertainment aspect, covered digital mobile music and TV/Video application. The analysis about end-user aspect covered the consumption and preference pattern of the mobile user to those services, in the global and Indonesian market perspective.

Based on the problem domains defined previously, it is clear that the scope of this thesis is

not to form new business models, but to assess the suitable options. Thus, the analysis was done with the pre-study and post-study analysis according to theoretical framework mentioned in 2.1. This pre-study analysis consists of three parts: • First, we saw the business model as value network of entities (Network centric

approach) [10], because there are multi-actors involve in the media business ecosystem. We identified these actors, what kind of roles they have, and the value network they form in the OTT music and TV/Video services provisioning. We used the actors, resource, and activities (ARA) point of view [7], where the identification of these actors was elaborated from the work at [5].

• Second, we analyzed common characteristics that lie in the business models used to

deliver media services worldwide. We evaluated different business models (Taxonomy point of view) [2], by elaborating the work at [6] and [5]. Here, we observed different forms of business model based on the service, distribution, and revenue model.

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!4 The interview questions for the representatives from companies based in Sweden are designed together with the other students that studied similar topic but different case study, which are Ekambar Selvakumar and Jin Huang 5 “Layanan Media pada Mobile Broadband di Indonesia”, available at https://www.surveymonkey.com/s/8LH9LLR

10!

• Third, after getting the understanding of these digital media industries landscape, we evaluated specific firms’ business model (Instance point of view) [2]. We chosen Spotify, as it is one of the current prominent OTT and mobile media service providers in Europe and have inter-actor relation within its business model. We framed our analysis by using six elements in the framework designed in [9].

These three analyses provide us the conceptual bridge to align the triangle of strategy, organization, and technology. We need to do those, as mobile broadband in Indonesia is still immature and mobile media services have not taken off yet. Therefore, there is not much to learn if we only see it from a local point of view. Therefore, it would be hard to analyze this target market, without first doing the pre-study analysis to the mature markets and get understanding of the landscape.

While doing this pre-study analysis, we developed hypotheses that later are then tested in

the post-study analysis, which was OTT music and TV/Video services provisioning in Indonesia’s mobile broadband market. These hypotheses and its tested variable are listed in Table 32 in Appendix C.1. Using this, we assessed the Indonesian mobile users’ acceptance and readiness toward media services through mobile broadband within the UTAUT framework [33], which was chosen as it is an improvement of the previous prominent user acceptance models and has widely used among researcher, as discussed in section 2.1.

UTAUT is based on belief that individual factors affecting user’s decision to accept or

reject a technology or service can be identified and measured. According to this, user’s expectancy for performance, effort for adoption, and social environment would influence behavioral intention. This intention, along with the facilitating condition factor, is fundamental determinant of the actual usage behavior. It also considers age, gender, experience, and voluntariness aspect, and it has been successfully applied in numerous studies. Therefore, this is suitable to be used in this study, to understand Indonesian mobile users’ characteristic toward mobile broadband media services.

3.2. Quality of the Study 3.2.1 Reliability Reliability refers to the consistency and/or stability of a particular result from procedures done in the different circumstances, assuming the variable factor/s has/have not changed [21]. As this thesis uses mixed approach, thus the reliability to both of the qualitative and quantitative study should be taken into account. The reliability of the qualitative study was ensured by letting the interviewee be informed about the interview protocol and the questions prior to the interview session.

In addition, because the interviews are semi-structured and open questions, the interviewer

might repeat and paraphrase the information relayed by the interviewee, to correct any misunderstanding. We mostly recorded the conversation on behalf of the interviewee’s permission, which makes us possible to re-analysis the information in its whole context later during the post-interview session. The main points from the interviewee’s answer to each question are listed in Table 24 to Table 30 in Appendix A.1.

The reliability of the quantitative study was ensured using internal consistency, by testing

the hypotheses in a single survey. Internal consistency refers to the degree of the relation among items measured [22]. This means, in this study we assessed the internal consistency by measuring how the results (variable) from different questions that are designed to measure the same hypothesis, yield to the similar construction. We assessed these results using two prominent types of Correlation tests, which are Cronbach alpha and Pearson product-moment correlation (Pearson correlation).

11!

Cronbach alpha shows us the consistency of the variables, which is approached within the value of its reliability coefficient (!) [22]. The assessment for the Cronbach alpha’s values is done by following the thumb mentioned in [23] as quoted in [24], as listed in Table 1. Meanwhile, Pearson correlation shows the correlation coefficient value (r), which tells us whether the variables correlate each other and the strength of their correlation. This at the end also represents the reliability. Nevertheless, different with the Cronbach alpha, it only can be used to measure two variables. Pearson correlation may show negative or positive value, which the negative value means the negative correlation between two variable (one variable, the other decrease), and vice versa. The assessment for the Pearson correlation‘s values are done by following the thumb as listed in Table 2.

Table 1. Interpretation of Internal consistency from Cronbach alpha

Reliability Coefficient Internal Consistency ! " 0.9 Excellent

0.8 # ! < 0.9 Good 0.7 # ! < 0.8 Acceptable 0.6 # ! < 0.7 Questionable 0.5 # ! < 0.6 Poor ! < 0.5 Unacceptable

Table 2. Interpretation of Correlation coefficient from Pearson Correlation

Pearson Correlation Correlation coefficient -0.09 to 0, or 0 to 0.09 None

-0.3 to -0.1, or 0.1 to 0.3 Small -0.5 to -0.3, or 0.3 to 0.5 Medium -1.0 to -0.5, or 0.5 to 1.0 Strong

3.2.2 Validity

Validity describes the extent to which a measurement accurately represents what it intents to measure [21]. In the qualitative study, the validity was enhanced using the triangulation strategy. Here, we use multi sources of information in the qualitative data collection, which are the secondary sources and interviews. The qualitative validity was also ensured through carefully choosing these data, including the report documents, companies and interviewees. The data taken from the secondary resources were compared and cross case analyzed with the information from the interviews session, and vice versa. Therefore, the bias and interpretation from author’s personal perspectives can be minimized.

The validity of the quantitative study was assured through both external and content validity. The external validity was done by performing a survey to the small random sample of Indonesians who are currently living in Indonesia, which means the study was carried out to the representative environment to which the study are intended to be applied. The content validity was achieved by conducting the pre-study to the reference markets, with the support of literatures review and qualitative data collections. Thus, we could design the questions that have a close connection with theoretical frameworks (UTAUT) by ensuring the aspects within the framework were measured. The questionnaire was made using Indonesian language (Bahasa) instead of English, to avoid any misinterpretation about the context of the questions.

12!

Chapter 4 – The Digital Music Industry This chapter presents the analysis of business models used in the digital music industry. We start with the current landscape of the industry, and then provide discussion of the actors involved, and relations and networks they form using “ARA point of view” as mentioned in section 2.1. In addition, we also evaluate business models used by these actors in the “taxonomy point of view”, where we analyze their distribution, service and revenue models. 4.1. Industrial landscape The digital music services have caused the revolution in the music industry, and it is obvious that the compact disk (CD) has been the largest revenue stream for the industry for more than two decades. This industry has also faced a shift from the physical to the OTT digital distribution era, since the debut of the OTT retail store a decade ago. It extends the music business models and reaches out to more users across the globe. This makes the music as the first media sector that gets full impact from the Internet and technology-empowered forces.

Nowadays, people do not need to purchase CD to enjoy the music, because the available

OTT services allow them to immediately and legally download or stream the music stored in the cloud. According to IFPI6, there has been a consistent rapid increase in the digital music revenue worldwide since 2004 and its percentage over the overall revenue keeps expanding, as shown in Figure 5. This number has far away passed physical format, and made digital penetration in music industry dwarfed all comparable creative industries, except game sector.

Figure 5. Digital music revenue and its percentage share over total revenue worldwide6

However, digital music business internationally is not a homogenous, but consists of

various service model portfolios that run at different speeds. Therefore, there would be no single business development road map applicable in all markets, due to the uniqueness and enormous interaction preferences in the way user in each territory consumes music. This is either due to the differences in economic situation, network connectivity, device penetration, digital development, etc. In addition, the obstacles across countries are also not just the piracy levels and low law enforcements, but also different publishing right mechanisms, user’s preference stages of online payment, etc. 4.2. Actors, Relations, and Networks Before we are able to analyze the network value of the music industry with others, it is important to identify the main actors inside this industry and to get an understanding of their !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!6 Author’s compilation from the data in the IFPI’s Digital Music Report documents, from 2004 to 2012

13!

roles. We then analyze these actors’ relations, from traditional era to the recent phenomena. First, we identify the prominent relations in the digital music industry that have run or are running worldwide. Second, we observe these patterns and model it using ARA point of view [7].

4.2.1 Actors Despite the dramatic changes in the industry’s landscape, the majority of the actors in the music industry still fulfill their traditional roles [48], as mentioned below. 1. Artist The term of artist in the music industry covers several parties, which are the recording singer and/or musician, composer, and songwriter. The recording singer is an artist that delivers the recorded music and/or stages the performance. The composer is an artist that writes the music sheet, while the songwriter is one who writes the music lyrics. Some of the artists in music producing industry might hold several roles. A common example is an artist who both works as a singer and songwriter, which makes him/her hold multi-properties for the copyright works. 2. Publisher and Record label These parties are concerned with the development, protection and valuing the copyright works. The publisher works with the artist based on the publishing contract. It handles the copyright registration, promotes, and monitors the usage of it, and then collects and distributes the royalties of the copyright work as written forms (music sheet, lyrics). The record label works with the artist based on the recording contract, where its main responsibility is in the production and enforcement of the copyright works as recorded forms (sound recording). These actors and the artist act as the content owners who hold the copyright works.

There are currently around 1400 music record labels and publishers worldwide [56], which

can be grouped into two types as below: • Major record labels

These big record labels are often as an umbrella organization (record/music group), and control ~ 75 % of the worldwide music market [48]. They also may consist of several subsidiary companies, such as the music publishers, record manufacturers, and distributors.

• Independent record labels (Indies) The music publisher and record label company, which do not have a corporate backer, might form business relationships with other companies to handle full tasks. However, since in the 1990s, the line between Indies and major record label has begun to blur, as the major record labels sometime also distribute some works of the large Indies as well.

3. Collecting society The collecting society (copyright collective, copyright collecting agency, licensing agency) is an organization that handles management rights on behalf of its members (publisher and artist). These management rights might include selling non-exclusive licenses, enforcing rights and prohibit the use of copyrighted works, collecting royalty payment and distributing it to the content owners, negotiating license fees for public performance and reproduction, etc. In other words, this organization acts as a point of contact for licensing music from its members, to avoid cumbersomeness for the third party to navigate multitude of independent rights holders.

In some countries, there may be also the collecting society that focuses on functions that are

more specific. Some common examples are the collecting society that grants license for performing the work (Performance rights organization / PRO), and collecting society that focuses on publishing management (Reproduction Rights organization / RRO). Indeed, according to the finding in [5], we can classify this actor as the content integrator, as they integrate the copyright works and bridge the content owners, which are mostly the Indies, to the third parties.

14!

4. Content aggregator The content aggregator is an organization that supplies and aggregates the music from the record label (mainly Indies record label) and distributes it to the music retailer and/or service provider. There are aggregators that offer direct-to-fan integration services (merchandise, event management, etc). In addition, some might also supply download services to other aggregators/distributor and act as a music retailer/service provider directly to the end-user. This is consistent with the finding in [5] that also mentioned a same role in mobile service provisioning.

5. Music retailer and service provider These actors are at the end of the supply chain line that holds range of licensed copyright works in various formats for distribution to other or to be sold in their own platform. There are currently ~ 520 legitimate digital music Retailers and Service providers worldwide, ranging in 58 countries [58]. The music retailer provides the music as a product, e.g. CD or downloadable track, or any physical packages that contain music. The music service provider provides the music as a service or any act that convey the music. Some parties provide music as product and service, which makes them as both music retailer and service provider, e.g. Spotify. 4.2.2 Value Network of the Core Actors As same as what have been discussed in the actors’ role previously, the core actors of the music industry’s relation (“Core partnership”) also do not change dramatically. As far as we can identify, their “core network” exist in every models that we discuss further. Meanwhile, the variation happens outside this network “box”. The core actors of the music industry and their relations can be visualized as shown in Figure 6.

Figure 6. Conceptual actors and relations of the music industry’s core network

When a music artist wants to distribute its artwork, it is quite frequent that they have to

deal with separate parties, which are the record label and publisher, through at least two basic contracts to each of them: the Recording contract and Publishing contract respectively. The music artist whose role is a songwriter mainly deals with the publisher, e.g. for the music composition, lyric sheet, etc. The publisher then, on behalf of the songwriter’s, brings that artwork to the record label in order to be used by the singer.

Meanwhile, a music artist that has a role as the singer, mainly deals with the record label,

e.g. for the music recording. The record label, on behalf of the singer, contacts the publisher for the music material that is suitable for this singer. Both the publisher and record label deal with the collecting society, which is mainly settled per country-based, that then acts on their

Other Actor(s)

&'()!*+,-.!/.0'(,1!2)03'(4!5'6!

Publisher

Collecting Society Content Aggregator

Artist

Indies Record label!Major Record label

Revenue stream / sharing line

15!

behalf to collect the royalties for the usage of their artist’s work from the aggregator. The Aggregator acts as a point of contact for the distribution of that work to the music retailer / service provider.

However, following the development in the technology and decreasing in the revenue, there

are several exceptions from the default strategy and working relation mentioned above [63]. This is mainly applied to the Indies environment, as the major record label uses to have a complete capability to fulfill these scoops by themselves. They merge the publisher’s and record label’s traditional roles, and turn as the “music company” by setting up compact all in one deal with the artist (“360-degree” / “multiple-rights” agreement). While in return, the artist gets more shares, compared with having separate deals with different actors (publisher and record label).

4.2.3 Value Network in OTT Distribution In the traditional era, there is only the music retailer that the core actors of the music industry deal with. The music retailer distributes the copyright works to the end-user as a physical product (e.g. phonograph record, cassette, CD, etc). Then since a decade ago, the music business mainly run under the OTT channel, where it is distributed using several packaging models, as discussed later in section 4.2. This then expands and actors from different fields are involved in the ecosystem. This evolution forms diversification to the music industry’s business model, brings music closer to consumer, and becomes the key route to the global mass-market music distribution. 4.2.3.1 “Direct-To-Fan” Model During the early digital music business, the record labels also rolled out their own digital retail channel. They act as the OTT music retail / service provider, under their own or artist’s brand [64][65]. Their offerings range in variances of product (e.g. track and album, merchandise, performance ticket, artist interview, etc) [90]. Some record labels also turn their channels as an advertisement platform for the brand partners. The decrease in the traditional advertising forms, makes the music industry offers a better opportunity for the international brands to reach their end-user. The actors and their relations in this network are visualized in Figure 7.

Figure 7. Actors and relations of “Direct-To-Fan” in the digital music

4.2.3.2 Telecom Actor’s Potential Positions A. “Point-To-Point” Model

The earlier relation variance under this model was struck worldwide around 2001, started first by the collaboration between music industry actors with Internet industry actors. Despite various variances and complex ecosystem, the relation under this model requires “external” actor partners to build their own distribution platform, positions themselves at the end of the Value chain. Meanwhile, core actors of the music industry act as suppliers at the back-end, as same as in the traditional distribution era. The Figure 8 shows the typical actors and their relations in the digital music industry using the point-to-point model.

Core music actors network box! Brand Partner

78)!)29:+,)(!)2;;/*)20!<-2)!

Customer

Revenue stream / sharing line

16!

Figure 8. Typical actors and relations of the “Point-to-Point” partnership in the digital music

As can be seen above, multitude of actor can also involve in the OTT music business, not

just the music industry actors we mentioned in the section 4.2.1 previously. These actors can range from end-user electronic (CE) provider, mobile device OEM and OS provider, telecom operator (CSP), etc. In the variance that involves the CSP, the music industry actors can give them an advantage over established billing structure, commercial network footprint, and mass delivery channel. Music content and service also has a potential as game-changer to the CSP. This can keep them competitive by increasing market share and ARPU, reducing the churn rate, and competing with the OTT, rather than just being the “dump pipe” [90].

In addition, the legal music service is delivered through the client/server and/or assisted-

P2P. This could help to control user’s bandwidth usage, as they have a potential to turn to this legal and easy-to-use service rather than the illegal one [66]. TDC7 and SK Telecom8 are some of the CSPs that were successful to deliver the “Point-To-Point” model. The data shows that TDC has reduced its churn rate approximately 40% in mobile and 60% in fixed broadband since the launching its download music service, YouSee PLAY, in 2008 [67]. By November 2010, TDC has recorded ~250 million downloads [99]. Meanwhile, SK Telecom has also grown its mobile data revenue up to 30%, far ahead of its competitors due to its music streaming service, MelOn9.

The fundamental reason of these successes was due to their value proposition that could not

be defeated by their local competitors. In TDC’s case, the unlimited free download was hugely attractive to the Danish market. It was able to offer this because of its success on setting up agreement with all major record labels and Danish Indies record labels [100]. In SK Telecom’s case, it has a subsidiary record label (Loen Entertainment) that is also the largest record label in South Korea10. This significantly removed the agreement licensing complexity, and reduces the cost to operate the service with updated contents. It should also be noted that the local music contents are much popular in South Korea, compared to the international contents [101].

However, there are prohibitive up-front cost and significant complexity that prevent others

to follow. There are range failures from “Point-To-Point” model implementations, delivered by types of high profile CSPs worldwide. An example is Virgin Media’s direct relation with Universal Music in the UK, French and Australian market back in 2009 [94], because they were unable to secure deal with other three major record labels. On the other hand, BSkyB music service also failed in the UK market due to the inability to reach a large end-user,

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!7 “TDC”, available http://tdc.com/ 8 “SK Telecom”, available http://www.sktelecom.com/ 9 MelOn http://www.sktelecom.com/eng/html/service/WirelessInternet/Music.html 10 LOEN http://www.iloen.com/

CE Provider CSP Mobile Device OEM & OS Provider

Customer

Core music actors network box!

OTT Retailer and Service Provider

78)!)29:+,)(!)2;;/*)20!<-2)! Revenue stream / sharing line

17!

although it had a direct backing from major and Indies record labels, large subscriber based and marketing strength [102].

The success of TDC and SK Telecom are not easy to be replicated by other CSPs in

different countries with their unique circumstances. These factors are due to several reasons [72], as below: • First, there are huge efforts to be put in developing and maintaining digital music service

that has excellent quality, enough affordability and sustainability of music contents, the same as with marketing and branding. During the interview, Lars Roth (TeliaSonera) also mentioned that despite the significant cost that have been invested to obtain license and operate their digital music service, “Telia Musikbutiken”11, they still failed on gaining values they look for (see section 6.4.2)

• Second, music service is a part of VAS, not the core part of CSP’s main capability and

revenue, as they only contribute small proportion to the annual non-voice revenue. Considering these aspects, the CSP might not like to put huge effort and capability to focus to the digital music business, as the OTT music service players would capable to do.

• Third, the CSP also has to struggle to compete with other well-known OTT music service

players, such as Apple with iTunes and iPod12 that give end-user more inferior listening experience. They have the capability to reach consumer globally that then creates a global trend. Their focus to the music business result on a high-quality OTT service, which may lead user to put off their music consumption from CSP’s retail channel.

B. “Point-To-Multipoint” Model The earlier variance of actor’s relation under the “Point-To-Multipoint” model was struck worldwide around 2007. This model involves the actor that acts as a “bridge” that has a settle link between the main music industry actors and several actors in different fields of industry. The typical actors and their relations in this network are visualized in Figure 9, which also might involve actors from other industries that previously have never been seen in the models before, e.g. the automotive industry [108].

Figure 9. Typical actors and relations of the “Point-to-Multipoint” partnership in the digital

music

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!11 “Frågor och svar för Musikbutiken”, available http://www.telia.se/privat/produkter_tjanster/noje/musik/musikbutiken/faq 12 “Apple - Play music and more on iPod.”, available http://www.apple.com/ipod/

Vehicle Provider CE Provider

OTT Retailer and Service Provider

CSP Mobile Device OEM & OS Provider

Core music actors network box!

78)!)29:+,)(!)2;;/*)20!<-2)! Revenue stream / sharing line

Customer

18!

In the collaboration variance that involves CSP, this model also has a potential to gain market share, increasing the smart phone selling with bundling data plan, and reducing churn rate. This variance can be used by bundling it with the CSP’s broadband and telephony subscription packages: hard bundling (e.g. Deezer – Orange in France), soft bundling, free trial (e.g. Spotify – 3 in UK), billing integration (e.g. Spotify – TeliaSonera in Sweden and Finland), etc. Collaborating with a well-known service provider would also bring extraordinary impact on CSP’s core business, as it could expand their values beyond the traditional “pipe” provider.

The CSP also can take advantage of their end-user based and network by cross selling the

music service with their existing products. However, it more makes sense for the CSP to use this model, instead of create it by them self. This is because this offers opportunity to the similarly compelling service, but with opportunity to gain it quickly, in a simpler way, and at a reasonable cost. Using this model, the CSP can share or even remove the effort on marketing and branding to deliver the service to the market. As also mentioned by Lars Roth (TeliaSonera) and Andreas (Spotify) during the interview, this kind of partnership gives the CSP an already established platform and existed license agreements with the content owners.

We can compare this with the fails that have been faced by the CSPs worldwide when

dealing with the licensing issues, as discussed previously. In return, the OTT retailer / service provider can expand its end-user based efficiently on a scaled size. Its tie with the big players gives them a significant advantage, as they are mostly new and small company. This offers a strong position to compete with competitors and deal for the license with the content owners, as also informed by Lars Roth (TeliaSonera) regarding TeliaSonera-Spotify partnership. Swedish market is an example: according to IFPI, 56 % of the consumer confirmed that the streaming service was their main reason to stop music piracy activities [66]. The local music industry saw 10% increasing revenue for the first time since 2000 and the file-sharing activities were reduced by 60%.

4.3. Service, Distribution, and Revenue Models The economics of the music value chain has faced several dramatic changes since two decades ago, moved from the physical to OTT era, and revenue from the OTT channel is expected to overtake the physical channel by 2015 [68]. The OTT distribution, with a support from the network technology, also changes the way the music served the end-user [90]. There are possibilities to consume music content in different ways, as it is delivered on top of diverse devices and platforms. The current business model used in the OTT music can be mapped into different combination of services, distributions, and revenues, as shown in Table 3 and Table 4.

Table 3. Revenue models in the digital music

Revenue model Note One-off

Payment / Retail The end-user buys the music content, and it might be downloaded or stored in the digital locker, e.g. iTunes’s revenue model.

Subscription The end-user pays for regular period of time to access (stream) /get (download) limited/unlimited music contents (based on amount of time or content), e.g. Deezer and Spotify revenue model (premium/unlimited subscription)

Subsidized The end-user does not pay for the music content consumption and other revenue channels support the business, mainly the advertisement fee, either as voice, picture (banner), or placed at the intervals within the content delivery, e.g. Spotify revenue model (free subscription).

19!

Table 4. Service and Distribution models in the digital music Service model Distribution model Revenue model

Personalization Ring tone (monophonic, polyphonic, real tone, user-generated tone), Ring-back tone / RBT

One-off payment, Subscription

Ownership DRM à-la-carte, Free-DRM à-la-carte One-off payment, Subscription, Subsidized

Access On-demand streaming, Internet radio streaming

Subscription, Subsidized

Integrated Combination of 2-3 service models above, mostly “Ownership” and “Access” model

Combination of revenue generation models above

1. Personalization service models Personalization service models were initially launched by Nokia, as one of its mobile device features, and marked the start of mobile music era [68]. In those early years, the monophonic ring tone has ever contributed as the biggest revenue for MNOs, where the payment was made either using premium SMS or telephony (interactive voice response). However, in line with the expansion penetration of advanced mobile device, the marginal revenue share from monophonic ring tone service decreased and until then it was replaced by the RBT service.

The RBT service was firstly introduced in South Korea [90]. It mainly relies on the network

rather than mobile device, as it is installed in the voice system of MNO’s core network. Because of its “close” delivery platform, this service does not relate to the piracy issue. However, it has a technical make up that causes the RBT costly to implement and relies to the regional marketability that makes this service very culturally specific. Thus, only MNO that is often motivated to promote the service, while the third parties mostly involve at the back-end.

2. Ownership service models The “ownership” service or à-la-carte download has been and currently the dominant digital business model. However, the lack of interoperability between services and devices has been a barrier to business development. Significant moves have been done to solve this, such example is DRM-free service that allows end-user to buy and download music, then play it in any devices [90]. It includes “cloud feature” to store the track that has been bought in the cloud.

The DRM-free model is currently still as the key in the digital music business. However,

there is still an issue in the interoperability of the content. In addition, the price competitiveness is highly important in this model, and piracy issues still become major barrier to the business growth. In the CSP’s point of interest, this service is only able to influence a fewer bond and loyalty with the user. This happens because user can just download the music content and then chooses on his/her own to experience it offline and/or elsewhere using other devices.

3. Access service models When seeing the business in the network point of view, the CSP can build an enormous strategic alliance with the music industry actors by using the streaming services. This service model has more ability to give an impact to the CSP’s key business metrics and compete with their competitors, compared with the download services, due to several reasons. • First, music is inherently emotional for people and if it is delivered using a streaming

service, the quality of experience becomes a matter. It potentially creates a genuine bond between user and service (“stickiness” aspect). This is a key to reduce churn rate and gain market share. The service makes end-user stay with their CSP, unless they would lose access and/or unique offering.

• Second, the revenue from streaming service is generated via repeatable billing cycles, so

provider keeps the revenue from both data traffic and service usage, as end-user naturally

20!

accepts the concept of bundling with a scheduled-billing (“recur” aspect). Although revenue for a stream is lower compared to a download transaction, the cumulative revenue might be higher, because each track is listened multiple times, that each generates revenue.

• Third, streaming service is delivered in a high-end mobile device and can be bundled

with a data plan, which is a significant potential for CSP to gain more revenue (“up-selling” aspect). The CSP can compete not mainly based on the low price of the offerings, but due to the attractiveness of the bundling.

• Fourth, streaming service is the “access type” consumption model. Therefore, it is natural

for end-user to get access into it across different devices (“cross-promotional” aspect). This could expand CSP’s position in supply chains, e.g. PC, mobile phone, digital TV/Audio system, etc.

In addition, the global market demand is also giving signs that the digital music business is

shifting to monetizing the “access” to the music. Ovum predicted the revenue from digital music subscription service would increase to more than 60% of its annual growth by 2015, due to rapid increase of paid-subscription to the streaming service [71]. It has been forecasted that the “streaming” service would likely become the main driver in the music distribution channel for the coming years. These revenue shares are shown in graph in Figure 10.

Figure 10. Revenue stream share from the digital music [68]

21!

Chapter 5 – The Digital TV/Video Industry This chapter presents the analysis of business models used in the digital TV/Video industry. We start with the current landscape of the industry, and then provide discussion of the actors involved, and relations and networks they form using “ARA point of view” as mentioned in section 2.1. In addition, we also evaluate business models used by these actors in the “taxonomy point of view”, where we analyze their distribution, service and revenue models. 5.1. Industrial landscape Video is a key driver for exploding mobile data traffic, bundled through ranges of product. According to Allot, video streaming was the largest service accessed worldwide that occupied 42 % global mobile bandwidth, derived from the booming of smart phone and tablet. Youtube was responsible for 57 % global video streaming traffic by Q2 2011 [78], where over 600 million views / day are mobile13. While, Neflix occupies ~ 95 % downstream traffic in Canada and ~ 20 % in US during prime time, from only 1.8 % subscriber.

In addition, video streaming business in mobile devices was also forecasted to catch other

businesses in the creative industry, especially game and music. However, video is not representing strong revenue for the telecom operator for its own right. This increasing media stream traffic also triggers demand for a better quality of the content accessed. This creates challenge to manage the traffic, and OTT players also bypass telecom operator’s substantial role in the digital TV/Video value chain. Thus, the telecom operator, especially MNO, poses threat from this, in line with their investment for mobile broadband network, as the digital TV/Video business grows across multiple and connected devices [92].

5.2. Actors, Relations, and Networks The same as the music Industry, before we are able to analyze the network value of the broadcasting industry with others, it is important to identify the main actors inside this industry and to get an understanding of their roles. We then analyze these actors’ relations, from traditional era to the recent phenomena. First, we identify the prominent relations in the digital TV/Video industry that have run or are running worldwide. Second, we observe these patterns and model it using ARA point of view [7].

5.2.1 Actors Here we also see a consistent pattern as similar as what we have observed in the digital music industry. Despite the dramatic changes in the broadcasting industry landscape, the majority of the actors inside still fulfill their traditional roles, as below: 1. Content Owner The Studios and TV broadcasters are match to this category. They own movie (video) and TV content, and traditionally distribute it directly to the widest audience and are compensated for the usage of that. Several examples of actors in this category are Movie-based content owner (e.g. HBO, Disney, Sony Pictures / Crackle, Universal, Discovery Channel), News-based content owner (e.g. BBC / Youview / iPlayer, ESPN, ABC, NFL, Showtime), and Sport-based content owner (NBA, Major League baseball), etc. 2. Collecting Society As same as in the digital music industry, there is also a same role of collecting society. This organization handles the management rights for digital TV/Video content, and usually operates per country based. In other words, it acts as a point of contact to acquire licenses and avoid cumbersomeness for third party to navigate multitude of rights holders. The work in [5] classifies this kind actor as content integrator in mobile service provisioning.

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!13 ”Product Metrics”, available http://www.youtube.com/t/press_statistics

22!

3. Content Aggregator The need for a content aggregator (content distribute) comes to fill the gap between the retailer and/or provider and various content owners. This party obtains range of legitimate intellectual property rights and aggregates the licenses to distribute it. In other words, this actor provides a one-stop-shopping service, which is the established negotiation with the studio and broadcasting firms and assures the availability of the content [29]. This is consistent with the finding in [5] that also mentions similar roles in mobile service provisioning.

4. TV/Video retail and service provider The service provider’s role was previously held by CSP that provides variety of TV/Video contents to their subscriber via their own service brand and delivered through their closed network infrastructure. This service is often bundled with other service features (e.g. triple play or quadruple play bundle). Some examples of actors in this category are the cable provider, satellite provider, and telecom operator. Nevertheless, nowadays, any parties from range of industries can also hold this role, e.g. Internet player, mobile device OEM and OS provider, CE provider, etc. 5.2.2 Value Network of the Core Actors The relation between actors inside TV/Video industry in the digital/OTT era is visualized in Figure 11. As what have been discussed in the actors’ role previously, the core actor’s relation (the “Core partnership”) in the broadcasting industry also does not change dramatically. Here, the copyright works belongs to the Studio/TV broadcasting. Therefore, artist’ roles are less relevant to be discussed, because they do not hold the copyright. As far as we can identify, this “Core network” exist in every models that we will discuss further, while the variation happens outside this network “box”.

Figure 11. Actors and relations in the TV/Video industry’s core network

In the case of movie content, the studio may also grants the licenses for distribution to the

collecting society. Märta Rydbeck (digital broadcasting expert) during the interview stated this as below,

“We (TV broadcaster) pay to [name of a collecting society] for a right that can not be cleared, for a right that is transferred by the owner to an exclusive organization that can negotiate with TV channel. So in order to distribute TV in Sweden you always have to have agreement both with the right owner of the content, but you also need the [name of a collecting society]”. In addition, content aggregator determines how many copies that have to be made, and do

the screening process to the prospective buyers, e.g. Cinema Theater or other content

Studio

Core broadcasting actors’ network box

TV Broadcaster

Revenue stream line

Other Actor(s)

Content Aggregator

Collecting Society

23!

aggregators. Put differently, the studio company is behind the wall. However, this is mostly not the case with the TV broadcasting industry, as they directly set up the engagement with user.

5.2.3 Value Network in OTT Distribution Previously in the traditional era, user only got access to the TV/Video content through TV broadcaster or cinema theater. Here, the contents are defined by specific schedule and user accepted the concept to adapt to it, because the service is offered by creating demand of the same content for many end-users at specific uniform time. Then, the retailer expanded these channels and rental (physical) outlet/kiosk in late 80s/90s, where the copyright work is bundled as a physical product (e.g. DVD, VHS, Blu-Ray, etc.)

The “virtual” digital distribution channel was started by the cable and satellite TV provider,

and then followed by the telecom operator through their closed network infrastructure, as discussed later in section 5.2. In addition, up until recently, there was rarely direct competition between these parties, as their operation area was different. In that era, the value chain was well established, where the CSPs act as gatekeepers to their own end-users, and there was geographical competition that borders their service range.

However, the technological advancement and change in the regulation across markets have

altered the picture, made each cross other’s primary business area. Here, the competition comes from any angles and through any devices. The cable provider offers similar range of services: obtaining voice, data, and existing TV/video service in a bundle. The satellite provider might also collaborate with ISP to provide cross-selling services. While these different CSPs are racing in the triple-play business and not just facing intense competition with their counterpart core competitors, the OTT TV/Video players also enter the market.

This brings even more pressure to the existing actors and triggers a second change in the

digital TV/video service competition landscape. The launching of Youtube in 2005 marks new generation of the distribution model, and makes the barrier to provide digital TV/video to the end-user shrink. Now, practically everybody can be a digital TV/Video service provider. The content owner has more choices to deliver and monetize their copyright work across different platforms. Meanwhile, user has more freedom on consuming the TV/Video contents. 5.2.3.1 “Direct-To-Fan” Model

Figure 12. Typical actors and relations of the “Direct-to-Fan” partnership in the digital

TV/Video Following the rising of OTT access through range of connected devices, the studio and TV broadcaster also seek opportunity for direct access to their end-users. There are two paths on the implementation, by either collaborating with the CE, Mobile device OEM and OS vendor and/or creating their own platform that enables TV/Video content delivery. Thus, the studio and TV broadcaster provide OTT TV/Video service in this model. The typical actors and their

Brand Partner

Customer

CE Provider Mobile Device OEM & OS Provider

Core broadcasting actors network box!

78)!)29:+,)(!)2;;/*)20!<-2)! Revenue stream / sharing line

24!

relations in this network are visualized in Figure 12. Some examples of this actor are SVT play14 in Sweden, or several TV broadcasters in the UK that launch open platform called Canvas15, which enables any broadcasting actors to use it as delivery platform. In these cases, the TV broadcaster acts as OTT TV/Video service provider and provides its own billing relationship with the viewer. Märta Rydbeck addressed this strategy as below,

“(Up until recently) a lot of distribution (channels) are by the operator, so obviously the operators are not so happy with [name of a TV channel] extending beyond its manage network, going over the top… And [name of a OTT TV service] where [name of a TV channel] builds its own customer is huge threat for the (telecom) operator. They obviously want that content in their own network. We said ‘if that is in your platform, we do not get any of the advertising (revenue)… It needs to be played in our service, so can calculate on the advertising, whether we have free channel or we have billing relationship with our own customer’”.

From the statement in the interview above, it is revealed there is still a main problem that

the TV broadcasters are facing when delivering the content in multi-screen strategy beyond the traditional Pay-TV distribution channel (e.g. mobile TV). Therefore, it is not surprising that the TV broadcaster, as a content owner, also attempts to reach directly their viewer directly in the digital distribution, by delivering their own online channel, by passes telecom operator’s traditional role that exists in Pay TV ecosystem. Märta Rydbeck called this as called as “advertising dilemma”, as she explained as below,

“(For TV channel) it (mobile channel) is of course revenue stream, but it does not bring more value to watch it on the mobile (mobile TV) rather than set up box. Actually currently it bring less value, because the TV channels have not sort it out how to deal with advertising on the mobile yet… because, if I watch [name of a TV channel] here (mobile phone), and I watch an advertisement that run on [name of a TV channel], It does not get calculated as a view, I do not get calculated as viewer. Because there is no measurement system today that takes into account the (number of) advertisement that is broadcasted here (mobile phone)… all the measurements set up today are at home (set-top-box). So it is big problem if you look at the increase of both on demand viewing or linier type of viewing on mobile today for TV channel. So people start to switch at their home, to watching only on their mobile and we do not sort this out… The TV channels are facing big problem”.

5.2.3.2 Telecom Actor’s Potential Positions Despite the opportunities offered in OTT distribution, not all studio and TV broadcaster attempt to shift to the “Direct-To-Fan” model. Some still see the advantages of establishing delivery service through collaborating with other actors. Telecom operator have several strengths this circumstance, e.g. billing and direct relationship with the end-user, capability to reach mass-market through infrastructure and marketing campaign, ability to offer end-to-end quality of the content, investment to acquire desirable content, and flexibility to leverage the service through a range of bundling and revenue streams. There are different strategies and positioning that telecom operator can take in OTT TV/Video business, as discussed below:

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!14 “SVT play”, available http://www.svtplay.se/ 15 “You view”, available http://www.youview.com/

25!

A. “Point-To-Point” Model At the beginning of the digital distribution, the only value chain exists was the ‘Pay-TV’ where the CSP occupied the market space. As in the traditional era, user only had a choice of what channel to subscribe, but not the content itself. The revenue is split between the CSP, content owners, and/or content aggregator [47]. Each typically receives 40-35%, 60%, and 5% respectively from total revenue share [92]. Then, multitude of actors from different industry also entered digital TV/Video market and distributed content directly to end-user with different kind of service models. Figure 13 shows the typical actors and their relations in the digital TV/Video industry using the point-to-point model.

Figure 13. Typical actors and relations of the “Point-to-Point” partnership in the digital

TV/Video

This is a same pattern we have seen in the music business as discussed previously. These actors bypass CSP’s role using technological advance spurred by the increase of broadband and mobile devices penetration. This was previously triggered by the content aggregator, e.g. BBC, HBO, etc and then leveraged by other Internet players, e.g. Neflix, Hulu, etc. CE provider, especially TV manufacture, also took a role in this ecosystem, e.g. Panasonic (Vieracast), Philips (Net TV), etc. Here, the revenue split is approximately 60/40 respectively between content owner and OTT player [47]. This market generates much smaller revenue, ~8 billion globally, and it currently offers a relatively lower quality and availability compared to traditional digital services offered by CSP.

However, this makes CSP as no longer sole provider in digital TV/Video business, as the

market is growing fast and the quality issues are addressed through different approaches. This issue becomes less important in following years, compared to the advantage over availability and simplicity. Soon, the increasing quality of the OTT service and the better network connectivity, which is built by the CSP to CSP to strengthen its own core business, will result to the disintermediation of the CSP’s value chain as a traditional digital TV/Video gateway to the end-user. Even though the CSPs’ digital TV/Video services are still not fully cannibalized yet, this definitely brings challenge, as it potentially becomes mainstream in the future.

There also several weak factors and challenges in this ecosystem that do not meet with the

CSP’s capability. Some examples are of these weaknesses are their closed and vertical system demands a large investment, but creates slower deployment life cycle. Their mass-market focus makes them slower to adapt to the emerging end-user’s demand and competitors that come from other industries. The effort for setting up and maintaining the license and delivering the service across multiple devices also cost time and money. This might not be worth enough for the CSP, considering the unproven business model and small revenue from this stream, compared with their core businesses.

OTT Retailer & Service Provider

CE Provider CSP Mobile Device OEM & OS Provider

Customer

Core broadcasting actors network box

The end-user enggament line Revenue stream / sharing line

26!

B. “Point-To-Multipoint” Model The content owners in this model still play behind the scene, but there is an actor that acts as a bridge between these content owners and other actors outside the core media industry. This actor typically also reaches its own end-user, thus it takes a central position in the value network. In other words, the actors form the horizontal strategic alliance in the ecosystem [18], as they collaborate with other actor that was previously competitor [11][12]. Figure 14 shows the typical actors and their relations in the digital TV/Video industry using the point-to-multipoint model, where the Internet player becomes the “bridge” actor, e.g. Google (OTT player) partnership with Sony (CE Provider), Voddler (OTT player) partnership with Tele2 (CSP), etc.

Figure 14. Typical actors and relations of the “Point-to-Multipoint” partnership in the digital

TV/Video

The telecom operator might consider collaborating with the OTT TV/video player, as it significantly reduces/remove the cost and complexity to roll out the OTT service, and telecom operator can still get some percentage from revenue generate. This is suitable because even though the role of mobile device in the OTT TV/video business is increasingly important, it is still only holding a small part of the whole picture. The OTT TV/video consumption is different from the OTT music consumption, which is naturally suitable for mobility feature. Nevertheless, this is not always the case with the TV/Video consumption, as end-user naturally wants to watch a video on the best available screen. The role of mobile device in the video business would still be a “second screen” device: enhancing experience, helping discovery and facilitating social activity.

This is supported by the research done by Ofcom to the mobile phone owner in the UK

about the activities done in the smart phone [76]. They found that despite the increase mobile data services usage and range of smart phone in the market, TV remains among the less consumed features. Therefore, telecom operator better to use this OTT TV/Video to complements the existing Pay-TV service. Telecom operator may take position in the marketing, integrated branding, product bundling, or enhancing the efficiency in the content delivery (resell CDN) [82][83]. It has a better technical capability compared with other actors (CDN provider), as they only able to reach the edge of CSP’s network, results on no guarantee in the final stage delivery of the content. In contrast, telecom operator’s CDN able to provide deep caching as the content is transmitted through their network infrastructure, by avoiding as much as possible upstream point of contention.

CE Provider

OTT Retailer and Service Provider

CSP Mobile Device OEM & OS Provider

Core broadcasting actors network box!

78)!)29:+,)(!)2;;/*)20!<-2)! Revenue stream / sharing line

Customer

27!

5.3. Service, Distribution, and Revenue Models The business model using the digital distribution channel is still unclear, and the actors in

the ecosystem are still testing a wide variety of business models. OTT TV/Video service means the content are delivered over open network/Internet. Different with IP TV and Mobile TV, Internet TV/Video provider cannot offer full QoS control as it uses public/unmanaged network. The current business model of OTT TV/Video can be mapped into different combination of services, distributions, and revenues, as shown in Table 5 and Table 6.

Table 5. Revenue models in the digital TV/Video

Revenue model Note Pay-per-view (PPV)/Rental

The end-user pays for the ability to view the TV/video content over period of time, e.g. iTunes rental, Amazon watch instantly, Vudu, Voddler, etc.

One-off Payment/Electronic

Sell-Through (EST)/Retail

End-user buy the TV/video content, and it might be stored in digital locker, e.g. Amazon VoD, iTunes, etc.

Subscription The end-user pays for regular period of time for the TV/video content consumption, e.g. Neflix, News Corporation (Hulu Plus), LoveFilm, etc.

Subsided End-user does not pay for the TV/video content consumption and the business is supported by other revenue channel, mainly advertisements fee, either as non-video advertisement (banner), pre-roll or placed at intervals within content delivery, e.g. YouTube, Hulu, etc.

Table 6. Service and Distribution models in the digital TV/Video

Service model Distribution model Revenue model Ownership Content download, physical

purchasing Rental, retail

Access Streaming (Live/TV, on-demand, catch up TV, end-user-generated)

Rental, subscription, retail, Subsided

Integrated Combination Combination

As can be seen above, the actors in OTT TV/Video are also still found to “mimicking” the traditional distribution model through physical bundle distribution. There is problem that might exist in download distribution model, e.g. the network might not be able to provide good delivery speed so user has to wait to download the content. Thus, it is not surprising that streaming is mostly a primary delivery model used [81], as it offers an instantly satisfying form of access compared to download, especially for long-form contents and in mobile device. There are also a different usage patterns, where the ‘on-demand’ or personalization aspect is prominent, e.g. sport competition or concert, as well as a tendency toward less viewing-time, especially in the viewing through mobile devices.

1. PPV and EST revenue model These models are classified in [6] as Utility model (pay-as-you-go approach). The availability of high interest and new release video contents are extremely important in EST model, rather than TV programs and older release contents [80]. In contrast, the availability of the high interest video contents are more important in PPV model rather than the release date aspect, as the transaction is driven more by interest in viewing content multiple times [80]. Here, user might get the copyright works as a physical bundle (e.g. VHS/DVD/Blue-ray disc, etc), beside the download or streaming based content. Some players have also offered digital-locker feature, where s/he can store the content has been bought/rented in the cloud and access it anytime later.

28!

2. Subscription and Subsided revenue model Both of these models are mentioned in [6], as ones of the generic forms of e-business model. The key success factor in the subscription model is different with both PPV and EST model. Here, the size of content library turns as the most important factor to attract subscriber [80]. Although some contents have a lasting popularity, the continuation of updated library and rotating the popular titles are necessary to help retain viewing interest. In the subsided based on advertisement-supported model, there is a mechanism called as CPM (cost per mile), where the fee is either based on the number of ads viewed or end-user reached by that ads.

Indeed, that is why the OTT TV/Video service providers that use subsided model as one of

its revenue streams, needs to get a large user based on as many as possible. Some OTT TV/Video service providers also offer pinpoint delivery based feature to its brand partner, either based on user’s location or favorable genres, in order to add value to advertisement delivery. It is also possible to use the placement model, where the advertisement is built as an integral part of the content, whether designed as subtly (used as product showed in content), or overtly (used as main content, e.g. appear in talk show as a product review).

Figure 15. Revenue share from the digital TV/Video16 [79]

However, despite the variety of these revenue models, the trend shows that the revenue

stream generations from the subsidized-based would potentially overtake the “paid-by-end-user” models in the OTT TV/Video’s revenue source worldwide, as shown in

Figure 15. This model is predicted to dominating this share in a long run, due to the better availability and quality of the network infrastructure across the globe, and less significant factor of ownership model (using EST revenue generation). Märta Rydbeck also mentioned the importance of revenue stream from advertising channel, as she stated,

“… The local advertising is very very profitable for [name of a TV channel] and it is a growing business. The local advertising market for TV is huge in Sweden, and the only competitor in term of local advertising today is the local newspaper or local direct mailing, and so on... It (local advertising) is considered more valuable, more expensive, and more (delivery) times... There are lots of TV channels that have regional version; they are facing the same (advertising dilemma) problem”.

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!16 In this figure, the revenue share from placement model is separated with the advertisement model

29!

Chapter 6 – Spotify Business Model This chapter presents an “instant point of view” analysis to the OTT media player’s business model, using theoretical framework mentioned in section 3.1.3. Spotify is an OTT music service provider that was founded in Sweden and commercially released in 2008. We also get in deep into the “cross-firm collaboration”, and as this study focuses to the business model in mobile broadband, we mainly cover Spotify’s partnership with the telecom industry actors (MNO).

6.1. Value Proposition Spotify actually does not offer new type of business in the music industry, as there is already numerous music retails and music service providers out there that also have the same main offerings. However, it has final touches apart from those main offers that make it more successful than the other competitors do. Unlike others that form the business only based on the download-purchase, Spotify is not just selling a product (music retailer selling unit), but also provide the streaming service. It closely resembles the radio, iTunes, and original Napster model, by recognizing the validity of the copyright and conducts its business accordingly. It covers 15 million licensed music tracks globally (size and availability varies in each country).

Convergence of the service across ranges of device is one of the major themes in the digital

music business. Spotify application is written in several programming languages and available at desktop clients17, mobile devices, tablet PC, and consumer electronics "=>%[96][97]. The service is delivered as a downloaded client application with proprietary protocol. It has much control over the network protocol as opposed to a web application18. It also has a device integration feature between these devices so user can synchronize music tracks between these devices, access it in “offline-mode”, and/or download it. Spotify uses different audio bit rates of OggVorbis and offers feature to the premium users to access the music tracks with higher quality [103][104].

Spotify works differently compared with other similar services as it relies on several ways

to make the music delivered seamlessly to the client, which is very important in the real-time service [44]. It uses TCP-based communication, which guarantees that each bit is correctly delivered. However, TCP is accomplished with timeout and retries mechanism, which this creates delay. This is minimized by Spotify using buffering in the client side [45][46]. It implements this with hybrid peer-assisted streaming per track (P2P), not per torrent as the Bit Torrent19 [40]. Here, the peer discovery strategy is the main concern, as the data is small and short. This is different with the P2P streaming for large data (e.g. Movie), where the concern is the download strategy.

To accomplish this, it uses a combination of three data sources: CDN/core streaming

servers, P2P and cache, where the data comes from each is 8.8%, 35.8%, and 55.4% respectively [44]. Using this, Spotify recommends 256 kbps bandwidth and free space for caching [105]. Spotify implements dual mechanisms to handle the first request for the track, e.g. when the user search for a new track and add it to the play list. In this case, the client asks to the tracker (CDN server), which will be replied with random OTT peers that recently played the track (tracker-based peer search). It also sends request to the neighbors in the overlay network if there are any peers between the 1-2 distances that have the track in the cache.

This dual mechanism is done to provide a better availability. If the tracker is inoperable, the

client still can rely to the overlay search. In addition, if the client is disconnected from the !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!17 “Download”, available http://www.spotify.com/se/download 18 “Spotify Web API” https://developer.spotify.com/ 19 “Bit Torrent - Delivering the World's Content”, available http://www.bittorrent.com/

30!

overlay network, it can find other peers from the tracker’s information. Every requested track is saved in the local cache, where the default policy is to use 10% free space (capped 10GB in the desktop version). Before the played track ends, the application starts pre-fetching the next track from the cache. This cache exists for 30 days and Spotify monetizes feature to access it for “offline-mode” usage20. This is what the combination of P2P and local cache is intended to: to make the streams data sent between the peers, and prevents the server and the network become bottleneck [43].

By using a combination of these data sources, the client’s cache is not just used to prevent

client to re-download every time it plays the track. Nevertheless, it also provides tracks that are shared with other peers in the overlay network. Spotify spreads the sharing effort for music data among the users and keeps the Uplink (UL) speeds slower than Downlink (DL) speeds. However, the ration of this UL/DL remains unknown and it does not enforce fairness (e.g. tit-for-tat) [40]. This actually raises issues, especially for the user that has a limited bandwidth, and peak restrictions or expensive data plan subscription where the UL is counted towards the monthly subscription allowance.

6.2. Revenue Generation and Market Segment Spotify cost structure are mainly to address the licensing cost of its music delivery to the core music industry actors, apart from the development and maintenance cost. These costs are funded through three revenue generation streams, as discussed below: 1. Streaming subscription

Since its commercial launching in 2008, Spotify has made several price scheme changes. It offers three tiers for the streaming service since then, where two are paid-based (“Spotify Premium” and “Spotify Unlimited”) [106]. It also recently confirmed the plan to launching the premium family plan, which is an incentive offering for members of a same household to get access on multiple based [107].

2. Download purchase Apart from the streaming service, Spotify also builds its own music store. It previously worked with 7digital to provide a-la-carte pay-for-track. Later on, it ended this relation and offered a pay-to-own option [109][110]. In addition, unlike most of other digital music stores, this integration allows user to re-download the purchased tracks several times.

3. Subsided Spotify offers a free subscription plan (“Spotify Free”) that is supported based on audio and graphic advertisements and sets relation with several brands21. Using the information that it has about its users, Spotify is able to generate value to deliver this advertisement in specific targeting pinpoints delivery, also the detail tracking and reporting for that advertisement delivery [111].

From the discussions above, it can be concluded that these revenue streams are build based

on two pricing models: • The “Freemium” subscription model, which is implemented to the streaming subscription

bundles: the “Subsided advertising-based” (“Spotify Free”) and “Premium subscription” bundles (“Spotify Unlimited” and “Spotify Premium”)

• The Retail model, which is implemented to its download for purchase feature

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!20 “Offline mode”, available http://www.spotify.com/se/about/features/offline-mode/ 21 “Advertising”, available http://www.spotify.com/se/work-with-us/advertisers/

31!

6.3. Competitive Strategy Spotify implements its pricing models using the “demand-based pricing” strategy, as it adjusts the price between countries by following the perceived value as the central element to set the price. Spotify also uses geographic segmentation, as what Facebook did in its early years, but Spotify expands its availability per country based. Its first commercial launching was in Sweden, and since then it has been available in 16 countries, while the expansion strategy that Spotify mainly targeted is the early adopter, which potentially drives benefits in a long-term engagement.

Although, there is no official information of the subscriber distribution per country-based,

Spotify currently has 10 million registered users, where 3 million of it is using paid-subscriptions [56]. This makes it as the world’s largest paid-based streaming music provider and the second largest source of digital music revenue in Europe. This is actually the core aim behind the “Freemium” revenue model: to attract end-users with the free offering before migrating them to the paid subscription. The introduction of mobile feature, along with the time-access limit, similar-track limit, offline access, etc is some of the key factors that significantly boost the paid-subscriptions, as shown in Figure 16.

Figure 16. Spotify’s user based by service type in Europe, Sept 2008 – March 2011 [73]

Figure 17. Spotify monthly revenue by service type in Europe, Sept 2008 – March 2011 [73]

32!

Spotify also uses feature differentiation in some markets, e.g. the implementation of “Private beta” launching model based on invitation. It also keeps changing the offering to the bundling tiers, which might be different between countries, e.g. radio service is available for the tree-based subscription users in the US, or there is no 5-track cap for the free-based subscription users in several European countries. All these strategies are not just mainly due to the copyright settlements that work different between countries, but also to make the product become exclusive and attractive to the early adopter users. Spotify’s revenue is analyzed to come ~ 23 % from advertisements and ~ 77 % from paid-subscriptions [73], as shown in

Figure 17 above. 6.4. Value Network In order to legalize its business above, Spotify mainly makes deal with actors in the music industry. Apart from that, it also has several strategic alliances with actors from different industry fields to build the delivery channels that expand and make its service closer to users. These actors range from the consumer electronic, Internet and telecommunication industry. The prominent actors and their relation with Spotify be visualized as shown in Figure 18 and discussed in the sections below.

Figure 18. Actors and relations in Spotify point-to-multipoint partnership

6.4.1 Relation with the core actors of the music industry Spotify opens its service channel for independent/self-published artist and small record label that does not have an existing delivery partner or platform22. They can make deliveries through several aggregators or collecting societies that already have an agreement and delivery process in place with Spotify23. Therefore, it can be concluded that in order to get license to deliver the music to end-user, Spotify deals with the record label, music publisher, collecting society, and/or aggregator, not with the artist.

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!22 “Artist Page”, available http://www.spotify.com/se/work-with-us/record labels-and-artists/artist-page/ and “Record labels and artists”, available http://www.spotify.com/se/work-with-us/record labels-and-artists/ 23 “Record label and Aggregator Page”, available http://www.spotify.com/se/work-with-us/record labels-and-artists/record label-page/

CE Provider

OTT Retailer & Service Provider

CSP

Customer

?08)(!,0(/0);-.!/<<-/2.)!<-2)!

Brand Partner Other OTT Player

Core music actors’ network box!

78)!)29:+,)(!)2;;/*)20!<-2)!Revenue stream / sharing line

33!

There is no official information that contains all of Spotify’s music industry partners. However, its agreements have reached the entire major record labels, as well as Indies record labels, international and local collecting societies, and music aggregators. Some tracks may not be available or disappear on Spotify because of several license reasons24. These music industry actors provide the music and metadata to Spotify, which will be encoded and uploaded to Spotify server, and integrated with Spotify delivery platforms. This means, Spotify tracks all the music played and provides in-depth reporting to all participating parties25.

The price and payment to these music industry actors are calculated based on the function

of the total number of streams and the total revenue of that particular month, result of the payout per stream. Thus, when the total stream grows more than the total revenue, the payment per stream might be less. According to Andreas Liffgarden (Spotify), the detail licensing arrangements between Spotify and the content owners are not published, as it is part of the Non Disclosure Agreement/NDA. The licensing price might also be different between these agreements, and some of the deals might have discount arrangements.

6.4.2 Relation with the Telecom actors Apart from the collaboration above, Spotify also sets up an alliance with several CSPs. Spotify started its first agreement with TeliaSonera in Sweden in 2009. It then expanded it to the UK and Dutch market with Hutchison Whampoa (3UK) and KPN, respectively. It also collaborated with the MVNOs, which are Virgin Media UK, Hi Netherland, and MTV Mobile Europe. In the sections below, we focus on analyzing the collaboration between TeliaSonera and Spotify, as one of Spotify’s biggest deals.

A. The Alliance Model TeliaSonera is a company formed from a merger between Swedish incumbent, Telia and Finish incumbent, Sonera at 200226. TeliaSonera telecommunication services operation is splinted into three business areas (Mobility Services, Broadband Services, and Eurasia), and operates in range of countries in Nordic, Baltic, and Eurasia region. Spotify and TeliaSonera started an agreement in 2009, which was a 2-year exclusive deal that makes TeliaSonera a sole exclusive telecom operator in Sweden to use Spotify brand within its products branding and marketing [112]. They boarded this to Finnish market in 2010, and then renewed it in 2011 for the next two years forward.

They call this strategic alliance as a “compiling service”, which is formed by bundling

“Spotify Premium” with TeliaSonera subscription plans, as below27: • Mobile Post-paid subscription

User gets a free access for 6 months to “Spotify Premium”. The offering is available for user who signs for 24 months contract for selected broadband packages on selected handsets. User can activate the offering using the same Spotify account in fixed broadband and IP TV connection.

• Mobile Pre-paid subscription

User gets a discount of “Spotify Premium” subscription for a year plus 1 GB mobile browsing bonus each month through “Spotify Annual pass”. These offerings are bundled through selected handsets purchased and/or refill/started kit28.

The data connection used from Spotify service is still counted toward the bucket consumption billing though, as TeliaSonera implements the “Volume-based” data plans, which !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!24 “The Spotify music catalogue”, available http://www.spotify.com/se/about/music-catalogue-info/ 25 “Record labels and artists”, available http://www.spotify.com/se/work-with-us/record labels-and-artists/ 26“About Us”, available http://www.teliasonera.com/en/about-us/ 27 “Spotify”, available http://www.telia.se/privat/minasidor/noje/spotify 28 “Kontantkort”, available http://www.telia.se/privat/mobilt/kontantkort/

34!

is caped based on the bandwidth volume. According to our interview with Lars Roth (TeliaSonera), TeliaSonera also does do the packet inspection to differentiate Spotify’s traffic with other OTT traffic. However, once a user reaches the limit, s/he still can get access to Spotify’s streaming service and other OTT services, as TeliaSonera offers free “Spotify premium” access during promotion early months using “Soft cap with throttling” policy [74].

Lars Roth (TeliaSonera) mentioned that TeliaSonera able to offers this there is a discount

arrangement they got from the music industry actors (content owners), resulting from its exclusive relation with Spotify. The detail of revenue sharing model is not officially announced though, but Lars Roth informed that TeliaSonera gets some chunk of the profit, which is got from the fee paid by every “Spotify Premium” subscribers that are acquired through TeliaSonera bundling offerings. It was also revealed that most of the profit goes toward the content owners. This is also mentioned in [18] as a possible characteristic in a strategic alliance, where distribution of the profit may not be fully equal among alliance partners.

B. Analysis of the Values Exchanged The offering opens possibility for both parties to expand their market and tight the bond with their end-users. Although the exact addressable market is not mentioned in the marketing campaign, Spotify would mostly suit the young end-users. Therefore, this offering would have a strong attraction to the first-time broadband and TV subscribers. In addition, TeliaSonera’s main competitor has launched similar service29 and the competition in Sweden in Finland pushes price and ARPU down. Lars Roth (TeliaSonera) addressed this issue as below,

“Spotify is extremely oriented building the brand through the product. We can't really do that, because our competitor is following us quite fast”.

Lars Roth also stated that they have tried to unlock music business using point-to-point

model, as discussed in section 4.2.3.1, but failed. He stated this as below, “We actually started with our own service; much more like Spotify…It was under brand, but it was not great... We tried that for half a year and realized that we needed to spend all our communication explaining ‘what the service was’, ‘how it works’, and ‘why you should get it’. (At the end) the perception and value we gained from (offering) six months for free was pretty low”.

Andreas Liffgarden also mentioned that there are “cultural differences” between telecom

and media industry that makes direct collaboration between telecom operator and media industry does not work. This brings a need for a “bridging actor” that aggregates the objectives of both telecom operator and media content owner. The main objective of the content owner in this case, according to Andreas Liffgarden, is to monetize their copyright works continuously in a scale based.

In Spotify’s side, its relation with TeliaSonera is proved to increase its own end-user based.

This strategic alliance drives end-users directly to higher value subscription (“Spotify premium”), which is still the core of Spotify’s revenue source (see section 6.3). At the time when the agreement was firstly launched, Spotify was still a young brand and had small market segment. Collaborating with TeliaSonera would definitely leverage to this and help on bargaining with music industry actors, who are taking control over the business ecosystem. Lars Roth mentioned about TeliaSonera’s role as a telecom operator to this strategic alliance as below,

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!29 Partnership between Telenor with WiMP

35!

“In order to get end-to-end deliver to their service, they (Spotify) need a good network... The other reason is, what we have done in giving trust to new company, it is enormous. Telia is the most trusted brand overall in Sweden. When we pick up a partner, we more and less say this is a partner you can do business with. And also, since we are billing provider for them, we charge our customer and get the money to them...Also our bills are one of the most paid in Sweden, with lowest bad debt as well. So this is a really good channel (in order) to have a continuous relation with customer, especially when they have problem when they do that with credit card”.

Indeed, during the first term of agreement, Spotify subscriptions through TeliaSonera have

increased by ~ 300 %, acquired up to ~ 25 % of “Spotify Premium” in Sweden. This is in line with the IFPI report about the substantial increase in the digital music sales, which is also due to the law enforcement for piracy (IPRED law) since that year [75]. The revenue sharing model is not officially informed though. Both of our interviews with Both Andreas Liffgarden and Lars Roth revealed that there is no direct arrangement between them and the music industry actors, but it was involved in the discussions to help on the licensing deal for the copyright works used under the alliance model.

The lines of reasoning of the partnership between these actors match with the three points

mentioned in [12]. The “Learning and internationalization” can be seen in Spotify side. As a new company in media market and to support its expansion strategy, Spotify needs the resources acquisitions, which in this case are large end-user based and good network coverage. In TeliaSonera side, the “Cospesialization” motive can clearly be identified. Using this partnership, it can focus on its core businesses and let Spotify deal with the licensing, product development and maintenance, and share or even avoid the effort for marketing and branding.

In addition, the “Cooption” reason appears for the both sides, as both gains benefits by

collaborating, instead of competing each other (“co-opetition”). For example, instead of delivering its own, which was not successful, TeliaSonera cooperates with the actor that previously was its competitor. Nevertheless, it avoids of just being a “dump pipe” provider. By using this agreement, TeliaSonera has built at least three strategic objectives as what the “Point-to-Multipoint” model could offer, which are to reduce the churn rate, expand market share, and increase the ARPU by up-selling the subscription to the higher value packages.

36!

Chapter 7 – The Indonesian Market This chapter presents a brief introduction to regulatory framework of the telecom and ICT sector in Indonesia, an analysis to the mobile telecom and media market, as well as the quantitative study to the Indonesian users. The first part of the quantities study is the background data, while the second is the test and analysis to the hypotheses. It was made based on the UTAUT framework and findings in the previous sections in related with the Indonesian mobile user characteristics. 7.1. Regulatory Framework Indonesia’s telecom regulatory regime can be divided into 3 phases, Pre-1989, 1989 – 1999, and 1999 – present. The current phase can be concluded as mentioned in [27], shown in Table 7. The catalyst of the last phase was fastened by the Asia’s economic crisis in 1997 that affected Indonesia as well. As part of the economic recovery program assisted by the International Monetary Fund (IMF), the government of Indonesia (GOI) started the deregulation and privatization of the telecom sector. This is one form of its commitment to the 1998 World Trade Organization (WTO) Basic Telecommunications Agreement30, translated as the “Telecom development Blueprint”31.

Table 7. Formation changes in Indonesia's Telecom Regulation

Phase Legal Form Field Form 1999 – present Telecommunications

Act No. 6 of 1999 • Duopoly in fixed-line operations, and private participation for other operations without need for a joint partnership with state-owned company • Network, and service provider structure, • Establishment of independent regulator

In 2007, the GOI issued foreign investment barriers in the several key sectors, including

telecommunication32. Under this negative list, the foreign participation is capped at 65% for the fixed-line business and 49% for mobile, satellite and Internet service business (ISP). This rule is not retroactive though, but it sends a signal regarding future investment policy. In 2008, the tower business was closed, and the existing foreign investors were expected to exit within two years. Thus, either the MNOs or domestic provider owns the tower. In 2010, there was also new policy that limits the establishment of the private broadcaster, which is capped 20 % for foreign capital33. 7.1.1 QoS and Technology Convergence There have been discussions in the Ministry of ICT (KOMINFO) to accommodate the harmonization to the telecommunication, Internet, media, and entertainment (TIME). It falls under the telematics issues, but there is still no specific act that has been formed to take account the convergence issue yet. Currently, the misuse of illegal content distribution, especially the pornography content, is subjected to the Electronic Information and Transactions Act [25].

The provisioning of the services in the cloud ("Aplikasi informatika"), including media, is

not handled neither by the telecom regulatory authority (Badan Regulasi Telekomunikasi Indonesia/BRTI) or broadcasting regulatory authority (Komisi Penyiaran Indonesia/KPI).

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!30 World Trade Organization (WTO) Agreement on Basic Telecommunications Services, or Basic Telecom Agreement 31 The blue print is based on Ministry of Communications (“MoC”) Decree No. KM 72 of 1999 32 “Regulation of President of The Republic of Indonesia No.111 of 2007”, available at http://www3.bkpm.go.id/file_uploaded/Perpres_%20No_%20111_%20Tahun%202007_en.pdf 33 “Regulation of President of The Republic of Indonesia No. 36 of 2010”, available at http://www3.bkpm.go.id/file_uploaded/PPres-36-2010.pdf

37!

According to Didik Akhmadi Usman (BRTI), there is a tendency to put this under BRTI control. The BRTI has willingness to address so the regulator can oversee the emerging technologies and the state gets something from the purveying of these services, exp. taxation issues34.

There is also no specific QoS regulation formed supplying of the broadband services yet

(e.g. minimum speed), though there are guidelines of the minimum requirement for QoS basic telephony services. Didik Akhmadi Usman (BRTI), argues that it should fall under consumer protection law, and the providers must give lucid information (e.g. in the advertisement, service-level agreement/SLA), so consumers aware about the terms and condition when they are using data connection. However, there is no obligation that forces the provider to submit their SLA to the regulator for approval. This is inline with the TRE score in 2008 that found the QoS of mobile and broadband service in Indonesia are below the average, due to lack of regulation [26].

7.2. Telecom Market 7.2.1 Structure and Characteristic Indonesia with its 248 million populations is currently the world’s fourth largest country. This market is among the other 120 that was classified to have a low Internet penetration, as the Internet penetration rate was ~35 % by the end of 2011 [61]. The fixed broadband penetration was only ~1.2%, because PC ownership was also rare with just ~ 6.9% penetration rate, and the fixed broadband plans are relatively expensive among Indonesians (~USD 100/months). This makes this country has a lower fixed penetration rate, compared with the neighbors in the region and among the BRICI countries, as shown in Figure 19.

Figure 19. Fixed broadband penetration in the BRICI countries [81]

This can be understood, as there were problems in the infrastructure deployment, due to the

geographic split of the country’s archipelago territory and complex terrain. Therefore, the mobile communication is assessed as a better suit to this market. Indonesia’s telecom market shows common trends like most of the developing countries. The mobile telecom dominates with penetration rate has passed ~113 % by the end of 2011 [61], which makes it as the world’s fifth largest mobile market. The market is heavily dominated by the prepaid subscription, which has reach more than 95% total subscriber [61], and the churn rate has reached ~ 146% annually [85].

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!34 For example, there have been also conversations between GOI with the international OTT players, such as Google and Blackberry (for the blackberry services), to set up a representative office and build the data server in the country.

38!

The mobile penetration is predicted to scale up to ~173% by 2016, because there are multiple ownerships of the Subscriber Identification Module (SIM) and the inactive subscriptions distort the true market growth [61]. This happens due to the four factors below: • First, the starter packs are cheaper compared to the top up fee. • Second, no numbering cost regulation for releasing a new cellular number. • Third, the perception between GSM and CDMA tariff, where the CDMA tariff

(especially voice) is perceived cheaper than the GSM tariff. • Fourth, the number portability is not implemented yet. Thus, people tend to reluctant

losing their existing GSM cellular number and keep their subscription plan, as it was rolled out earlier.

Although Indonesia is perceived as one of the riskier destinations in Asia [61], there has been presence of the key investors in the telecom sector. The market is supported by a total nine MNOs, consists of five GSM-based and four CDMA-based operators, which most of them have several brands targeted to the specific end-user segments, as shown in Table 31. This is quite a number of players and most of them are granted national licenses, though some have started consolidation35. This is a different situation compared to the other emerging markets, e.g. India, that also has a lot of MNOs and subscribers, but most of them are granted regional license.

The competition has been intense since the introduction of the new interconnection policy in 2006, which decreased almost all of the interconnection tariffs for voice36. It leads to the massive load in the network utilization, especially for the basic services (traffic volume and minutes of usage for voice, and SMS). The blended ARPU shows a decreasing trend since several years ago, which is currently ~26,000 IDR [61], and forecasted to keep down in future [90]. This is not just affected by the competition, but also due to the large low-income based subscribers, following the MNO expansion to the rural and densely populated areas [86].

The subscribers and the revenue are mainly concentrated in Java and Sumatra, which are

the most populated islands [60], as well as relatively have stronger purchasing power and dominate the mobile ownership in the country [93]. The big three (incumbents) become more prosperous in this circumstance, as they have been supported by the well-established network coverage and good record of accomplishment. Therefore, they are more attractive to the high-value and loyal subscribers, compared with the green field operators. They also acquire more spectrums that make them have a better flexibility for expansion and price (see Table 31).

7.2.2 Mobile broadband The 3G services have been available in the country since 2006, but its uptake was low with less than ~ 20% penetration [87]. This was not just due to the slow roll out of the network, but also the devices price was not affordable and attracted scale subscriber yet. Thus, the GOI has made a target to increase the country’s broadband penetration, especially to push the Internet literacy in the remote areas. Mobile broadband has been launched in 2009 and since then it has been served by all the GSM-based and two CDMA-based MNOs (see Table 31). In addition, The GOI has released additional 3G spectrums to raise the network capacity.

Indeed, there is an enormous opportunity for the MNOs to expand through mobile

broadband. There would be increasing subscribers in Indonesia that would use the underlying technologies in the following years [85]. The mobile device has also been as the preferred access to the Internet among Indonesians. It makes this market unique, as the consumer

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!35 The recent example is between Mobile 8 and Smart Telecom, which forms as Smartfren Telecom 36 Except the interconnection tariffs of fixed or FWA originated from / terminated to mobile, VoIP, Satellite and International. See http://www.postel.go.id/content/ID/regulasi/telekomunikasi/kepmen/pm%20nomor8-2006-interkoneksi.pdf

39!

normally gradually changes from the “no connection” stage, to adopting landlines and then subsequently to the mobile devices. In contrast, there is a tendency of “leapfrog trend” as in Indonesia, where people head straight to mobile [86]. We can compare this phenomenon with several years ago where Internet café (“Warung Internet”/“warnet”) was a preferred way among people to get access to Internet [113].

Figure 20. Mobile phone penetration in the BRICI countries [81]

As also shown in Figure 20, mobile penetration has been increasing quite fast in Indonesia,

even compared to some other BRICI countries [86]. This is in contrast with the landlines penetration that has dropped since recent years [93]. Mobile broadband plans in Indonesia are among the cheapest in the world [88], also in contrast with the situation in fixed broadband discussed previously. The price of smart phones is also falling and more affordable for Indonesians, which is also supported by the attractive Blackberry service and device bundling. Blackberry devices and service (e.g. Blackberry messenger and mail services) have been booming since 2008, and Indonesia is currently RIM’s largest market [89]. This makes this market as a contrarian market for the smart phone.

Furthermore, the share of the data ARPU from the total ARPU is predicted to keep

increasing [80]. In the future of the mobile data revenue, the infotainment would contribute 25 % in 2016 compared with the 19 % in 2009, while the revenue from the messaging would decrease from 61 % to 52 % respectively. The MNOs saw a slight decline with their mobile data revenue in 2011, but it was more related with the suspension of VAS services by the regulator at that year, as a response to the massive complains regarding fraud in premium services (mainly SMS). Didik Akhmadi Usman (BRTI) also highlighted this case during the interview.

Moreover, Indonesia is also forecasted as one of the leaders in the emerging Asia-pacific

markets for the LTE deployment by 2016, after China [89]. These are mainly driven by at least two underlying reasons below: • First, the country’s economic growth turns to be the fuels for its telecom industry. There

is a huge growth in the domestic consumption and Indonesia is assessed relatively resistant to the global economic crisis. There is a booming of the middle-class income in this country, and the GDP per capita is predicted to keep increasing [98].

• Second, the demographic plays a major part, as the country’s population is dominated by

extremely young people with a median age of 27 [60]. These young people normally have

40!

tendencies to be open for new technology, which can be seen from the popularity of the instant messaging, social media, bulletin boards, and blog in Indonesia37.

7.3. Media Services Market Several players have offered the music and TV/Video services in Indonesia, both as ownership, access model, and personalization model. This has been initiated since 2006, in line with the 3G launching. There are some of the dominant characteristics, which vary this market with others. Indonesia is one of the few markets worldwide that still sees personalization services dominating the digital music landscape, as 95 % of revenue comes from this (mainly RBT, followed by Ring-tone activation).

Several prominent retailers and/or service providers in OTT music and/or TV/Video business in Indonesia are the MNOs, the TV broadcastings, the mobile device OEM & OS provider (Nokia), and local OTT players (Jatis mobile, Mivo TV, Kapan Lagi, etc). There is limited focus and availability of the prominent international OTT music services (streaming and download services), which create a hole in the ecosystem, e.g. iTunes. Some of the MNOs focus to specific devices, such as Blackberry or locked low-end mobile devices launched by the MNO itself, e.g. “HP Esia” phone.

The revenue generations used are mostly retail (for music, e.g. unlimited download/week) and subsidized-based (for TV/Video, e.g. advertising based). However, the piracy level in Indonesia is extremely high, as people currently can easily buy the illegal CD/DVD that contains music, TV series, or movie contents for less than ~10 SEK or simply download it through the illegal websites/file sharing. In music business, the piracy market was reported twice than the legal market value [69], and reached 88% in 2006 [90]. This makes Indonesia as one of 10 priority countries outlined by the IFPI according to the degree of threat posed by the music piracy [91].

The local players in the market also see the competition has been aggressive, and they have

tried to fine out other revenue stream and/or to boost the demand to its main service offering. Although there is a partnership done by Telkom Indonesia that set up Joint Venture (JV) with SK Telekom (Melon) [114] for provisioning media services, the majority MNOs in this market typically deal directly with the content owners by acquiring the license to distribute the copyright works. In other words, they use point-to-point model, which was also typically seen in European markets before the trend moved to the point-to-multipoint, as discussed in section 4.2.2 and 5.2.2.

The main reason of this dominance point-to-point model done by the MNOs is due to the

limited appearance of international prominent OTT players38. Hengky P. Ginting (NST) stated that usually 50-50 revenue sharing models are used, which is done after the MNO takes out the cost to deliver the service. The roles of aggregator/collecting society/ publisher are minimum or bypassed. The MNOs mainly handle the marketing, branding, platform development and maintenance. There is a strong appearance of the mobile content providers (CP), but they run business in monetizing non-audio/video content/merchandizes, and only few exceptions where the CP bridges the MNO with the content owners39.

Although several local and international parties have tried their attempt, their service

offering does not take off yet, due to the piracy issues and user’s willingness to pay is also low (discussed later at section 7.4). Therefore, it is natural for the MNO to choose this strategy,

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!37 Indonesia is one of the world’s largest facebook and twitter the end-users, and Indonesian is also one of the most used languages on Wordpress. 38 Apple decided to skip Indonesia along with China and India for iTunes music store 39 Several examples are wallpaper download, premium SMS content subscription/quiz.

41!

because they have strong appearance in the market and stronger bargaining position toward the media industry actors. This is in contrast with the situation we have seen in the developed countries. These are also as result of the several factors, which were also mentioned during the interview with Jockie Heruseon (Telkomsel), Yose T. Arizal (Axis) and Hengky P. Ginting (NST), as below: • First, there are high dependencies among the people to access the Internet through the

mobile connection (as discussed in section 7.1). • Second, the MNO billing capability offers the easiest payment mechanism, because the

bank account penetration is extremely low, covering less than 20% of the population [59]. • Third, the personalization services (RBT, Ring-tone activation, etc) are assessed to have

the simplest content delivery channel (e.g. in term of billing) and are safe from the piracy issues.

• Fourth, the licensing fee in monetizing the media content is much lower compared with the developed markets. Therefore, MNO has less issue to deliver their service directly to end-users, instead of collaborating with the OTT player (to reduce the cost).

7.4. Mobile user Understanding the end-user, including anticipating their needs and exceeding their expectations is the key in building any business. Apart from the country’s regulation and competition field, Indonesian user has a unique OTT music listening and TV/video viewing behavior. User acceptance is one of the critical aspects that influence the success of the new mobile services provisioning. Thus, it is important to include the analysis of users’ expectation and behavior, in order to understand what influences the Indonesian users’ acceptance toward the mobile media services.

We have formulated the hypotheses related to the Indonesian user and OTT media services,

according to our findings about the Indonesian market, both its regulation, telecom and media market situation, as discussed in section 7.1, 7.1, and 7.2. This was also formulated after we saw the trends in the global and reference market (Sweden), as discussed in section 3 - 7. These hypotheses and explanation of it variables are available in Table 32 at Appendix C - Quantitative Survey. We used the UTAUT framework and 119 responds to test the hypotheses, as discussed in section 2.1 and 3.1.3. The respondents’ answers are compiled as the graphs, which are available in section C.2 at Appendix C - Quantitative Survey. 7.4.1 Sample Characteristics The first initial questions are to identify the respondent’s background. As shown in Figure 21 and Figure 22, the majority of our respondents are men (~ 76 %) and the average age is ~ 29 years old. All of these respondents are identified to have mobile device, and although this proposition is not varied, it correlates with mobile penetration in Indonesia that has passed 100 % penetration rate, as discussed in section 7.2 previously.

Furthermore, as can be seen in Figure 24, all of these respondents have ever used Internet

through mobile device. 28 % of them confirmed that mobile device is their primary way to access Internet, as shown in Figure 25. These findings are in line with the discussion in section 7.2.1 regarding a high dependency among Indonesians toward the mobile Internet. Therefore, according to these findings, we conclude that the hypothesis 1 as mentioned in Table 8 is accepted.

Table 8. Hypothesis 1 result

Moreover, as can be seen in Figure 26, the most popular services used by these respondents

in their mobile phone are email, Instant messenger, and social media. In addition, more than 60

Hypothesis 1 Result Indonesian users have already used with the mobile internet ACCEPTED

42!

% of them confirmed that they have used OTT music and TV/Video services through the mobile device, although personalization services are still a major revenue stream for the actors in the market, especially for digital music content distribution. Thus, according to these findings, we conclude that the hypothesis 2 as mentioned in Table 9 is rejected.

Table 9. Hypothesis 2 result

Hypothesis 2 Result Indonesian users are not exposed with the mobile media services yet

REJECTED

7.4.2 Technology Acceptance 1. Behavior intention The first hypotheses that we test are related with the behavior intention aspects of the Indonesian users to adopt the OTT media services.

a. Willingness to use the service The first hypothesis is the Indonesian users’ attention to adopt the OTT media services. There are three questions that gather the information whether the respondents will use/keep using the services (Q10), will have/keep having the services’ application in their mobile devices (Q13), and will use/keep using the services when there is opportunity to use it for free (Q17). The respondents’ answers are shown in Figure 30, Figure 33, and Figure 37.

Table 10. Average and Internal consistency of respondent’s willingness to use Mean Services

Q10 Q13 Q17 Music streaming 2,924 3,261 4,118 Music download 3,092 3,235 4,106 Video streaming 3,218 3,429 4,176 Video download 2,916 3,176 4,106

The data in Table 10 reveals that the respondents generally have high intention to use video

streaming service, followed by the music download service, as the average responds values for these services in Q10 are above the neutral value (3). The music streaming and video download service have quite lower and below the neutral value. However, the average responds values in Q13 and Q17 show us that the respondents’ intention to have music streaming service and video download service (Q13) and to use it for free (Q17) is high.

Table 11. Internal consistency of respondents’ expectancies

Cronbach Alpha

Pearson correlation Services

Q10, Q13, Q17 Q10, Q13 Q10, Q17 Q17, Q13 Music Streaming 0,759 0,709 0,386 0,430 Music Download 0,769 0,701 0,463 0,401 Video Streaming 0,783 0,735 0,418 0,479 Video download 0,714 0,652 0,361 0,329

We tested the reliability of these data, as shown in Table 11. The Cronbach Alpha was used

to check the consistency in Q10, Q13, and Q17. The Pearson correlation was also used check the consistency in Q10, Q13, and Q17, by pairing them each other to get the correlation coefficient. The results from Cronbach Alpha show us that all of the consistency values are in the range of acceptable.

Meanwhile, all the correlation coefficients from Pearson correlation between Q10 and Q13

are in the range of strong value, which means the respondent’s willingness to use / keep using

43!

the services have tight correlation with their willingness to have the service in their mobile phone. All the correlation coefficients between Q10 and Q17, and Q13 and Q17 are in the range of medium. This means there is also correlation between respondent’s willingness to have the service in their mobile phone with their willingness to use / keep using the service and the opportunity to use the services for free, but this is not as strong as the correlations between Q10 and Q13. It can be seen that the opportunity to use the services for free is positively valuable, but this is not perceived to be so much significant for Indonesian users.

However, these all supports the hypothesis about Indonesian mobile users’ attitude toward

the media services, as the average responds values and internal consistencies among the answers are good. Based on this, we conclude that the hypothesis 2 as mentioned in Table 12 is accepted.

Table 12. Hypothesis 3 result

Hypothesis 3 Services Result Music streaming ACCEPTED Music download ACCEPTED Video streaming ACCEPTED

Indonesian users have positive intention to use the mobile media services

Video download ACCEPTED b. Willingness to pay for the service The next hypothesis that we tested was the intention of the Indonesian users’ to pay to use the OTT media services. We use the data from Q9 that gathers the information whether the respondents think they would like to pay to use the services. As can be seen in the raw data in the Figure 29, majority of the respondents think they would not like to pay, which applied in all of the four services. The music streaming service has the highest proportion of rejection to pay (~ 75 %), followed by the video download (67 %). This is in line with the lower intention from the respondent to use both services, as revealed in Q10 previously. Based on this, we conclude that the hypothesis 4 as mentioned in Table 13 is accepted.

Table 13. Hypothesis 4 result Hypothesis 4 Services Result

Music streaming ACCEPTED Music download ACCEPTED Video streaming ACCEPTED

Indonesian users have negative intention to pay for the mobile media services

Video download ACCEPTED 2. Expectancies The next hypothesis that we tested was related with the effort and performance expectancy. There are four questions that gather the information whether the respondents think the services are useful and interested to be used compared with physical distribution (Q15), whether using the services save their money compared with physical distribution (Q11), and whether it saves their time compared with physical distribution (Q20), and if the services are complicated to use (Q18). The respondents’ answers are shown in Figure 31, Figure 35, and Figure 40.

Table 14. Average of respondents’ expectancies

Mean Services Q15 Q11 Q20 Q18

Music streaming 3.210 3.244 3.252 2,017 Music download 3.202 3.437 3.310 2,076 Video streaming 3.395 3.403 3.369 2,084 Video download 3.277 3.311 3.327 2,113

44!

The data from Table 14 above reveals that the respondents generally have a good attitude and expectancy towards all the services. This can be seen from the average responds values of Q11, Q15, and Q20 that are above the neutral value. In contrast, all the average responds values of Q18 are below the neutral value, which means that the respondents do not think the OTT media services as complicated.

Table 15. Internal consistency of respondents’ expectancies Cronbach

Alpha Pearson correlation Services

Q15, Q11, Q20 Q15, Q18 Q11, Q18 Q20, Q18 Music Streaming 0.792 -0,391 -0,182 -0,359 Music Download 0.780 -0,319 -0,158 -0,286 Video Streaming 0.797 -0,432 -0,308 -0,409 Video download 0.766 -0,330 -0,215 -0,331

We tested the reliability of these data, as shown in Table 15. The Cronbach Alpha was used

to check the consistency in Q11, Q15, and Q20. Meanwhile, the Pearson correlation was used for Q18, as it leads to different direction with the three questions previously. We get the correlation coefficient by pairing each data of Q15, Q11, and Q20 with Q18.

The results from Cronbach Alpha show us that all of the consistency values are in the range

of acceptable. Meanwhile, all the correlation coefficients from Pearson correlation show us the negative value in the range of medium and small, where the negative correlation in video streaming is the highest compared with other services in all of the Pearson correlation values. These all supports the hypothesis about Indonesian mobile users’ attitude toward the media services, as the average responds values and internal consistencies among the answers are good. Based on this, we conclude that the hypothesis 5 as mentioned in

Table 16 is accepted.

Table 16. Hypothesis 5 result Hypothesis 5 Services Result

Music streaming ACCEPTED Music download ACCEPTED Video streaming ACCEPTED

Indonesian users have positive attitude toward the mobile media services rather than physical products

Video download ACCEPTED

3. Facilitating condition The next hypotheses that we tested were related with the barriers for Indonesian mobile users to embrace the OTT media services. a. MNO’s network and mobile billing Two questions gather the information whether the respondents think their MNO network coverage and capability do not support them to use the services (Q5), and whether accessing the services causes them to pay a lot for mobile account billing (Q19). The data from the Table 17 reveals that the respondents generally do not think of the mobile network as a major barrier for them to use the services. They also believe they will not have to pay a lot for mobile account billing because of using the services. This can be seen from the low average responds values in Q12 and Q19 that are both below the neutral value.

Table 17. Average and Internal consistency of respondents’ perceive to MNO network &

billing Mean Cronbach Alpha Pearson correlation Services

Q12 Q19 Q12, Q19, Q10 Q12, Q10 Q19, Q10 Music streaming 2,328 2,555 0,167 0,134 -0,027

45!

Music download 2,294 2,521 0,271 0,097 0,051 Video streaming 2,454 2,647 0,186 0,143 -0,004 Video download 2,454 2,630 0,239 0,066 0,112

We tested the reliability of these data, as shown in Table 18. The Cronbach Alpha was used

to check the consistency in Q12, Q19, and Q10, as it contains the data of the respondents’ willingness to use the service. Meanwhile, the Pearson correlation was used by pairing each of the data in the Q12 and Q19 with Q10.

The results from Cronbach Alpha show us that all of the consistency values are in the range

of unacceptable and the coefficient correlations are in range of small to none. These values show us that there is no correlation between respondents’ willingness to use the service with the barriers from mobile network quality and billing. This makes sense, as the low average values above show us that the respondents do not think these barriers exist. Based on this, we conclude that the hypothesis 6 as mentioned in Table 18 is accepted.

Table 18. Hypothesis 6 result

Hypothesis 6 Services Result Music streaming REJECTED Music download REJECTED Video streaming REJECTED

Indonesian mobile users have positive perception that the mobile network and billing are a major obstacle to use the mobile media services

Video download REJECTED b. Mobile device Two questions gather the information whether the respondents think their mobile device support for using the services (Q21), and whether they can afford to buy mobile device that support to use the services (Q23). The data from Table 19 reveals that the respondents generally think their mobile device support in using the services, and they can afford to buy a mobile device that supports to use the services. This can be seen from the above average responds values that are below the neutral value.

Table 19. Average and Internal consistency of the respondents’ perceive in mobile device

feature and ownership Mean Cronbach Alpha Pearson correlation Services

Q21 Q23 Q21, Q23, Q10 Q21, Q10 Q23, Q10 Music streaming 3,992 3,798 0,540 0,480 0,127 Music download 4,025 3,782 0,469 0,169 0,042 Video streaming 4,025 3,798 0,550 0,054 0,182 Video download 4 3,782 0,405 0,176 0,061

We also tested the reliability of these data, as shown in Table 19, whether there is any

correlation with respondent willingness to use the services. The Cronbach Alpha was used to check the consistency in Q21, Q23, and Q10, as it contains the data of the respondents’ willingness to use the service. Meanwhile, the Pearson correlation was also used by pairing each of the data in the Q21 and Q23 with Q10.

The results from Cronbach Alpha show us that all of the consistency values are in the

range of poor and unacceptable. The coefficient correlations are small to none, except for music streaming. These values also show us that there is no correlation between respondents’ willingness to use the service with affordability to buy a mobile device, as well as whether respondent’s mobile device support the usage of the services or not (music download, video streaming, video download). However, as the average responds values are above the neutral value, we conclude that the hypothesis 7 as mentioned in Table 20 is rejected.

46!

Table 20. Hypothesis 7 result Hypothesis 7 Services Result

Music streaming REJECTED Music download REJECTED Video streaming REJECTED

Indonesian mobile users have positive perception that the mobile devices is a major obstacle to use the mobile media services

Video download REJECTED 4. Social influence The next hypotheses that we tested were related with the social influence to Indonesian mobile users to adopt the OTT media services. Two questions gather the information whether the respondents think the services would be popular among Indonesian mobile users (Q16), and whether they would get positive attention from their social environment because of using the services (Q22).

The data from Table 21 reveals that the respondents deem the services would be popular

among the Indonesian users, but they do not think they would get any positive attention because of using the services. This can be seen from the high average responds values in Q16, where all of that are below the neutral value. In contrast, these are low in Q22 for all of the four services.

Table 21. Average and Internal consistency of the respondents’ believe about social

environment Mean Cronbach Alpha Pearson correlation Services

Q16 Q22 Q16, Q10 Q16, Q10 Q22, Q10 Music streaming 3,445 2,765 0,761 0,617 0,226 Music download 3,613 2,748 0,748 0,603 0,336 Video streaming 3,613 2,832 0,661 0,501 0,136 Video download 3,504 2,815 0,633 0,464 0,279

We also tested the reliability of these data, as shown in Table 21, whether there is any

correlation with respondent willingness to use the services. The Pearson correlation was used to check the consistency of these questions, by pairing each of the data in the Q12 and Q19 with Q10, as it contains the data of the respondents’ willingness to use the service. The coefficient correlations between Q22 and Q10 are small to none, except for music download. In contrast, the coefficient correlations between Q16 and Q10 show us the contrast value, which are in the range of medium and strong. The Cronbach Alpha between Q16 and Q10 also show correlation (especially music streaming and music download), as the values are in the range of acceptable.

According to these findings, we can conclude that the Indonesian mobile users’ willingness

to adapt to the service have a correlation with a perception that the services would be popular among other Indonesians. However, this does not correlate with respondents’ motivation to get any positive attention from their social environment, except for music download. This makes senses, as they do not believe they would get any positive attention from their social environment because of using these services (music streaming, video streaming, video download). According to these finding, the hypothesis 8 as mentioned in Table 22 is accepted.

Table 22. Hypothesis 8 result

Hypothesis 8 Services Result Music streaming ACCEPTED Music download ACCEPTED Video streaming ACCEPTED

Indonesian mobile users have significant influence from their social environment to use the mobile media services

Video download ACCEPTED

47!

Chapter 8 – Implications to Indonesia This chapter presents the discussion of implication from mobile media services in mobile broadband in Indonesia. The discussion is formulated based on the findings and analysis derived from the global market trend, lesson learned from mobile media service pre-study in Swedish market, as well as Indonesian regulatory, telecom and media market situation. It is presented by answering the research questions mentioned in section 1.2. 8.1. Actors, Relations and Business models In this section, we formulate the answer for the first research question: “Who are the actors that involve in serving the media services, and what kind of relations and business models do they use?”

In the digital music, these actors are the content owners (music artist, publisher and record

label), which hold the right to the copyright work, as well as the retailer/service provider, which is directly connected to the user. In some cases, the collecting society and aggregator actor may involve at the back end. In the digital TV/Video, there are also the content owners (studio and TV broadcaster), but the artist is excluded, as it does not hold the right to the copyright work. There is also a role of the retailer/service provider, and the aggregator may be involved at the back end of the service provisioning. The CDN provider’s role is increasingly important here, compared with the music business, as the data distributed is much larger.

Furthermore, the technology advance makes it possible for multitudes of actor, which were

previously never seen in the traditional music and TV/Video business and come from outside the media industry, to also play role as a retailer/service provider and serve media services directly to the end-user. These actors can be categorized into three types, which are indentified in [90] as well, as below: • The Pure player

This kind actor has the music and TV/Video service as its core business. It simply mimics and expands the role in the traditional era into the digital and OTT era. Such examples of the pure OTT music players are Spotify, Deezer, Napster, etc, and the TV/Video players are Hulu, Neflix, Voddler, etc.

• The Diversified player This kind actor previously runs non-music and/or TV/Video business. It enters the media industry in order to expand its brands and broaden consumer presence, which is possible following the advance of the digital and Internet technology. This actor comes from various kinds of industry, such as Amazon, Wal-Mart, Google, etc.

• The Vertically integrated player This actor previously also runs non-music and/or TV/Video business. It owns the platform or infrastructure, as the based of its core product or service offerings. It actually integrates the music or TV/Video services into its business in order to increase the value and drive the demand to these core offerings. As same as the actor in the Diversified player category, this actor may also come from the various industries, such as CSP (TeliaSonera, TDC, etc), mobile device OEM and OS provider (Apple, Nokia, etc).

The relation between actors inside the music and TV/Video industry in the digital/online era

has been visualized and discussed in 4.2.1 and 5.2.1. This is the value network of the “core actors”, which is still same with the traditional distribution era. Our findings show that the change and variation in the value network, both in the music and TV/Video service provisioning, happen outside this “core network”. This value networks can be grouped into three categories, as below:

48!

• The Direct-to-fan model Here, the content owners directly reach their end-users through their own service or product platform. This model is visualized in section 4.2.3.1 and 5.2.3.1 for the music and TV/Video service provisioning respectively.

• The Point-to-point model In this model, the content owners do not directly engage with the end-user. Multitude of retailers/service providers, as discussed above, obtains the license, either directly from content owners, or via collecting society and content aggregator. They then provide the service or product platform to deliver the music and TV/Video content to the end-user.

• The Point-to-multipoint model The content owners in this model still play behind the scene, but there is an actor that acts as the bridge between these content owners and other actors outside the core media industry. They usually at the same time also reach their end-user, thus it takes a central position in the value network.

This finding above is also consistently mentioned in [16], which categorized constellation

of the current and emerging business models used in the converged mobile internet services: “Telco centric”, “device centric”, “aggregator centric”, and “service/content provider centric”. The “Telco centric” and “device centric” are typically found in the point-to-point partnership in the OTT music and TV/Video business, where each of the while the “aggregator centric” and “service/content provider centric” is the constellation that can be found in the point-to-multipoint partnerships.

From the discussion in section 4.2.3.1 and 5.2.3.2, we can see that there are two potential

positions that the telecom operator’s (CSP) can take in provisioning the media service, either as a central in the network configuration (“Telco centric”), or as a part in the “aggregator centric” or “service/content provider centric” constellation. After analyzing the recent trend, we conclude that the pure OTT retailer/service provider has more capability to be at the central of the value network. Their focus to the music and/or TV/Video and capability to work in a horizontal partnerships make them has a better flexibility to adapt to the emerging business, compared to the vertically organization approach that is typically used by the telecom operator.

We have also analyzed the business models used by these actors in the taxonomy point of

view, as mentioned in [1]. Despite the diversity of the business model used in the OTT music and TV/Video business, we found the patterns in the revenue, service, and distribution models used, as discussed in section 4.2 and 5.2. The typical revenue models used are retail, subscription, subsidized, or rental (in the TV/Video content distribution). The service and distribution models used are ownership (DRM/Free DRM à-la-carte, or physical packaging, e.g. OTT CD/DVD purchasing), access (on-demand, live, radio streaming), or personalization (music content distribution, e.g. as ring tone, RBT, etc). Mostly, the actors combine these models, e.g. the “Freemium model”, used by Spotify as discussed in section 6.4.1.

49!

8.2. Indonesian Mobile Media user’s characteristic In this section, we formulate the answer for the second research question: “What are the mobile user’s characteristics, especially in Indonesia, toward the media services through mobile broadband?”

The typical mobile user’s characteristics in Indonesia are pricing conscious and generally

have low willingness to pay for the media services. The ARPU has been decreased over several years following the MNO expansion to the rural and densely populated areas. However, there is still a room for the music and TV/Video service using one or combination of the access and ownership service model. The extremely young user segment dominates the population and there is a boosting of the middle-income individuals (see section 7.1). It means this market actually has a significant amount of potential end-user with luxury need beyond the basic expenditures.

This leads us to the conclusion that there is a huge opportunity for the media services

business on top of mobile broadband in Indonesia. The Indonesian mobile users are mostly concentrated in certain areas, mainly in Java and Sumatra Island. These two islands have the most populated cities in the country, and generally have higher purchasing power, and dominate mobile phone ownerships. Our respondents that represent a sample of Indonesian mobile users have also confirmed to have an intention and expectation to adopt OTT media services. The video streaming and music download are the top services that they are using, while music streaming turns as one of the most attractive services, if it is offered free.

Indeed, these aspects, combined with the fact that the social media, bulletin boards, blog,

and location-based services are hugely popular among the Indonesians, would offers the opportunity for advertising (mobile ads). It offers a momentum to use the subsidized-based revenue model, presents a way to reach out consumers directly (Word of mouth /WOM) [49], and able to create a presumption of the popularity of the service. As we have sees at the user survey in section 7.4, there is a positive correlation between willingness to use the services with the perception of the popularity of the service.

It means, the more the people are aware or assume that a particular media service is popular

among other Indonesians, the more they are likely to try to use it. This does not correlate with an individual’s motivation to get positive attention from social environment though, as the Indonesian users do not believe they would get any positive attention because of using these services. The discussion in section 5.2 also supports this finding, that the advertising would be the key to the TV/Video business growth. The right composition between paid and subsidized-based subscription is also found as a vital point for music services, as discussed in section 4.2.

There is a low average TRE score for mobile and broadband QoS in Indonesia, and

dominance of the pre-paid and low-ARPU subscriber. However, the Indonesian users generally have a positive perception of the network quality and do not think the mobile billing would be a barrier for them to use the OTT media services. This was found in the user survey, where respondents generally believe they will not have to pay a lot for mobile account billing because of using the services. Mobile device is also not a major barrier for them to use the services. Instead, smart phone penetration has risen since years in this market, which would be as the driver for these mobile media services, as discussed in section 7.2.1.

Mobile device has been as the preferred access to the Internet among the Indonesians

(“leapfrog trend”), compared to several years ago where the Internet café dominated. We have seen the mobile penetration has been increased fast, even compared to some other BRICI countries. In fact, the mobile broadband plans in Indonesia are among the cheapest in the world. This is in contrast with the fixed broadband, where its availability is still limited and the

50!

penetration has dropped. It is still quite expensive to the majority of Indonesians and the PC penetration is low.

8.3. Potentials and Challenges for MNOs in Indonesia In this section, we formulate the answer for the third research question: “What kind of potentials and challenges would the MNOs in Indonesia face in providing media services on top of mobile broadband?

There are several potential drivers and obstacles aspect ranges in the policy and regulation,

supply and demand side in Indonesia market. As discussed in section 7.1, there is no entry barrier exist in OTT media services provisioning in this market. It means in this circumstance, the media service providers only need to consider the licensing issues to distribute copyright work in the market, both for music and TV/video contents. This market also offers a huge challenge, due to high digital media piracy, both as illegal digital download, file-sharing, and physical transaction. In general, there was still low law enforcement, although there have been undergoing actions done by GOI to address these issues and pornographic content distribution.

Furthermore, we have found that the Indonesian users generally have a low willingness to

pay and are pricing conscious, and this market has high piracy level. This makes the media providers cannot just directly apply the strategies that they have founded successful in the developed markets to Indonesia. Determining the price is a substantial factor in the success of the media services provisioning, thus the providers have to consider the right prices using different pricing tiers, in order to attract different consumer segments away from the illegal content and use the provider legal services. This is also important to ensure the healthy profit for both the provider itself and the content owners. Indeed, the lesson learned we have seen in the success of media retailers/service provider’s global expansion covers these three factors: • Pricing tiers

The tiered pricing system is a differentiation made to suit types of end-user segment (value-based pricing). This can be done either based on the different amount or popularity of the content that can be accessed, range of time to access the content, or kinds of feature (e.g. supported device) in several pricing packages.

• Flexible pricing The flexible pricing is done by adjusting the price based on the “demand-based pricing” and “geographic segmentation”. According to Lars Roth (TeliaSonera) in the interview regarding the lesson learned from the success of Spotify expansion, the price between the market territories should not be always same. Instead, it should follow user’s perceive value in that market.

• Differentiated features The flexible feature offered in different market territories is also the key to make the service take off in a particular market. The personalization services (ring tone download, RBT activation, wallpaper download, etc) are huge popular in Indonesia, and as the main source of revenue for the music industry. Therefore, it is considerably valuable for the provider to include this feature in their OTT retail or service delivery channel. Such example is to make it possible for the end-user to activate particular content (music) as RBT.

The provider should also aware that the availability of the service in the mobile device is

much important in Indonesia, as the majority of the people only have access to the Internet through mobile, as discussed previously. Thus, the mobile device feature should be available in all price tiers, or otherwise the service would not take off. This is a same case when Spotify adjust its offering in the certain territories, e.g. availability of radio feature, 5-track cap implementation, etc.

51!

8.4. Positions and Roles for MNOs in Indonesia In this section, we formulate the answer for the fourth research question: “What kind of positions and roles can the MNOs in Indonesia take to make media services as a profitable business?” 1. Analysis of the current situation The price war has been going on for several years, including in the data service. In contrast with the western markets, the MNOs have low opportunity for triple/ quadruple play because the fixed penetration is not significant. As our findings discussed in section 4.2.3.1 and 5.2.3.2, the triple/ quadruple play is one of the CSPs’ key motives when offering the OTT media services bundling. In addition, the revenue share from the data has just started increase in Indonesia mobile market, and is predicted to keep increase in the next following years.

Indeed, it is not surprise that the MNOs in Indonesia are mainly interested to monetizing the data as access, not yet on monetizing the media service on top of it (bundling), though there has been bundling offering with few other services, e.g. Blackberry services (see section 7.2.1). The threat of being the “dump-pipe” provider in this market is not as obvious as we have seen in the developed markets. In those markets, the MNOs have been forced to figure out the strategic decision regarding the music and TV/Video business, which cause of the data traffic explosion in their network (e.g. Neflix phenomena in the North America markets).

This is also supported by the fact that the personalization service has been proved as

profitable to the MNO, and is still as the main revenue channels from VAS. It is even currently dominating the revenue of the digital music market, which makes Indonesia a contrasting situation with other markets. This is caused by at least three factors. • First, there is a limited appearance of prominent OTT media service players that results on

a hole in the ecosystem (e.g. iTunes). • Second, there is high piracy level in the physical and OTT media distribution. • Third, there is a limited opportunity for online payment, because bank account penetration

is low.

These factors make the MNO in Indonesia has a better bargaining position in the value network toward the media services provisioning, compared with the situation in the western markets, as they are able to offer a simple payment solution. The low fixed-line/fixed broadband penetration and high dependency among Indonesians to the mobile Internet also give the MNO a strong appearance and ability to reach a large end-user scope. This is in contrast with the developed markets, where the mobile connection is as a substitute to the fixed connection. In Indonesia, mobile device is often not just for basic communication, but also may also offers the first opportunity and main way to connect to the Internet. 2. Analysis of the potentials for MNO in Indonesia The CSP, especially the MNO, is confronted with at least three possible options or positions in the OTT music and TV/Video business, as also discussed in [77]. a. The “Do nothing” option

This means the MNO does not involve at all in the OTT music and TV/Video business, and leave the provisioning of the service to the other player. This option is suitable for the MNO that operates in the markets where the network infrastructure (mobile broadband) and/or mobile device penetration (smart phone) do not take of yet. It means the market as overall is not ready yet for these services.

In the OTT music business, although the market as overall is ready, the MNO still could

tolerate this situation if the opportunity for the MNO to gain benefit from the service is too low, which makes it not significant with the effort they should do. This can happen because either the revenue share is too small, or the potential revenue gained from the market is too

52!

low. In addition, the impact of the traffic from the music content delivery for the overall traffic is also likely limited; because the bandwidth requirement for this service is low (high-quality streaming music is likely 320 kbps).

However, this is not the case in the OTT TV/Video business. Even tough the revenue share

is small; the impact of the TV/Video content delivery to the traffic load is significant (see Neflix traffic in the North America markets in the Chapter 5 – The Digital TV/Video Industry). Therefore, if the OTT TV/Video service takes off in the market, the threat to the MNO of being the “dump-pipe” provider is high. In addition, in the both of the business fields, the MNO would miss out the opportunity to monetize, as well as the indirect benefits to be involved in the provisioning of these services in their market. b. The “Do it alone” option

This means the MNO launches its own digital music and/or TV/Video service brand, and compete with the OTT players. Using this, the MNO becomes the retail and/or service provider (The Vertically integrated player) and partnership directly with the music and/or broadcasting core actors (point-to-point model) This option might feasible for the operator that has or able to gain sufficient resource to develop, maintain and deliver its own service. This includes the effort to obtain the license from the content owner, product development, maintenance and marketing, etc. The success examples are SK Telecom and TDC with their music service (see the part A in section 4.2.3.1 and 5.2.3.2).

However, this option is a high-risk approach for the MNO because the media services

(entertainment) are not the MNO core business. In addition, the potential benefits and direct revenue share from these services to MNO’s total revenue are not significant to justify the effort to deliver the service to the market. Moreover, the typical vertical system of the MNO’s organization impacts to the slower deployment life cycle, and their mass-market focus makes them slower to adapt to the emerging user’s demand and competitors, compared with the nimble culture of the OTT music and/or TV/Video service provider (The Pure player).

c. The “Collaboration” option

The MNO could choose to collaborate with other player (e.g. the Pure player) and get some percentage of shares from the total revenue, as return to the new users gained from their end-user based, the usage of their brand, and associated services (e.g. billing).

Our findings about the MNO and OTT player partnerships in section 6.4.1 show us that this

MNO typically set up the short-time agreement with the OTT player (non-equity mode) [15]. This means they can review this agreement and exit at the end of the term, if the partnership and/or the service do not take off or profitable. The MNO can reduce the effort to deliver the services and share the risk with this partner, compared if they choose the “Do it alone” option (see the part B in section 4.2.3.1 and 5.2.3.2). This is also confirmed by [17] that stated that the two critical success factors in multimedia business are company’s core competencies and user acceptance.

They can also take an active role in this partnership and create a respective bargaining

position in the value network. Using this, they may get a larger percentage in the revenue sharing and/or bulk discount in the cost to acquire the license (e.g. discussion between TeliaSonera and record label regarding its partnership with Spotify). Some operators also set up a joint venture (equity mode) with other player and/or become part of the shareholders of the OTT media company [15]. This strategy also offers a possibility of an active role in the partnership, and shares the effort to roll out the media service.

However, this demands a higher investment and riskier, compared if the MNO use the

short-term agreement strategy [15]. Our analysis in section 4.2.3.1 and 5.2.3.2, and the lesson learn from global market landscape, Spotify and other OTT media players’ business model

53!

lead us to the conclusion that, it is strategic for the MNO to use the point-to-multipoint model, rather than the point-to-point model. We recommend the operator disregard the first and second option, and instead form a partnership with other player using the short-term agreement under this point-to-multipoint model.

It should also be note that, through any of those available options, the MNOs should not use

these services to increase the revenue from the media. Our findings reveal that, even if the number of the subscribers acquired is high, the majority of the revenue generated from these services will go to the “core actors” (the content owner). In contrast with other businesses, in the music and broadcasting business, the provider only has choices regarding what kind of the content they would like to get the license to, not from whom they would like to get the license (e.g. Virgin Media fail in acquire the license from the major record labels).

Indeed, although it is still possible for the MNO to acquire some chunk of the revenue from

music and/or TV/Video service provisioning, the key benefits for the MNO are actually to: • Reduce the churn rate

This can be achieved by monetizing the higher value of the access to the content and feature in the service to user, which will be canceled once user stops their subscription with their MNO.

• Acquire market share

The market share can be increased by exposing music and/or TV/Video service to user and making it exclusively bound with the MNO’s brand.

• Increase the ARPU The MNO can increase the ARPU from the data by acquiring or turning the subscriber to the higher value subscription (the music and/or TV/Video service is bundled into it).

54!

Chapter 9 – End of Discussion This chapter provides the assessment of limitation and advantage points of this study, as well as the suggestions for future investigations related to the topic of this study. 9.1. Criticism of the Study

There are several limitations of this study. The first criticism is the expert interview that only covers small amount of stakeholders in the industry. It could be better for the study to cover more experts, so we can get a broader point of view and do in-depth analysis from their information. Nevertheless, due to the time and resource limitation, only the most relevant representatives that could be contacted for the interview.

The second limitation is the user survey, which is the based of the quantitative study. It only

covers small amount of respondents and there are even smaller valid data that can be used for analysis. The fact that the questionnaire is formed as an OTT survey also limit the respondent that can be reached, as naturally it is not convenient to do this survey in the small screen and limited connectivity from the mobile device.

The third limitation is in the limited analysis in the quantitative data. This study did not

conduct the regression analysis, which is very valuable to measure the most significant values that affect user’s decision to adopt the service. This cannot be covered by the correlation tests. Fourth, the survey itself only explores the acceptance factor for mobile media usage in mobile devices. In real scenarios, the media services might be consumed through other types of device (e.g. laptop), but still use mobile broadband as underlying connection.

However, there are also several strengths of this study though. First, our strategy to use the

mixed method approach, by using both quantitative and qualitative study, brings variety of perspectives to the study. This helps to enrich our analysis and ensure the robustness of the findings. Second, the reliability of the quantitative data was assured through several types of correlation tests, and there are more than one questions used to test the hypothesis through internal consistency.

Third, the questionnaire is distributed using Bahasa (Indonesian language). This removed

language barrier, helped to attract broader respondent, and make the questions clear to avoid risk of misunderstandings. Forth, the study covers the representative from well-known industry actors and is currently playing vital role in media/telecom business, both in reference market (Sweden) and target market (Indonesia). This help to ensure the results to has practical and influential implications for the marketplace.

9.2. Future Work

This study reveals several issues that can be subjects of further study. Indonesian mobile users generally have positive intention and perceive to use the OTT media services through their mobile device. In addition, the mobile device and connection is not a barrier for them to adopt the service. However, there is still little focus from the local actors, and especially the MNO, to fulfill and monetize this need. We have identified that these MNOs are currently playing dominant role due to high dependency among people to mobile, and low PC ownership and bank account penetration.

Indeed, It would be interesting to study this issue further in the economic and market-based

system point of view. One of the examples is predicting the outcome of complex interactions among actors, if the prominent (international) OTT players enter the market. The approach could be to do a micro-economic analysis, e.g. the game theory modeling. This can be used to predict and analyze the further forms of strategy and interaction between these identified actors in the value network of the OTT music and TV/Video business.

55!

Bibliography

Paper, Journal and Book • Business Model

[1] C. Zott, R. Amit, L. Massa. (2010, June). The Business Model: Theoretical Roots, Recent Developments, and Future Research. IESE Business School - Working Paper [Online]. WP-862. Available: http://www.iese.edu/research/pdfs/DI-0862-E.pdf

[2] A. Osterwalders, Y. Pigneur, C.L. Tucci. (2005, May). Clarifying Business Models: Origins, Present, and Future of the concept. Communications of AIS [Online]. 15. Available: http://www.softwarepublico.gov.br/5cqualibr/6-publicacoes-e-artigos/view/vetor-ecossistema/sobre-modelo-de-neg-cios/Claryfing-Busines-Model.pdf

[3] N.K. Elnegaard, K. Stordahl, J. Lydersen, T. Gunnar. (2008). Mobile broadband Evolution and the Possibilities. Telenor [Online]. Telektronikk (104) no $. Available: http://www.telenor.com/en/resources/images/063-073_MobileBroadbandEvolution-ver1_tcm28-36180.pdf

[4] E. Bohlin. (2007, August). Business models and financial impacts of future mobile broadband networks. Telematics and Informatics [Online]. 24(3), pp. 217-237. Available: http://dl.acm.org/citation.cfm?id=1266280

[5] P. Kallio, “Emergence of Wireless Services. Business actors and Their Roles in Networked Component-based Development”. Espoo: Technical Research Centre of Finland, VTT Publications 534. 118 p. + app. 71 p. ISBN 951-38-6386-7; 951-38-6387-5.

[6] Rappa, M. (2001). "Managing the digital enterprise - Business models on the Web," North Carolina State University. Available http://digitalenterprise.org/models/models.html

[7] J. Markendahl, “Mobile Network Operators and Cooperation - A Tele-economic study of infrastructure sharing and mobile payments services,” Ph.D. dissertation, CoS, KTH, Stockholm, 2007.

[8] E. Faber, P. Ballon, H. Bouwman, T. Haaker, et all., “Designing business models for mobile ICT services,” in 16th Bled Electronic Commerce Conf. eTransformation, Bled, Slovenia, June 9 - 11, 2003, http://www.mozaiek.dds.nl/docs/PanelBled03.pdf

[9] J.H. Chesbrough and R.S. Rosenbloom. (2002). The role of business model in capturing value from innovations: Evidence from Xerox Corporation’s technology spin-off companies. Industrial and Corporate Change [Online]. 11(3), pp. 529-555. Available: http://icc.oxfordjournals.org/content/11/3/529

[10] H. Håkansson and I. Snehota. (1989). No business is an island, the network concept of business strategy. Scandinavian Journal of Management [Online]. 5(3), pp. 187-200. Available: http://dx.doi.org/10.1016/0956-5221(89)90026-2

[11] A. Osterwalder, “The Business Model Ontology - a proposition in a design science approach,” Ph.D. dissertation, University of Lausanne, Lausanne, 2004.

[12] F. Li ans J. Whalley. (2002). Deconstruction of the telecommunications industry: from value chains to value networks. Telecommunications Policy [Online]. 26, pp. 451–472. Available: http://dx.doi.org/10.1016/S0308-5961(02)00056-3

[13] S. Bendahan, G. Camponovo, Y. Pigneur. (2003, May). Multi-issue actor analysis: tools and models for assessing technology environments. Journal on Decision Systems [Online]. 13(2), pp. 223-253. Available: http://jds.revuesonline.com/article.jsp?articleId=5378

[14] Y.L. Doz and G. Hamel, Alliance advantage: The art of creating value through partnering. Boston, Harvard Business School Press, 1998, ch. 1, pp. 1–32.

[15] Y. Pan, D.K. Tse. (2000). The Hierarchical Model of Market Entry Modes. Journal of International Business Studies [Online]. 31(4), pp535-554. Available: http://www.jstor.org/stable/155660

[16] P.T. Farda, T Wong, V.C.M. Leunga. (2010). Access and service convergence over the mobile Internet – A survey. Advances in Wireless and Mobile Networks [Online]. 54 (4), pp. 545–557. Available: http://dx.doi.org/10.1016/j.comnet.2009.08.017

56!

[17] B.W. Wirtz. (2009). Convergence processes, value constellations and integration strategies in the multimedia business. International Journal on Media Management [Online]. 1 (1), pp. 14-22. Available: http://dx.doi.org/10.1080/14241279909384482

[18] Teece. (1990). Competing, cooperation, and innovation: Organizational arrangements for regimes of rapid technological profress. Journal for Economic Behavoiur and Organization [Online]. 18 (1), pp. 1-25. Available: http://dx.doi.org/10.1016/0167-2681(92)90050-L

• Research Methodology [19] J.W. Creswell. Research design: qualitative, quantitative, and mixed method approaches.

London: Sage Publications, Inc., 2003 [20] W.M.K. Trochim and J.P. Donnelly. The Research Methods Knowledge Base. Ithaca,

NY: Atomic Dog, 1999 [21] P. Roberts and H. Priest. (2006, April). Reliability and validity in research. Nursing

Standard [Online]. (20) No 44, pp. 41–45. Available: http://www.sanpad.org.za/sanpad2011/index.php?option=com_phocadownload&view=category&download=95:reliability-and-validity-in-qualitativeresearchpdf&id=52:reading&Itemid=483

[22] J.M. Cortina. (1993). What is Coofficient Alpha? An Examination of Theory and Applications. Journal of Applied Psychology [Online]. 78 (1), pp. 98-104. Available: http://psychweb.psy.umt.edu/denis/datadecision/front/cortina_alpha.pdf

[23] J.A. Gliem and RR. Gliem. “Calculating, Interpreting, and Reporting Cronbach’s Alpha Reliability Coofficient for Likert-Type Scales,” in 2003 Midwest Research to Practice Conference in Adult, Continuing, and Community Education, 2003, http://scholarworks.iupui.edu/bitstream/handle/1805/344/Gliem+&Gliem.pdf?sequence=1

[24] D. Goerge and P. Mallery. (2003). SPSS for Windows step by step: A single guide and reference. 11.0 update (4th ed.). Boston: Allyn & Bacon

• Telecom Regulation [25] _. “Indonesia” in Telecoms and Media, An overview of regulation in 46 jurisdictions

worldwide 2012, 1st ed, Vol 1(1). London: Global Competition Review (GCR), 2012, pp. 237 – 244. [Online]. Available: http://www.gettingthedealthrough.com/shop/books/39/telecoms-media-2012/

[26] C. Wattegama, J. Soehardjo and N. Kapugama. "Telecom Regulatory and Policy Environment in Indonesia: Results and Analysis of the 2008 TRE Survey", available at http://lirneasia.net/wpcontent/uploads/2009/07/TRE_Indonesia_2009Mar18.pdf

[27] R. Lumanto, T. Kosuge. (2005). Telecommunications reform: Resolving performance problems in Indonesia. Technology in Society [Online]. 27, pp. 517–529. Available: http://dx.doi.org//10.1016/j.techsoc.2005.08.004

[28] D. Setiawan. (2010, Jan). Alokasi Frekuensi, Kebijakan dan Perencanaan Spectrum di Indonesia. (2). [Online]. Available: http://denysetia.files.wordpress.com/2011/09/alokasi-frekuensi-edisi-2-januari-20101.pdf

[29] “ICT Regulation Toolkit”, available http://www.ictregulationtoolkit.org/en/Sections.html

• Technology Acceptance [30] F.D. Davis, “A technology acceptance model for empirically testing new user

information systems: theory and results,” Ph.D. dissertation, Sloan School of Management, MIT, Massachusetts, 1985.

[31] L.D. Davis, R.P. Bagozzi, P.R. Warshaw. (1989, August). User acceptance of computer technology: A comparison of two theoretical models. Management Science [Online]. 35(8), pp. 982–1003. Avalable: http://www.jstor.org/stable/2632151

57!

[32] S. Hong, J.Y.L. Thong, K.Y. Tam. (2004, Nov. 19). Understanding continued information technology usage behavior: A comparison of three models in context of mobile Internet. Decision Support Systems [Online]. 42(3), pp. 1819–1834. Available: http://dx.doi.org/10.1016/j.dss.2006.03.009

[33] V. Venkatesh, M.G. Morris, G.B. Davis, F.D. Davis. (2003, Sept.). User acceptance of information technology: toward a unified view. MIS Quarterly [Online]. 27(3), pp. 425-478. Available: URL: http://www.jstor.org/stable/30036540

[34] H. Nysveen, P.E. Pedersen, H. Thorbjørnsen. (2005, June). Intentions to Use Mobile Services: Antecedents and Cross-Service Comparisons. Journal of the Academy of Marketing Science [Online]. 33(3), pp. 330-346. Available: http://www.springerlink.com/content/n15480270470w32r/

[35] M. Pagani. (2004). Determinants of Adoption of Third generation Mobile Multimedia services. Journal of Interactive Marketing [Online]. 18(2), pp. 46-59. Available: http://dx.doi.org/10.1002/dir.20011

[36] J. Lu, J. E. Yao, C. S. Yu. (2005, Sept.). Personal innovativeness, social influences and adoption of wireless Internet services via mobile technology. Journal of Strategic Information Systems [Online]. 14(3), pp. 245-268. Available: http://dx.doi.org/10.1016/j.jsis.2005.07.003

[37] R. Bamforth and C. Longbottom. (2011, July). Optimizing the online customer experience in Telecommunication [Online]. Available: http://www.oracle.com/us/products/applications/atg/quocirca2-opt-online-cust-exp-telco-520367

[38] E. Ponnert, Z. Rodriguez, “Video goes Mobile: -What Consumers Expect of Mobile Video Services.” Master Thesis, Stockholm School of Economics, Stockholm, 2007.

[39] M. Lindqvist, J. Svensson, “Mobile Multimedia Services – Surveying the preferences of young people,” in Proceedings of the 30th Information Systems Research Seminar in Scandinavia IRIS, 2007.

• Mobile Broadband and Digital Media [40] Y. Kim, T. Kelly, S. Raja. “Rethinking Broadband as an Ecosystem,” in Building

broadband: Strategies and policies for the developing world, Washington, D.C: The World Bank, 2010, pp. 15–25.

[41] G. Blennerud. (2010, July). Mobile broadband – busting the myth of the scissor effect. Ericsson Business Review, No 2, 2010 [Online]. Available: http://www.ericsson.com/ericsson/corpinfo/publications/ericsson_business_review/pdf/210/210_strategy_mobile_broadband.pdf

[42] M. Ljungberg, and A. Bolle. (2011, Feb). Mobile broadband second wave – differentiated offerings. Ericsson Review, No 1, 2011 [Online]. Available: Available:http://www.ericsson.com/res/thecompany/docs/publications/ericsson_review/2011/er_mobile_broadband_2nd_wave.pdf

[43] G. Kreitz and F. Niemelä. (2010, August.). Spotify - Large Scale, Low Latency, P2P Music-on-Demand Streaming. Proceedings of IEEE P2P'10. Available at: http://www.csc.kth.se/~gkreitz/spotify/

[44] G. Kreitz and M. Goldmann. (2011, Oct.). Measurements on the Spotify Peer-Assisted Music-on-Demand Streaming System. IEEE International Conference on Peer-to-Peer Computing (P2P). Available at: http://www.csc.kth.se/~gkreitz/spotify-p2p11/

[45] C. Krasic, K. Li and J. Walpole. (2001, Sept.). The Case for Streaming Multimedia with TCP. 8th International Workshop on Interactive Distributed Multimedia Systems (iDMS 2001). Available at: http://www.cc.gatech.edu/projects/infosphere/papers/iDMS_krasic.pdf

[46] A. Goel, C. Krasic, K Li, and J. Walpole. (2002, May). Supporting Low Latency TCP-Based Media Streams. 10th International Workshop on Quality of Service (IWQoS). Available at: http://www.cs.ubc.ca/~krasic/publications/

[47] P. Lopez. (2012, March 15). “Mobile video optimization 2012”. Core Analysis

58!

[48] I.L. Pitt. Economic Analysis of Music Copyright: Income, Media and Performances. New York, USA: Springer, 2010. Available: http://www.springer.com/economics/microeconomics/book/978-1-4419-6317-8

[49] Jansen, B. J., Zhang, M., Sobel, K. and Chowdury, A. (2009), Twitter power: Tweets as electronic word of mouth. J. Am. Soc. Inf. Sci. [Online], 60(11), pp. 2169–2188. http://dx.doi.org/10.1002/asi.21149

Report [50] “Cisco Visual Networking Index: Global Mobile Data Traffic Forecast Update, 2010–

2015”, Cisco, CA, 2011 [51] “Global Mobile broadband Traffic Report”, Allot Communications, Geneva, Rep.

D261013 Rev 6, 2011. [52] “The Future of Voice”, ITU, Geneva, Rep. FoV/00, 2007. [53] M. Newman. “Recognizing revenue opportunities from video services”, Informa

Telecom and Media. [54] M. Aguiar et. al, ”The Internet’s New Billion”, BCG., Boston, MA, 2010. [55] “The Mobile Internet Report Key Themes”, Morgan Stanley, 2009 [56] “Digital Music Report 2012”, IFPI, London, 2012 [57] N. Covey, “How People Watch: A Global Nielsen user Report”, Nielsen, 2010. [58] “OTT Music Stores”, Pro Music. [59] “Mobile Banking in Indonesia”, IFC The World Bank Group, Jakarta, 2010. [60] “Sensus Penduduk 2010”. BPS, Available http://sp2010.bps.go.id/ [61] “Indonesia Telecommunications Report Q2 2012”, BMI, London, 2012. [62] “ICT Facts and Figures. International Telecommunication Union”, ITU, 2011 [63] L. Abdo, “Digital Age Shifts - Roles of Record labels, Music Publishers”, Entertainment

Law & Finance. [64] “Digital Music Report 2007”, IFPI, London, 2007. [65] “Digital Music Report 2006”, IFPI, London, 2006. [66] “Digital Music Report 2011”, IFPI, London, 2011. [67] “Digital Music Report 2009”, IFPI, London, 2009. [68] C. Pettey, “Media IAS OTT Music Forecast, 2011-2015: Social Media, Subscriptions

and the Cloud”, Gartner, Stamford, Connecticut, 2011. [69] “The Digital Music Market Outlook, Evolving business models, key players, new

challenges and the future outlook”, Business Insight, 2009. [70] S. Pandey. “Mobile Music Market Overview”. Informa Telecom and Media, 2009. [71] “Digital music industry to hit revenues of $20bn by 2015”, Ovum, London, 2010. [72] G. Cottle, A. Blair. “Unlocking the Real Value of Mobile Music”, Informa Telecoms

and Media, 2010 [73] M. Scott, “Spotify shows that mobility is the key to successful ‘Freemium’ services”,

Analysis Mason, 2011. [74] “Allot Mobile Trends Charging Report”, Allot Communications, Geneva, Rep. D261013

Rev 6, 2011. [75] “TV and Media user Insights”, Ericsson, Stockholm, 2011 [76] “Communications Market Report: UK”, Ofcom, 2011 [77] C. Bachelet, “Digital music services: strategies for operators”, Analysis Mason, 2011. [78] Allot Mobile Trends Report H2, 2011”, Allot, 2012. [79] S. Rosen and J. Blackwell, “Over the Top (OTT) and Through the Middle (TTM) Video

- Advertising, Subscription, Rental, and Sale Markets”, ABI Research, 2011. [80] K. Nissen. “OTT Video: Content, Devices, Retail, and Pay-TV Collide”, In-stat. 2011. [81] G. Escofet. “Video will hog a third of handset traffic but earn less than 1% of enduser

mobile data revenue”, Informa, 2012 [82] A. Kishore. “OTT Traffic Control: Technology Options for Network Operators”, Heavy

Reading. 2011. [83] A. Kishore. “Operator CDN: Making OTT Video Play”, Heavy Reading. 2011.

59!

[84] “The Connected Archipelago: The role of the Internet in Indonesia’s economic development”, Deloitte, Sidney, 2011.

[85] E. Smith, “Indonesia: 3G and HSBB Network Expansion Create New Opportunities for Vendors”, Pyramid Research, Cambridge, Massachusetts, 2012.

[86] “Mobile Phone Penetration in Indonesia Triples in Five Years”, Nielsen, 2011 [87] “Asia-Pacific to Be Global Leader in LTE by 2014”, Pyramid Research, Cambridge,

Massachusetts, Vol. 3, No. 2, 2011. [88] “United Kingdom, France, Indonesia Among Cheapest for Mobile Broadband”, ABI

Research, 2010. [89] “Telecoms, media and technology M & A in Asia–Pacific: trends and prospects”,

Analysis Mason, 2012. [90] “The Digital Music Market Outlook: Evolving business models, key players, new

challenges and the future outlook”, Business Insights, 2009 [91] “The recording Industry 2006 Piracy Report”, IFPI, London, 2006 [92] M. Galli, “From OTT Video to Multi-Screen: Connected Home & Mobile Video Device

& Services Global Forecast & Strategy Review-2010-2014”, MRG, San Jose, 2010 [93] “Buku Putih 2010”, Komunikasi dan Informatika Indonesia, Jakarta, 2010 Article [94] A. Laughlin. (2009, June 16). “Virgin unveils new music download service, June 2009“.

Magazine [Online]. Available: http://www.digitalspy.co.uk/music/news/a160050/virgin-unveils-new-music-download-service.html

[95] “The music goes where you go”.FAQ [Online]. Available: http://www.spotify.com/se/mobile/overview/

[96] (2012, May 02). “Introducing the lean, green, music machine – Spotify for iPad”. Blog [Online]. Available: http://www.spotify.com/us/blog/archives/2012/05/02/spotify-for-ipad/

[97] “Use Spotify with your Sonos, Squeezebox, Oknyo, WD or Boxee Box system”. FAQ [Online]. Available: http://www.spotify.com/se/help/faq/devices/connect-devices/

[98] W. Mellor, F. Adi. (2012, May 1). “Indonesia Chases China As Middle-Class Consumption Soars”. Magazine [Online]. Available: http://www.bloomberg.com/news/2012-05-01/indonesia-chases-china-as-middle-class-consumption-soars.html

[99] (2010, Nov. 10). “TDC Makes History As The 1st Telecommunication Company Worldwide to Hit 250,000,000 Downloads Via A Music Subscription Service”. Magazine [Online]. Available: http://www.businesswire.com/news/home/20101110005799/en/TDC-History-1st-Telecommunication-Company-Worldwide-Hit

[100] (2008, March 31). “TDC: More than 30 music companies in the PLAY agreement”. Press Release [Online]. Available: http://tdc.com/publish.php?id=16268

[101] L. Farrar. (2010, Dec. 31). “'Korean Wave' of pop culture sweeps across Asia”. Magazine [Online]. Available: http://edition.cnn.com/2010/WORLD/asiapcf/12/31/korea.entertainment/index.html?iref=NS1

[102] J. Halliday. (2010, Dec. 6). “BSkyB closes Sky Songs music subscription service”. Magazine [Online]. Available: http://www.guardian.co.uk/media/2010/dec/06/bskyb-closes-sky-songs-music-service

[103] “What quality does Spotify stream in?” FAQ [Online]. Available: http://www.spotify.com/se/help/faq/tech/codec-quality/

[104] D. Murphy. (2012, Feb. 12). “Spotify Adds 320 Kbps 'Extreme' Streaming to iOS App”. Magazine [Online]. Available: http://www.pcmag.com/article2/0,2817,2400162,00.asp

[105] “What bandwidth does Spotify need?”. FAQ [Online]. Available: http://www.spotify.com/se/help/faq/tech/bandwidth-usage/

60!

[106] (2011, April 14). “Upcoming changes to Spotify Free/Open”. Blog [Online]. Available: http://www.spotify.com/se/blog/archives/2011/04/14/upcoming-changes-to-spotify-free-open/

[107] N. Patel. (2012, March 13). Exclusive: Spotify family plans 'definitely coming'. Magazine [Online]. Available: http://www.theverge.com/2012/3/13/2867880/exclusive-spotify-family-plans-definitely-coming

[108] (2011, Sept. 29). Pandora shifts into high gear!. Blog [Online]. Available: http://blog.pandora.com/pandora/archives/2011/09/pandora-shifts.html

[109] (2009, August 12). 7digital downloads now available in Sweden. Press Release [Online]. Available: http://www.spotify.com/se/blog/archives/2009/08/12/7digital-downloads-now-available-in-sweden/

[110] (2011, May 4). Spotify says hello to the iPod. Blog [Online]. Available: http://www.spotify.com/se/blog/archives/2011/05/04/spotify-says-hello-to-the-ipod/

[111] E. Harrison. (2009, May 2). Spotify review: The iTunes killer?. Magazine [Online]. Available: http://www.neowin.net/news/software/09/05/02/spotify-review-the-itunes-killer

[112] (2009, Oct. 8). Telia and Spotify sign exclusive cooperation agreement”. Press Release [Online]. Available: http://www.spotify.com/se/about-us/press/telia-and-spotify-sign-exclusive-cooperation-agreement/

[113] A. Miftachul. (2012, June 28). Yahoo! TNS Net Index: Advertiser Needs to Strengthen Their Mobile Strategy. Magazine [Online]. Available: http://en.dailysocial.net/post/yahoo-tns-net-index-advertiser-needs-to-strengthen-mobile-in-their-strategy

[114] S. Cuya. (2011, January 11). SK Telecom and PT. Telekomunikasi team up to launch digital music content in Indonesia. Magazine [Online]. Available: http://www.rcrwireless.com/asia/20110111/carriers/sk-telecom-and-pt-telekomunikasi-team-up-to-launch-digital-music-content-in-indonesia/

VIII!

Appendix A – Qualitative Interviews A.1. Interviewee List

Table 23. Interviewee List Interviewee

Name Actor Type Country Interviewee Role

Lars Roth MNO Sweden Vice President Consumer at TeliaSonera

Andreas Liffgarden

OTT music service

Sweden Global Head of Telecom business development at Spotify

Märta Rydbeck TV Broadcaster Sweden Head of broadcast managed services (Global Services business unit) at Ericsson. She was interviewed due to her previous role as Director of Distribution and Affiliate sales at TV4 until December 2011.

Jockie Heruseon MNO Indonesia Manager of Broadband Products PIM at Telkomsel

Didik Akhmadi Usman

Regulator Indonesia Member of BRTI, Indonesia telecom regulatory body

Yose Tireza Arizal

MNO Indonesia General Manager of Business development and support system at Axis Telekom Indonesia

Hengky Philip Ginting

Mobile content provider

Indonesia Director at New Solution Technology (NST). Hengky Philip Ginting was also interviewed due to his previous role as Head of Business development at Mobafone Indonesia

IX!

A.2. Interview Transcripts and Notes

Table 24. Interview with Lars Roth40 Question Answer

What is the main motivation behind collaboration of Spotify with Telia?

To be able to talk about the value of our products instead of talking about technology, it is one thing. Because if you say to some you going to get an MDU packet for your phone, they would like 'why?' But if you say, you get free music, bundle in our offer and MDU is also part of this offer because you need that in order to make it work, (it is) become more none technical way of saying it, and more on customer value way. So we want to actually to have marketing and communicating things that are not speed, products and technology, but rather focus on the end user values. So that’s why we were lucking at this to bundle with that. And looking at something to bundle with, it has to have value and it has to broad enough, (because) we have 44% of the market, we can just choose one small game, for example. Maybe it could be books, it could be music, it could be movie, something like that, it has to be that big.

We actually started with our own service; much more like Spotify, unlimited download for music, and 99 kr per month. It was under brand, but it was not great. We tried that for half a year and we realized that we needed to spend all our communication explaining, “What the service was”, “How it works”, and “why you should get it”. (At the end) the perception of the value that we gained from (offering) six months for free was pretty low. What we realized from Spotify that they came from nothing and conquered the market for half a year. They have a super strong brand, we do not need to explain anything, just said Spotify and people got it. So we realized that we can get a lot of values of it and we can actually move our brand little bit, be more interesting and dynamic by connecting our self with Spotify. So that I would say, that is the main thing. Also in this collaboration, we have exclusivity agreement that no one else can bundle it in our market and communicate Spotify within our territory and segment. This agreement was only in Sweden, but we have now similar agreement in Finland, and we are working in other countries as well.

And that part, we call it compiling difference, what differentiate our self from our competitors is quite important, because it is quite hard to find this differentiation factors, and that is why it is also very strategic. It (bundled product) has to be relevant for a lot people, and obviously can be bundled. Example [a name of operator] they campaigned last summer that you could choose between music, audio book, and movies, for example. But the message is not clear. And I think music is actually the most relevant for most of people, you can see that. In some segment, activation of Spotify is not that big, so probably we are looking at audio book... You can do other bundles, but if you do the marketing game to differentiate our self, it needs to be very wide, when you such a big operator as we are. If you are

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!40 Transcript from recording

X!

small operator, entering the market, you can pick small thing or market.

Who are the other actors involved in this partnership?

We started out the collaboration three years ago. So actually just between Telia and Spotify, but we renewed the contract during the autumn. And in that partnership, in order to... what we actually want to do was to deepen the collaboration. We saw that when we just gave 3 months for free, the value was limited. It was nice for the customer, but it did not really get them to move from another competing operator to us. We realized, we need at least 6 months or probably a year included, to be good enough, to (make customer) move from [a name of Telia's competitor] to Telia, for example.

So in order to actually afford this, we can't really afford the price we had per month that we bundle in the allowance contract. That would be too expensive for us in our accusation of new customer. So I more and less went to Spotify and said “OK, I want a massive discount, but in return I will put quarter of billion kronor in the table, are you interested?”. That is substantially increasing the whole business in Sweden for music. Obviously Spotify can't afford discount I asked for, so they need to involve other partners in the chain further behind, that would be the record labels, the collecting societies and the publishers. They would have to give Spotify the discount for us to accomplish the discount we asked for. So in order to make this happen, we have had extensive talks with all the hands of record labels in Sweden and US, and also all the collecting societies to actually the partners that own the music, the record labels own the recording, but not the song and lyrics.

So we have to talk with all of them, and obviously they can't invite them in one room. You have to take them separately, and they have different opinions about different things and they have the most favorable kind of deal. So the hard part is to get something that is common, because you never have them in the same room. And obviously has their own wisher, we have our own wishes and the music business wants to control everything, how we do it, how we report it, what customer can have what, and the big problem is that when we talk about such a big thing and such discount, they come to our territory, as they call it hard bundle. They want you to make it available to everyone you will sell it too, no exception. On the other hand, they have a lot of specific requirement on how and to whom you can bundle it, so with this they actually more and less control all our offerings. But in order to make them stand a bit, and make a room to maneuver, it is a hard work, with all the details.

It would meant that for example we would have to differentiate to what kind of offer to existing customers and new customers. And that is not acceptable for us, because we know, if we do it, we would start churning customers, for example. It would mean, we would have to say in our commercial state that you can get Spotify one year for free if you have got it before, for example. Because they do not want to cannibalism demand from high margin to low margin. So those kinds of thing, we have to get through. We were almost finished with the contract with apple, as you might be aware… They did not make reality out of this thread, more and less but we have got in the contract, that if

XI!

we have risk thing and if this become reality, we will get out from the contract, I guess it isforce major thing. But big problem is that to make this happen I have put all the money on the table in front, so I have already paid it, and it is already distributed to the artist, and so forth. So if this force major actually happens, they have to go back to all the members and say 'Hei, pay back'. So to make this happen it needs 2/3 months negotiation on how to make this goal in the contract.

Normally, the record labels we deal with are OK, with one exception, and that is [a name of a major label]. We try to base this on win-win kind of thing. “If we build our customer based to you (music actors), you will make a lot of money, long term, and it would be good for us… If we have a low price and not tying your hand on our back, we can use this in all our accusations, regardless the products and segments”. They agree in theory, but anyway in practice they try to tie they hands behind our back and you know, try to not give us the possibility to bundle 6 months, because then you will get so much values out of so little money, so that is kind of fundamental difficulties we have in this collaboration. We truly believe in win-win but they do not, they say they do, but they do not.

It is so strange, because the have full monopolized, all the record labels, because If you have Michael Jackson, no one else can compete with other Michael Jackson. You own it and you have paid for it already. You own it; you have paid for it already. Your catalogue is filled, and you need to just make out money out of it. And you will try to maximize everything that comes to that stream. And it is quite difficult to handle. This is primarily legal discussion, but based on business discussion. I would say we probably have the highest discount that everybody ever had from the music industry. So that is big, and it is definitely bigger that Spotify ever got. On the other hand, if I look at the marketing that we had during 3 years, for most of our business areas, and also taking consideration discount that we have by using media, that actually has market value within, would be 3.5 billion kronor spent in Sweden alone, on marketing Spotify. So it is obviously a huge impact on the whole (music) industry. The interest on music, also Spotify for their-own.

And looking at Spotify now, their main target is not actually to have an increase premium customer, they want to build customer in general. Because that is the fundamental in deciding the value of the company. If they do the IPO in 2 or 3 years, if they have 200 million customers, it would be the biggest in the world, then you get value for it. But from our perspective is premium based for them... and the conversion rate from free to premium is extremely good in Sweden. I think they have closed on a million paid customers in Sweden alone. And they though the room for half a million, and they still competing. So if nothing happen, my guess we could grow up in Sweden before the market fully penetrated.

We spent 14 months (negotiation). But on [a name of operator], they started their own, but ended with Spotify. They spent 3 years negotiating. This is the big problem. You cannot really negotiate if you are not in the level of where I am (Telia). Because then you have to redesign your businesses during the contract... But you can do,

XII!

something without big discount, but kind of partnership, but they will tight you hand in your back and it would quite hard to get profitability. That you can do quickly, but it has limited affect. Maybe KPN can copy us actually, they were negotiating with them. I think this is actually quite fundamental to get working. So quite risky, quite complex, quite time consuming, but for us it is rewarding as well.

What is the revenue sharing model used between these involved actors?

What they do is they have a rate card, so typically 5-6 euro goes to labels, maybe 1-2 Euros go to publisher and collecting society, and the rest is Spotify. So, you have statistic saying what type of music that customer visit. So if I am [a name of record label], in general have 40% market share, so I as customer with 70 % of my stream goes to [a name of artist of a record label], and then 70% of that 5-6 euro rate card goes is given to [a name of record label] and they distribute it to the artist. We just partner Spotify in our contract, but we have continuos conversation especially with record labels, and we have relation as well, especially with the marketing activities and sponsorship, and so fourth. So I would say, if we have in Sweden 100 kr, then I would say somewhere around, I cannot say you the exact figure, lets say 40-60 kr go to labels, and small part, typically 12 % of the end user revenue goes to collecting societies.

How is the music industry players interact in this cooperation?

Our agreement is only between us and Spotify, but in the agreement it is stated that we should have regular meeting with all the record labels to see what kind of activities we could do to live their artist their music, but also for us to find how to help the music industry.

What value do you think Telia, as a mobile operator, add to this mobile music streaming service and vice versa?

In order to get end-to-end delivery to their service they need a good network, and I think we are the best, especially if it comes to fixed and mobile broadband. So one of the reasons. The other big reason is, what we have done to give trust to new cool company is enormous. TeliaSonera is the most trusted brand overall in Sweden. So, when we pick up partner to go with us, we more and less say this is availed partner that you can do business with. And also, since we are billing provider for them, we charge our customer and get the money to them, also our bills are one of the most paid in Sweden, with lowest dept as well. So it is really good channel to have continuous relation with customer, especially when they have problem when they do that with credit card, so if you do not have 99 kronor in your account, you will switch off, and we have to deactivate you as customer, you do not get up if we are not involved.

What is the major value proposition that Spotify shares with Telia?

Well, in order to make streaming service available and work, they value propositions are then connected quite a lot. And I think in strategic level, we say that in order to be able to a service company, we should put our customer experience very high. We could not do that really in reality, unfortunately, but Spotify really focus on experience, from that perspective, it is quite aligned. We had similar service that on the paper would be the same offer. But the

XIII!

user experience is on the brand. So brand is extremely important, if they do not have a good brand, we see in other market when they do not have that strong brand, for example in Denmark, they just launched. We need to have a totally different plan to communicate them, because they are not quite known. They were not 100 % awareness when we launch, so it was really problem to say what Spotify is.

What do you think are the obstacles when music-streaming service immigrates from desktop to Mobile device?

The biggest obstacle in day-to-day experience is that it does not work well in the subway, for example. It requires quite a lot of bandwidth. Event ought the mobile network in Stockholm, Sweden is one of the best mobile networks in the world, and it is not sufficient for continuous online good experience. You have to use caching, online mode. We also realized, in some spot you need to bounce, because it is too crowded. Telia do not have priority for Spotify traffic. If we give that, PTS would consider us to not being neutral, and we cannot do that. So we do not any differentiate any traffic what so ever. We maybe see the P2P traffic, but other than that every service is threatening the same.

What are elements of consideration for the pricing and bundling?

For the bundling, we need to make the value strong enough, to make people choose us, and that we realized 6 months… We have tried with 29 kr per month but it was not quite successful. How we end up with 99 kronor, I think it isquite interesting. Because the psychological level is extremely important. Looking at the prices around the world, it is always 99 kronor Norway but toward us (Sweden) it is actually 20 % more expensive in Norway. In Europe it 9.9 euro, which is actually 10 kronor cheaper than in Sweden. In the US, it is 9 dollar, and that is 60 kronor… So that is why, if you say 108 kronor, the perception would be more expensive. So psychological is important, more and less.

In long term, I would guess the price would be higher if you get 100% penetration of the country, then you can multiply the population by 99 kronor, you see what is maximum revenue is in that country. And that is problematic, because all businesses want to grow, and it senses the cap. I think we would see new type of subscriptions, price increase, differentiation, probably when you get access to new content, so fourth.

We are looking off somewhere else cases (become part of OTT business). One example that we distribute concert in video on the demand service, and we do things with their artist, which is actually what the labels would like to sell to us, like using song of artist our commercials… That is what they are looking at. The mobile wallet is actually what we are looking at as well... We are looking at different things as well, but we try to keep our core.

What are the elements of Spotify’s service marginal cost per subscriber?

Spotify consumes somewhere in between 100 GB a month. So that we have production cost of that data (sending the music content on the network), it cost us bad debt, in average we have around 1% bad debt where we do not get paid. We have cost teaching them in the store, administration and getting started, the customer care. Typically we (our customer) are always in the bundle, so if you own 1GB of data (allowance), it cost you the same regardless of what

XIV!

you consume. And if you do not have Spotify, you consume less. So, that is obviously cost us as well. It is quite substantial traffic that is Spotify related on the network

What things need to be modified if the model similar to Spotify-Telia partnership is used in these 3 markets?

If you are looking at exactly launching Spotify in those markets with the same kind of condition now… I mean to start to make it get to work, the willingness to pay for music on that level, needs to be there. Otherwise, it would not be success, because they (consumer) would only get the free (offering) and then they quit. And then they (label) would not interest on giving you any discount, as they want a continuing state. So that is obviously the first lesson, to understand what level is actually possible there.

The other is, as I mentioned, the reason why music is relevant for us, is that we are that big, we need to address big part of that market. If you only just go to 0 – 5% of the market share, you could pick a small specific segment and try to get know what they exactly need. You could pick some thing else to work with, maybe audio books, educational courses... But the core thing is you make your service relevant by simplifying the value, the customer value has to quite clear, and also be consistent in over time within your message that this is why you should choose us. What we are doing more and less sending the same message for 3 years now. Now people start to understand and remember that you can get Spotify from Telia, it takes couple of years. And typically, if we change our message after few months, people will forget.

What are major consequences for Telia to get involved in the digital music industry, in term of network load perspective point of view?

We (Telia) did not think so much about what are the effects could be. In the first contract, we promised 20 thousand customers and we have 350 thousand customers... We know that in general our 3G networks is really congested especially in big cities, and this due to a lot of things, apps communication, not necessary data volume, but rather how much they communicate... We are not talking about problem in our network due to Spotify, but we realize that it is quite extensive part on the traffic, not the major part, but big part.

What kind of network capabilities in terms of billing, traffic and user management can improve end user experience in this service?

I think if you ask customer, the value more of a really good network because they need to have access to the music regardless of where they are. Therefore, from that perspective, it will add the perception of the quality to the customer. We have big problem three weeks ago, and it was difficult to know whether it is network problem, IOS problem, Spotify problem, or Facebook problem. So we spent many hours figuring out what was the problem, and it was Spotify problem connected to Facebook, and we though it was our problem. So that is quite complex if you want to have a good user experience. So that is also needed when you have collaboration, in order to actually digging to what is not working.

XV!

Table 25. Interview with Andreas Liffgarden41 Question Answer

What is the main motivation behind collaboration of Spotify with telecom operator, e.g. Telia? • What are the major consequences after mobile operator get involved with the digital music industry?

The telecom industry is now seeing and over regulated and over competition in their market, and price eruption. Thus, there is a need for differentiation. Telia think music fits with everybody and it can be used to create this differentiation. In addition, in Telia’s case, there is discount arrangement. But, Spotify does not have data for conversion rate due to partnership with Telia yet

What are the benefits of Spotify’s service compare to the other incumbents, or future services that might use the same value proposition and chain?

Spotify has user that is sponsored to access the service for free. However, Spotify also needs conversion rate to paid-based. Telecom operator can distribute the bundle offering for trial period. They need to have aggregator partnership to delivery music service, because the direct partnership with the music label has been found to not work (first generation model). There is a cultural difference between the music industry and telecom industry, and Spotify as aggregator bridge this difference.

What do you think are the drivers and obstacles when Music Streaming service immigrates from desktop to Mobile device?

The network bandwidth and intense communication. Spotify has offline mode feature, but actually they found that the consumption in mobile device is much more from Wi-Fi rather than mobile network. Spotify tract user behavior, and guarantee user experience. Meanwhile, telecom operator mange the network

With service launched in 13 EU countries and US, does Spotify consider to enter Asian markets (e.g. China, India, and Indonesia)?

• If yes, what are the things that you consider to modify in order to be successful in these markets?

• How does Spotify see premium vs. freemium model in these markets?

New market is a driver but also challenge. There are still a lot of rooms to expand, e.g. middle east. Spotify considers Internet penetration, smart phone penetration, and Facebook penetration to expand to a new market, as well as willingness to pay of the user in that market. Spotify is the answer for piracy, and user does not like DRM-locked content.

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!41 Information is based on the note made during the interview, not recording

XVI!

Table 26. Interview with Märta Rydbeck42 Question Answer

In 2010, TV4 has launched a package of four major channels available to subscribers to the Telia, 3 and Tele2 mobile networks • What are the significant values

to deliver TV4 program as mobile TV services? Why do you think that so?

• What business and technical values does TV4 get from the mobile operators in this cooperation?

• c. What other distribution channels are TV4 using right now? And how is the priority of mobile broadband?

I would like to point out that I am no longer working for TV4. I think this my opinion; this is not TV4 opinion anymore. So it is from someone that has working in the industry for long time. It is my point of view.

If you look at distribution (model), it does not really matter the device you are accessing your TV content. For TV channel, it is always a question of the negotiation with the operator. How much is the operator is going to pay, because the content would always drive the accessibility to the network. In addition, this has been the whole battle in the business. When I started 93, at the Eurosport... Eurosport was the first channel to ask distribution revenue from the operators, all channels until then have been free or charge, or even (several) channels have to pay to get distributed. So we reversed the business model at that time, and said “If you want to bring content to your subscriber, you have to pay for it as an operator”. This was such a big thing, and they were very upset with Eurosport at that time. They threw Eurosport out of the network, and you will not get distributed anymore. But the thing is people like the channel, they wanted the channel, so the operator came back and said “OK, how much do you want?”. The negotiation started.

And I think, to deliver TV4 programs or TV4 channels on the mobile services is the same thing. We are delivering content that is the significant value to mobile TV of mobile operator or mobile operator delivering TV. Look at [a name of operator], for example, they would obviously love to use the TV4 channel in delivering TV everywhere experience, where they have it on mobile. “You can have it on the PC, you can have it everywhere”. But you know, I think what we are delivering as a TV channel is consumer experience, it experience in watching TV, that is the significant value. Whether it is in set up box at home, on a PC, or mobile phone, it does not really make a difference.

(For TV channel) it is (mobile channel) of course revenue stream. It is a value, but it does not bring more value to watch it on the mobile rather than set up box. Actually currently, it brings less value, because the TV channels haven't sort it out how to deal with advertising on the mobile yet. So actually, an increase of mobile distribution for TV channel is actually not very good for TV4 as advertising vehicle today. Because if I watch TV4 here, and I watch an advertisement that run on the TV4. It does not get calculated as a view, I do not get calculated as viewe arer. Because there is no measurement, system today that takes into account the advertising that is broadcasted here. Because all the measurements set up today are at home (traditional)… So it isbig problem, if you look at the increase of both on demand viewing or linier type of viewing on mobile today for TV channel. So people start to switch from watching at their home, to watching only on their mobile and we are do not sort this out… The TV channels are facing big

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!42 Transcript from recording

XVII!

problem. That is why the TV channel, at least TV4 and also Eurosport said, “OK if you want to distribute channel on the mobile, you better pay us a lot more than you do when you have it at home”. So obviously, they did not really agree with us, because they think… I mean, “You already broadcasting and we are just put it in another device, so why are you bothering us with this”. And we are we were saying, “Of course, this is really important.”

So what are business and technical value we are get from mobile operator in cooperation? We are do not get any business or technical value at all today (in delivering content on mobile). Even on the PC today, if you watch live streaming at TV4 channel, you cannot really calculate the advertising and cumulate it with the advertising that you from normal viewing, in the set up box... Because those measurement systems are not correlated. You have to differentiate it with the linier TV advertising, the 7 minutes break... and what you see on the PC, where you have we areb advertising… The one-minute break, but those advertising are not the same advertising that you see in normal linier TV broadcast signal. So if you watch TV at home, you get the 7 minutes break, if you watch on PC, the same program, you will get the same program but with different break and advertisements. So there is huge challenge for all TV channels today to get the linier TV broadcast synched with we areb advertising. Because they are not measured the same way, you cannot cumulate the audiences… And you add the mobile, and you cannot measure anything today… This is a priority for the TV channel to actually able to handle this kind of advertising here (mobile).

(But) I always see this (mobile) as the next leap of revenue for TV channel; in the sense that today that the TV channel charges for operator is per house hold (not individual). So a typical [a name of operator] would pay, for example, for [a name of TV program] [a number of payment] per month per subscriber. And subscriber is a household. Obviously, the operator says “If you want to charge us for this (mobile), you have to divide the household price by 1.2”. And we ware saying “No, we are can increase the price”. So it is next gigantic leap of revenue for TV channel to go from household subscription to individual subscription. Obviously operator does not agree with this, because they said “It should be 3 times less, if you charge individual subscription”. There is a lot to be done, in the negotiation between the operator and TV channels today. When you come to, what to charge, how to charge, to whom, and what kind of compensation is that for the fact that there is no advertising revenue connected here (mobile), or can be measured. So, when it comes to other distribution channel, the priority is of course toward more traditional distribution channel. But we, as a TV channel, always interested in providing the content where the consumer is. So if the consumer is here (mobile) you want to provide the content here, but we are need to fine the right business model. I am not saying TV4 can only survey on advertising, because already all the channels are financed both distribution and advertising.

The TV channels do not want to re-do the mistake they did at the beginning, which is being controlled by distributors. The fact TV4 wants to distribute here is more to want to get upper hand of toward the mobile operators. Because mobile operators, they want to get real TV content here. And there is a right issue… TV4 has gone around to

XVIII!

get collective agreements, together with [a name of collecting society] and managed to clear the linier right for mobile. Meanwhile, all other channels are not dare to do this, they do not dare to pay to clear the right, because they do not think they would get sufficient revenue from operator for this. But TV4 was sort of beating, because this is attractive, in the sense that the operator will have to pay the cost for the clears the right. Because to clear the right for mobile is not free of charge. There is additional right to be cleared. And we are managed to do that with [a name of collecting society], which is collective agreement in Sweden, but for lots of channels it is not possible to do that... We are managed to these rights, but not going to the right owners, saying to have right to distribute on mobile device. Because 3G, it will be one right, 4G it will be another right. It would not surprise me if there would be a Hollywood lawyer say, “If you want (to distribute the content in) LTE, that would be 100 thousand dollar per hour”. So obviously, it is difficult today to understand what are we are really distributing and to what devices. Because a PC that is connected to wireless network is a mobile device, but we are have right for PC.

Who are all the players involved with TV4 in providing contents on mobile broadband?

For the other distributions, we are clear the right at the source (content owner). (But in mobile) we pay some of the fee to [a name of collecting society in Sweden] for distribution, for rights that cannot be cleared. We are cannot clear it by our self. Rights that have been transferred by the owner to exclusive organizations. And the only organization that can negotiate with TV channel in Sweden is [a name of collecting society in Sweden], the collective agreement. So there are small additional rights.

So in order to distribute TV in Sweden you always have to have agreement both with the right owner of the content, but you also need the [a name of collecting society in Sweden] agreement, if you are Swedish channel, at least. TV4 has decided to take that cost for [a name of collecting society in Sweden], but normally it should be the distributor as well. Distributor should be paid in redistribution of the content. But we are paying [a name of collecting society in Sweden], and we are re-charging the operators for this cost, of course. Although they do not like that.

And for technical standpoint, we are have changed the distribution. In that sense that, I think in the beginning when TV4 took mobile distribution, they made specific mobile fees for the operators. Then we are said, “If you want to distribute the channels in your mobile network, you have to do what all the other operators are doing, take that by satellite. So there is no specific mobile feed going out from TV 4 in that sense”. So they receipt the channel by satellite, encode them and decode them by them-self. So it is different

There is another thing I was thinking, if you looking at the specific of TV4 which is very challenging when comes to mobile distribution, is that fact that... you may now TV4 has 30 regions, with local advertising in 25 of them... Actually it is 25 editions, with 30 advertising region. So we are get regional advertising, mixed with the national advertising all over Sweden. And local advertising is very profitable for TV4 and it is growing business. The local advertising market for TV is huge in Sweden. Where the only competitor in term of local advertising today is the

XIX!

local newspaper or local direct mailing, and so on. And if you look at TV4, how do you manage that on mobile distribution? To get which feed should be received depending on where you are in the mobile network… You have to distribute 30 versions (of local advertising) in the network. In mobile it is not done yet, you could do it. You have Telia's network, 3's network, Telenor's network, and they each have to carry 30 versions of TV4, and if they want the HD version, it would be another 30 versions, and then you have other channel, that is called TV4 plus... it is also 30 versions. And that are a lot of version of TV need to be distributed in the network...

In term of mobile broadband distribution, I do not know how they manage that. Because, as TV4, we are do not have to replace that. It is (local advertising) considered more valuable, more expensive, and more (delivery) times... At the end, individual advertising that is requested by individual viewer is much more valuable than group, but we are so far from that kind of individual distribution of TV, still. That might be happening in 20 years. So there are a lot of technological issues to be though. So this is why TV4 is not pushing greatly in term of mobile distribution. There are actually a lot of TV channels... that have regionalized versions; they are facing the same (unsolved) problem.

What supporting systems and supporting business roles are needed in TV4’s mobile TV business?

Measuring who is watching, that is really needed. Currently, a lot of distribution (channels) is by the operator. So obviously the operators are not so happy with TV4 extending beyond their managed network, going over the top… And TV4 play premium that exist in Sweden, where TV4 builds its own customer is huge threat for the operator. They obviously want that content in their own network, in label [a name operator in Sweden]. They said, "We are can have TV4 corner, but it has to be (under) [a name of operator in Sweden]". We said, "Listen, if that is in your platform, we are do not get any of the advertising... I mean we are talking PC distribution" (digital online distribution in fixed broadband). It needs to be played in our service, so can calculate on the advertising that we are have in free channels and we are can have billing relationship directly with the customer". I mean, "You could build for us if you want to, but still they still have to get access to TV4 by their own, not within you services".

If you look at on the mobile distribution, where is the advertisement land really? I mean how can you monazite it. The advertising you see here, it could be the same as in the traditional TV, the 6 minutes break or you could take the web stream. But then how do you calculate that one? The huge challenge for TV channels today is to be able to add all these viewing together in the same measurement system. Obviously, one the question of having it (TV program) as an app (OTT application) on [a name of internet player's application store] is, TV4 does not want to give 30-40% to [a name of internet player's application store]. So do not want to share your revenue to some one who does not really bring value in that sense. So that is obviously one of the reasons (to not launching channel as application), but I think you can go to browser, by logging to TV4 play premium. So that would be mobile web stream, not mobile (application stream). And in the mobile packet we are still do not resolve the issue of advertising.

For TV, I do not think it is valuable (to add banner in display while viewer watching the content), I think that

XX!

would be disturbing in that sense. I also think there is a problem on the rule of what you can do with the video, in term of overlies on the screen, when you buy the content from the content providers. There are very strict regulations on what you can do with the TV content. You cannot do overlies; you cannot be allowed to do a lot of things to be on the screen. It could possibly, you can reduce the screen, but the think is that how you manage that, how do you monetizing that from technical point of view. There are already today on the PC, if you look at [a name of TV channel in PC] they have some extra (banner). Probably you can do that as we arell on mobile TV, but you cannot allow the operator to take signal on satellite. Then you have to manage kind of signal sent out from TV4 connected in different way. Today, if you look at mobile TV reception it is not a manage signal in that sense. And yes (you do not have that much screen to work on mobile), the think is that contrary to Spotify, here you actually watching the screen, and in there (Spotify) you are listening.

If TV4 operates in the emerging market countries (e.g. China, India, and Indonesia), what do you think about the VoD and streaming service options? • As content owner, what strategies are TV4 using to address the piracy problem from Internet?

I think it would be really important... I would understand, that maybe you will leap some of the distribution channels that we are have in Europe, and you will go probably direct to the mobile, rather than set up box and traditional house hold viewing. You will maybe go faster. I suppose the VOD and streaming services will be much more important in the emerging markets... because you really need to be able to provide a full service in mobile network in a different way. I am seeing in the countries, for example France, where you go straight to IP TV, they have been very much advance in mobile TV and multi screens solutions, much more than what are currently available in Sweden, which is much more traditional cable market, or Holland, or Germany.

I think people do not get use to pay for content. The whole TV business is actually people paying for technical access. That is what the cable operator has been selling, not the content. So you get paid for (providing) technical access to something. There is no worse market for cable operator in term of content. When you look at the telecom operator today is that they do not charge the right thing for the video today. There is a huge problem. When I was at TV4, we were having some scary discussion for [a name of operator in Sweden], about who will pay for the content distribution in the network. And they said, "You have to pay as TV channel, because you want to get your content out to the viewer". And I was saying, "No way, we are not will pay, why we are have to pay? You are the once who have to charge, the customers are willing to pay, they want to pay for the content. You have to make sure they get the content, and you have to find the right business model. And you have to continue paying us, we are will not pay you for this". But the problem is that, TV channel may not as strong in the future, because there is propitiation in the content (distribution). They are filling threaten in the traditional value chain, when content aggregators and distributors, and devices (manufactures) take that video value chain, there are too many people are jumping from aggregators, and jumping from distributors and go the end customers, and so on. So TV channel may feel threaten, not feeling so secure when negotiates this deal.

XXI!

Table 27. Interview with Jockie Heruseon43 Question Answer

Worldwide market now sees change in the revenue structure of the digital music. The revenue from access-based service (e.g. on-demand / radio streaming) slowly chases the revenue from ownership-based (e.g. DRM/free DRM à-la-carte). While Indonesia is one of the few markets that still see the revenue from digital music market is occupied by personalized-based service (mainly RBT), which is 95 %. In term of demand, supply, and policy enforcement, why do you think this happen?

He confirmed the revenue generation from the music business is occupied by the personalized-service (mainly ring-back tone). Telkomsel does not see any prominent OTT music/video service (like neflix in north America market, or Spotify in Europe), thus their deliver own music service brand called 'Langit Musik' (Service model: WEB/WAB/SMS/UMB, RBT activation, wallpaper etc and download based DRM-protected. Revenue model: purchasing, e.g. unlimited download / week, paid through customer's mobile account ('mobile pulsa', 'tcash'). Even tough Telkomsel campaigns for unlimited download, they are actually trolling the speed, and the regulator does not regulate this.

Seeing the performance of revenue generation from download service and weak law enforcement for digital piracy protection in the country, do you see the streaming-based music service, especially in the mobile broadband market, can take off in Indonesia compare to download-based music service?

He was not sure the streaming-core service is provable business model yet in the near future, as the minimal law enforcement becomes significant barrier for any players to monetize music trough this channel. Thus, there is very low focus to provide the download-core/streaming-core service, so that the players mainly bundled the service trough their personalized-core service channel. However, he agreed there is a need to have clear picture of the demand-side through customer survey, specific to measure technology acceptance for media service (music/video streaming/download) on mobile broadband.

There is consistent pattern in the competitive strategy that the operators worldwide take, to positioning themselves in the digital music market business: rolling out their own music service (point-to-point partnership with music industry), or collaborating with the OTT player (point-to-multipoint partnership). We identify a trend that, the operators those previously went to direct-partnership struggle to make its service sustainable, and then switch to point-to-multipoint. And Telkomsel is currently running

Telkomsel's revenue from providing the content from music personalized-service / RBT (VAS) still cover the overall cost to provide the service / content to the market.

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!43 Information is based on the note made during the interview, not recording

XXII!

its own service music called “Langit Musik” and considering the fact that only 5.5. % Telkomsel’s revenue comes from digital music. What is your view about this strategy in the cost vs. benefit point of view? Who are the other actors involved with Telkomsel in rolling out “Langit Musik” and “Lipstic TV”? And what is the revenue sharing model that Telkomsel currently uses with these involved actors?

In personalization service, Telkomsel handles the marketing and branding, as well as product platform development and maintenance for their music service. However, the contents licensing for merchandizing is handled by the local content providers partners, means Telkomsel does not deal directly with the music industry actors for that.

In line with the mobile broadband development across the country, how does or will Telkomsel position itself to address local / global OTT TV/Video player?

Although Telkomsel built their own brand for the music content delivery, they have initiated the partnership with the OTT players for other service and have revenue sharing for that (e.g. VoIP with Skype, IM messenger with Nimbuzz). He agreed Indonesians have much more tendency to open for the trend from International media (compare to South Korean, China, and India market where local music/video contents are so popular). In addition, in order to react on the possibility of the international media OTT players penetrate Indonesia, He said: Telkomsel might choose to partnering or competing, this positioning decision might chance trough time and kind of service, depend on the OTT popularity in the market, etc.

We identify that these operators use digital music channel, not directly to increase ARPU from data-service from music service, but to differentiate themselves by delivering it trough unique bundling offering strategy using point-to-multipoint partnership with OTT, that at the end to decrease churn, increase blended ARPU, and acquire more market share. Do you see this strategy apply to Indonesia’s market?

Jockie’s assessment for mobile broadband is that, Indonesia's market still has huge perceive to the data as access. The market mainly competes based on those values (mainly speed, coverage, and price) and the price sensitivity is high and competitive. Regarding survey mentioned in [1] about Indonesian user willingness to pay more for better speed and service, Telkomsel also recently did small survey to their customer, and assess the differences on the preference on the segments below: - Teenager / high school student segment, prefer network (e.g. 3G coverage) then price - Colleague student segment, prefers price then network (e.g. 3G coverage) - Professional segment prefers speed and willing to pay more. He argue the research reported there was decreasing trend in their blended ARPU [2]. However, He does not able to reveal detail information of how much exactly their blended ARPUs, nor the specific data ARPU.

XXIII!

Table 28. Interview with Yose Tireza Arizal44 Question Answer

Do you think media services delivered on top of mobile broadband would have significant value for you as MNO, either to gain market share, increase ARPU or reduce churn rate, etc?

AXIS's strategies are mainly focus to increase ARPU and then market share, and does not care so much about decreasing churn rate (AXIS’s blended avg. churn rate was about 15%.)

Media services that have been launched in order to monetizing mobile broadband pipe?

Axis does not offer streaming service yet, both for music and video. AXIS built own brand for the media content delivery and centralized it trough online channel in http://net.axisworld.co.id/. This channel is delivered in WEB-based and WAP-based, mainly for RBT activation (music content) [3]. AXIS handles the marketing and branding, as well as product platform development and maintenance for this service. There are local partners (content provider) that handle the licensing deal with record label.

Indonesia is one of the few markets that still see that revenue from digital music market is occupied by personalized-based service (mainly RBT), which is 95 %. - In term of demand, supply, and policy enforcement, why do you think this happen? - Drivers and obstacles in delivering digital music service in Indonesia’s mobile market, using streaming / download-based service models? - Do you see the streaming-based music service would take off in Indonesia compare to download-based music service, in term of demand side analysis point of view?

As same as what has been told by Jockie from TELKOMSEL, He was not sure nor the download and streaming-core service is provable business model yet in the near future. Obstacles: • Piracy issues: Minimal law enforcement becomes significant barrier for any players to

monetize music trough these channels, as it offer the easiest and cheapest way to get access to media contents (music, video, game, etc). Thus, the players, including AXIS mainly bundled the service trough their personalized-core service.

• Limited bandwidth and poor quality: although AXIS has roll out their mobile broadband, the availability of bandwidth is not suitable for their network. Personalized-core service offers the easiest and cheapest delivery channel.

Driver: The booming of young generation, middle-income and large usage of social medias that is currently happening in Indonesia is significant factor to raise the needs for entertainment segment. However, it is not enough to be driver to the download / streaming-based business, even if it is connected trough social media channel, due to the obstacle points above. The law enforcement is the first aspect that could drive this kind of business.

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!44 Information is based on the note made during the interview, not recording

XXIV!

There is consistent pattern in the competitive strategy, that the MNO worldwide take one of the positioning strategies below for the digital music services market business: rolling out their own music service (point-to-point partnership with music/broadcasting industry), or partnership with the music service provider / OTT player (point-to-multipoint partnership). What is your view about each of these strategies in the cost vs. benefit point of view?

As same as worldwide market, AXIS's VAS revenue is small if compare to overall revenue. However, the cost to offer media service (e.g. content licensing) in Indonesia is also small, and the MNOs are currently have significant bargaining position toward the media monetization (especially music over the record label), as result of low law enforcement in piracy. Thus, there is no urgent need to partnership with the OTT player. For revenue sharing from the media service offering (mainly RBT activation): The price for RBT subscription is around 6000 IDR, which is 50% is for AXIS, and then other 50% is divided between record label and content provider.

We identify that the MNOs in Europe offer digital media service not directly to increase data ARPU (especially from music) but to differentiate themselves, that at the end to decrease churn rate, increase blended ARPU and acquire more market share. • Do you see that strategy would apply to Indonesia’s

market? • How does AXIS differentiate itself when entering

music and video business, considering the fact the other MNOs in the market are also delivering the same type business model (e.g. music delivery trough RBT activation and/or download-based service)

•

AXIS differentiates itself trough price offering, not mainly trough the service quality, especially from the media services. He does not think the bandwidth (and quality) offering would be significant value for customer in Indonesia in the current situation. Also, the price war in Indonesia is beneficial for them, especially to play in the low-segment customer (AXIS ARPU is around 8 SEK, compare Indonesia's avg. ARPU 40 SEK). Also, there is currently minimum tariff regulation from government, except for interconnection fee. AXIS is one of the challengers MNOs who previously gained advantage from free SMS offering. There was ‘sender keep all’ for SMS interconnection fee that is just changed to 25 IDR interconnection fees to SMS-receiving MNO).

Do you think the MNO network capabilities e.g. billing, traffic and user management, are be the major value in delivering digital music and video service? And how does or will AXIS position itself to address local / global OTT Music and OTT TV/Video player?

Billing, traffic and user management, etc are already as major values for AXIS in the current situation, especially when dealing with the music industry. Considering the bandwidth availability in Indonesia, AXIS does not see OTT music / video players are significant thread yet in the near future, tough the top 10 traffic in Indonesia is dominated by global websites.

XXV!

Table 29. Interview with Didik Akhmadi Usman45 Question Answer

Law enforcement There is currently high digital media piracy in Indonesia, e.g. the business value of the physical piracy in the music industry itself reached 2x legal market, as result of low law enforcement. As recently government orders ISPs to start anti-porn filtering and in line with the broadband development across the country, Does this enforcement also happen for media piracy, including the illegal peer-to-peer,? Is there any rule that require the Communication Service Provider (CSP) to intercept the communication data on behalf of government request for such that purpose, e.g. like IPRED act in Sweden?

The GOI has handled this piracy issue it in ITE act, but there is no further government regulation as a follow up for this issue. It is still under process now.

Licensing vs. Internet service provision Is the online digital multimedia contents provisioned subjected to licensing requirements for broadcasting? If yes, how does it work, is that in the broadcasting or telecommunication regulation?

Didik does not really have a complete knowledge about this. However, as example, there was a problem previously about the spam in premium SMS activation. Now it is still waiting for new ministry regulation, which is still under process to make it more detail. The regulation for media provisioning in internet might be included in this regulation as well, but he does not have information yet. The supervision of the online services rolled out by mobile content provider might be under BRTI, instead of KPI.

Mobile Broadband Pricing Indonesia is also among the countries where the broadband price is the cheapest in the world. How does BRTI currently regulate the pricing for the data access mobile broadband offered by the MNO?

Didik believes that the price would be good if the competition in the market is healthy. However, in his experience in KPPU, he have penalized and markdown of price, because there have been excessive pricing in voice and messaging (cartel), as there are two operator in Indonesia are owned by two companies with a same holding company. Therefore, KPPU decided to markdown the price, in order to protect customer, which was calculated to have caused loss during 4 years. As sectored, there is no regulation for pricing yet, including for

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!45 Information is based on the note made during the interview and transcript from recording (translated from Bahasa to English)

XXVI!

broadband. However, he agreed that there should be ex-ante and ex-post regulation, in both BRTI and KPPU, for this.

Content and Advertising What kind of regulations in the media contents and advertisings that is provisioned in online multimedia services, especially mobile communication?

Didik said that it would be regulated in the new ministry regulation, which is still in the drafting process

Ownership restriction There is a policy that the establishment of private broadcaster in Indonesia is limited to domestic, 100 % initial domestic capital, and for development purpose there is max 20 % of the total capital. How that will affect the international-based OTT multimedia player (TV/Video on Demand, Radio/Music on Demand streaming) in order to roll out their service in Indonesia?

Didik believes the provisioning of cloud services should be regulated, concerning the taxation issue as well. However, he believed, as a regulator (in Indonesia), they should create regulation that includes licensing as well as market regulation. What they are doing would affect the number of actors in the market. Now he said that they have to calculate who the actors are, but he was not sure whether the new players (OTT) have been considered or not, in this draft regulation. Nevertheless, in business perspective, he have been aware that there are many players come from many directions in the market already. He thinks that they have to regulate this as a competitive market based. In the telecom industry, there are many players, and they cannot just use articles in the competition act that is related with (abusing of) dominance position, but the analysis would be related to the conduct of fair business provisioning. Therefore, it is rather to market behavior analysis rather than market structure.

Table 30. Interview with Hengky Philip Ginting46

Question Answer Who are the actors and their relation in provisioning mobile digital music business in Indonesia?

a. MNOs in Indonesia mainly deal directly with the Record labels regarding usage of publishing royalty for monetizing music content (download/streaming, personalized service: e.g. RBT/Ringtone activation, etc). The roles of content aggregator/collecting society/ publisher in this business model are minimum or bypassed. It means they use "multiple-rights"/blanket agreement and direct partnership. b. Mobile Content providers (CP) business areas in Indonesia are mainly in monetizing merchandizes (non-music/video

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!46 Information is based on the note made during the interview, not recording

XXVII!

content, e.g. download wallpaper, premium SMS content subscription/quiz). There are few exceptions:

• If there is any overseas CP partner that would like to distributes their content in Indonesia (e.g. bollywood song), but have no connection with the local MNO. In this case, the local CP bridge the deal, and take 10% from total revenue.

• The local CP acts as content aggregator/service provider (rolling out their own download / streaming music service ), as what we commonly see in other markets, e.g. Jatis mobile with "kantong musik", etc.

• The CP is also the Record label.

What is revenue sharing model between them?

a. Merchandizing & non-music content Content Provider (CP) does not pay any licensing fee to the Record label for monetizing merchandizes, as the license for that content is hold by the CP. This CP works with MNO for delivery channel of these contents, e.g. via SMS, USSD/UMB (e.g. *xxx#), WAP, STK (SIM card), etc. The sharing are one of this 3 possibilities:

• Gold (60% for CP – 40% for MNO) if the content could generate total revenue more than 100 million IDR/month. • Silver (50% for CP – 50% for MNO) if the content could generate total revenue 50-100m million IDR/month • Basic (40% for CP – 60% for MNO) if the content could generate total revenue less than 50 million IDR/month

This performance is evaluated per 6 month, where the deal might chance following the evaluation result. b. Music contents The MNO directly set up deal with the Record label with Revenue sharing model 50-50 (after take out the cost), e.g. subscription fee 9.000 IDR/month, the MNO would take ~2000IDR for cost, then the rest 7000 IDR is split between MNO and Record Label

What do you think about digital music download and streaming service

Hengky thinks the streaming and download business model might take some more time to take off in Indonesia. The personalized services (RBT, Ringtone) is currently the cheapest and easiest way of delivery channel, due to the 3 reasons: - the are huge prepaid customer base, - customer habit toward online media services is not ready yet, - The network capability (bandwidth availability and coverage that effect the service provisioning).

XXVIII!

Appendix B – Mobile Network Operator in Indonesia

Table 31. MNOs in Indonesia

MNO License Type/Coverage

Subscriber (Million)

Principal Shareholder Brand Name Technology Spectrum band) [28]

(MHz)

Telkomsel

Cellular/National (Operating since

1995) 104.2 Telkom, SingTel,

public

kartuHALO, simPATI, Kartu As

GSM-based (GSM, GPRS, EDGE, UMTS, HSPA)

• GSM-900, 7.5 MHz • DCS-1800, 22.5 MHz (un-contiguous 3 segments47: 1x7.5 MHz, 1x5 MHz, 1x10 MHz) • IMT-2000, 10 MHz (contiguous segments: 2x5 MHz)

Cellular/National (Operating since

1995)

Matrix, Mentari, IM3

GSM-based (GSM, GPRS, EDGE, UMTS, HSPA)

• GSM-900, 10 MHz • DCS-1800, 20 MHz (un-contiguous segments: 1x5 MHz, 1x10 MHz) • IMT-2000, 10 MHz (contiguous segments, 2x5 MHz)

Indosat

FWA/National (Operating since

2004)

51.5

GOI, QTel, public

StarOne CDMA based (CDMA 2000-1x, EV-DO Rev 0) CDMA-850, 2.26 MHz

XL Axiata

Cellular/National (Operating since

1996)

43.4 Axiata, Etisalat, public XL

• GSM-900, 7.5 MHz • DCS-1800, 7.5 MHz • IMT-2000, 5 MHz

Hutchison CPT

Cellular/National (Operating since 19.5 Hutchison, CP

Group Indonesia 3

GSM-based (GSM, GPRS, EDGE, UMTS, HSPA)

• DCS-1800, 10 MHz • IMT-2000, 5 MHz

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!47 Non-contiguous means the segment is not contiguous and split for several blocks and/or with different size.

XXIX!

2004) Axis Telekom

Indonesia

Cellular/National (Operating since

2004) 16

Saudi Telekom, Maxis

communication Axis GSM-based (GSM, GPRS,

EDGE, UMTS) • DCS-1800, 15 MHz • IMT-2000, 5 MHz

Telkom

FWA/Limited (Operating since

2002) 16.8 GOI, public Telkom Flexi CDMA-850, 4.92 MHz

Bakrie Telecom

FWA/Limited (Operating since

2003) 14.4 Bakrie group Esia CDMA-850, 4.92 MHz

Smartfren Telecom

Cellular/National (Operating since

2007) 10.6 Smart, Mobile-8 Smartfren

CDMA based (CDMA 2000-1x, EV-DO Rev B)

• CDMA-850, 4.92 MHz • CDMA-1900, 6.15 MHz

Sampoerna Telecomunikasi

Indonesia

Cellular/National (Operating since

2004) 0.5 Sampoerrna Ceria CDMA based (CDMA

2000-1x, EV-DO Rev A) CDMA 450, 3.69 MHz

XXX!

Appendix C - Quantitative Survey C.1. Summary of the Hypotheses and Questions

Table 32. Hypothesis and variable

Hypothesis Variable Explanation of variable - Q1 Question 1: Identification of the respondents’ age - Q2 Question 2: Identification of the respondents’ gender

Q3 Question 3: Identification of mobile device ownership

Q4 Question 4: Identification of mobile device usage to connect to the internet

H1 - Indonesian users have already used the mobile internet

Q5 Question 5: Identification if mobile device is respondents’ primary access to the Internet

Q6 Question 6: Identification of Internet services that respondent has ever used through mobile device

Q7 Question 7: Identification of how often does the respondent use uses the services

H2 - Indonesian mobile users are not exposed to the mobile media services yet

Q8 Question 8: Identification of how long does respondent uses the services for each usage

Q10 Question 10: Respondent’s believe that s/he will use or keep using the services

Q13 Question 13: Respondent’s believe that s/he will have or keep having the services application in their mobile device

H3 - Indonesian users have a positive intention to use the mobile media services

Q17 Question 17: Respondent’s believe that if s/he has opportunity to use the services for free, s/he will use or keep using it

H4 - Indonesian users do have a negative intention to pay for the mobile media services

Q9 Question 9: Respondent’s believe that s/he would like to pay in order to use or keep using the services

Q11 Question 11: Respondent’s believe that the services will save his/her money rather than buying CD/DVD/other physical format of music and video

Q15 Question 15: Respondent’s believe that the services are useful and fun to use rather than buying physical products

Q18 Question 18: Respondent’s believe that the services are complicated to use

H5 - Indonesian users have a positive attitude toward the mobile media services rather than the physical products

Q20 Question 20: Respondent’s believe that using the services will save his/her time rather than using CD/DVD/other physical of music and video

Q12 Question 12: Respondent’s believe that his/her MNO’s network capability does not support to accesses the services

H6 - Indonesian users have a perception that the mobile network and billing are a major obstacle for them to use the mobile media services

Q19 Question 19: Respondent’s believe that s/he will pay a lot for mobile billing account to access the services

Q21 Question 21: Respondent’s believe that his/her mobile phone supports the use of the services

H7 - Indonesian users have a perception that the mobile devices are a major obstacle to use the mobile media services

Q23 Question 23: Respondent’s believe that s/he can afford to buy a mobile phone that supports these services

H8 - Indonesian users get Q16 Question 16: Respondent’s believe that there are or

XXXI!

Q16 will be people in Indonesia using the services significant influences from their social environment to use the mobile media services

Q22 Question 22: Respondent’s believe that s/he would get or will get a positive attention from his/her friends because of using the services

C.2. Summary of Results

Figure 21. Question 1 result

(Identification of the respondents’ age)

Figure 22. Question 2 result

(Identification of the respondents’ gender)

Figure 23. Question 3 result

(Identification of mobile device ownership)

Figure 24. Question 4 result

(Identification of mobile device usage to connect to the internet)

Figure 25. Question 5 result

(Identification if mobile device is respondents’ primary access to the Internet)

Figure 26. Question 6 result

(Identification of Internet services that respondent has ever used through mobile device)

XXXII!

Figure 27. Question 7 result

(Identification of how often does the respondent use uses the services)

Figure 28. Question 8 result

(Identification of how long does respondent uses the services for each usage)

Figure 29. Question 9 result

(Respondent’s believe that s/he would like to pay in order to use or keep using the services)

Figure 30. Question 10 result

(Respondent’s believe that s/he will use or keep using the services)

Figure 31. Question 11 result

(Respondent’s believe that the services will save his/her money rather than buying CD/DVD/other

physical format of music and video)

Figure 32. Question 12 result

(Respondent’s believe that his/her MNO’s network capability does not support to accesses the services)

XXXIII!

Figure 33. Question 13 result

(Respondent’s believe that s/he will have or keep having the services application in their mobile

device)

Figure 34. Question 15 result

(Respondent’s believe that the services are useful and fun to use rather than buying physical products)

Figure 35. Question 16 result

(Respondent’s believe that there are or will be people in Indonesia using the services)

Figure 36. Question 17 result

(Respondent’s believe that if s/he has opportunity to use the services for free, s/he will use or keep using

it)

Figure 37. Question 18 result

(Respondent’s believe that the services are complicated to use)

Figure 38. Question 19 result

(Respondent’s believe that s/he will pay a lot for mobile billing account to access the services)

XXXIV!

Figure 39. Question 20 result

(Respondent’s believe that using the services will save his/her time rather than using CD/DVD/other

physical of music and video)

Figure 40. Question 21 result

(Respondent’s believe that his/her mobile phone supports the use of the services)

Figure 41. Question 22 result

(Respondent’s believe that s/he would get or will get a positive attention from his/her friends because of

using the services)

Figure 42. Question 23 result

(Respondent’s believe that s/he can afford to buy a mobile phone that supports these services