1

Taxi express delivery service in Lisbon – a new business model

Tiago Leite

Keywords:

Express delivery services Taxi service Urban logistics Business model

Abstract

Introduced to Europe in the mid-1980s, at a time that the requirement of European companies for

time-definite, guaranteed delivery could not be met by either postal services or freight forwarders

(Oxford Economics, 2011), express delivery services are still considered one of the «most

expensive, less efficient and more pollutant legs of the logistic chain» (Ehmke and Mattfeld, 2012:

622). The importance of this leg of the logistic chain should not be understated, mainly when we

attend to the companies´ sales revenues dependent on express delivery services which, in some

cases such as in the United Kingdom, can reach circa 28% (Oxford Economics, 2011).

It is in the context of needing to find an alternative solution capable of, at least, facing some of the

existing limitations in the express delivery services in Lisbon that this study was prepared,

developing a business model based in taxi services.

Once the state of art is reviewed, in specific on taxis, methodology to evaluate different urban

solutions and evolution of the business model concept, the current service of taxis and express

delivery services in Lisbon is studied. Based on the methodology presented by Osterwalder and

Pigneurs and comparing the results achieved against the current offer in the city of Lisbon, it is

concluded that, with regards to express delivery services in the city of Lisbon, there is an economic

and capable alternative.

1. Introduction

The last mile is, according to experts, the «most expensive, less efficient and more poluent legs of the logistic chain» (Ehmke and Mattfeld, 2012: 622).The importance of this leg of the logistic chain should not be understated, mainly when we attend to the companies´ sales revenues dependent on express delivery services, which, in some cases such as in the United Kingdom, can reach circa 28% (Oxford Economics, 2011).

«The movement of goods between a transport hub to a delivery address. Commonly known as the last mile, is becoming increasingly competitive and innovative. Strong relationships between retailers and logistic providers have never been more important in ensuring a positive experience for end consumers» for whom «cost, speed, flexibility, reputation and service», in this order, represent the most important factors when choosing a delivery option. (Conlumino, 2014: 3 to 10)

On the other hand, this leg of the logistic chain creates and faces the «traffic congestion of roads in city centers and historic areas with narrow streets». It is also needs to «delivery more frequently smaller parcels, reducing the efficiency of the operations and reducing the profit margin of the service providers» (TMPL, 2011 Aulas Teóricas, IST).

Thus, it seems necessary to find ways and services capable of reducing these limitations and, consequently, improve the existing offer.

In this context, this study considers an express delivery service ,based on taxis and its free luggage compartment capacity, that uses its vacant time to deliver small parcels in the city of Lisbon. In particular, this study will evaluate its business model and the advantages compared to the existing offer in the market, not only in a cost perspective but also looking at quality of service. The next scheme presents the proposed service.

Image 1 – Proposed service; Source: Author

Once the state of art is reviewed, in specific on taxis, urban logistics and evolution of the business model concept, the current service of taxis and express delivery services in Lisbon is studied. Based on the methodology presented by Osterwalder and Pigneurs, I will analyze the business model of the purposed service, compare it against the current offer in the city of Lisbon and conclude on the viability of this new service.

Image2 – Study methodology

2. State-of-the-art in research on taxis

The taxi market is divided into three market segments: 1) «rank/stand», 2) «street hail/cruising» and 3) «pre-book» (Salanova et al., 2011; OECD, European Conference of Ministers of Transport, 2010).

2

«Ranks are designated places where taxi can wait for passengers and vice versa. Taxis and customers are forming queues regulated by a FIFO system» (Salanova et al. 2011: 152).

«In the hail market clients hail a cruising taxi on the street. There is uncertainty about the waiting time and the quality/fare of the service customers will find» (Salanova et al. 2011: 152).

«In the pre-booked market, consumers telephone a dispatching center or a taxi drivers asking for an immediate or scheduling a later taxi service». (Biggar, 2001). In this group we can also include «contract taxis» that generally offer more regular scheduled services. These markets overlap to some extent with non-taxi industries, but this will vary with the definition being used (Aarhaug and Skollerud, 2013: 9).

The way this segment operates is not the same everywhere, being dependent on legal and regulatory aspects in each country and region (Aarhaug and Skollerud, 2013).

Therefore, taxi markets are dependent of local conditions. For instance, «rank/stand» and «hail» taxis are more popular in big cities while in suburban regions, the «pre-booked» taxis are dominant (Salanova et al. 2011: 152).

2.1. Market regulation

There is no general agreement with regards to the way the taxi market should be regulated, particularly, quantitatively (Salanova, 2011; Loo et al., 2007; Fernandez et al., 2006; Aarhaug and Skollerud, 2013)

Salanova (2011), based on the study from other authors, concludes that the situation of the market at a specific moment can be key to the result of market regulations, not only when defining the number of licenses, but also the price of the service.

In this light, it is also interesting to analyze the paradoxical behavior of the industry when operating in a non-controlled market. Instead of the expected reduction in prices, studies done in the USA (Dimpsey, 1996; Nelson and Nygaard, 2008), in Scandinavia (Longva, 2010) or in the Netherlands (Longva and Bekken, 2003) show exactly the opposite (Salanova, 2011: 160).

2.2.1. Legal and economic environment – revision of

models

Since the beginning of the 70s, the number of studies on taxis as been increasing. The firsts, were focused on the economic and financial performance of the market and its need to be regulated. For these, aggregated models were used.

With the start of the 90´s, these studies started including more realistic models on the sector: from the more simple ones, focused on small taxi fleets (Wong, 1997) to some more sophisticated, capable of modeling congestion, market elasticity, different client segments or non-linear costs (Wong, 2008 and Yiang 2010a). For the purpose, different market considerations were applied. (Salanova, 2011: 155).

2.2.1. Aggregated models

In fact, many studies on taxis used aggregated models, such as Orr (1969), Douglas (1972 e 1975), Beesley (1973), De Vany (1975), Manski and Wright (1976), Daganzo (1978), Foerster and Gilbert, (1979), Beesley and Glaister (1983), Schroeter (1983), Frankena and Pautler (1986), Hackner and Nyberg (1995), Cairns and Liston-Heyes (1996), Arnott (1996), Yang and Wong (1997), Chang and Huang (2003), Daniel (2003),

Ferdandez et al. (2006), Schaller (2007), Chang and Chu (2009) or Massow and Canbolat (2010), among others.

The problem with market regulation, and the main reason for the initial studies, was brought by Friedman (1962: 282), when he studied why taxi drivers are against the increase in number of licenses. Taxi drivers believe that by increasing the competition, revenues (already considered too low) will decrease. Friedman concludes questioning: Is this economic evaluation correct?

In another important work, Douglas (1972: 116-127) identifies the challenges in the problem formulation, considering the «dilemma among regulators in selecting a price and defining the level of services», mainly due to the large group of clients exerting pressure. Douglas concludes stating: «the price generated by the “competitive equilibrium” may be clearly inefficient, being higher than the set of prices that could be defined as efficient, given the constraints if this market». Similar conclusions were reached by De Vany (1975) and Foerster and Gilbert (1979).

One of the main objectives of these studies was solving, in a clear way, two important issues: a) restrictions in the taxi market and b) control of service prices.

The models developed to analyse these issues (price and entry controls in the taxi market) were based in aggregate demand and supply models and tested in different markets (monopolistic and competitive). They were based in three main assumptions: a) the relation between the waiting time and the total number of vacant taxi hours; b) constant operating cost per hour; c) demand estimation based in fares and waiting time of passengers (Salanova et al. 2011: 154).

Assuming that regulations to market entry are redundant when price regulations are in place, Fernandez considers that the necessity for regulation should be carefully considered, depending on the situation (Salanova, 2001: 160 and Fernandez, 2006: 254).

In order to provide information to regulatory authorities, other studies sought to understand how supply and demand were balanced in the presence of regulation (Beesley and Glaister, 1983: 595). However, taxi services have characteristics that distinguish them from traditional competitive markets. It is through the behavior of different variables of the taxi service (client´s waiting time) and the occupancy of the vehicle (fraction of time while taxi is in service) that the demand for a service is interconnected to supply. In general, the amount of service offered (empty and busy taxi per hour) is usually superior than the amount actually used (busy taxi per hour), which means, there is always a capacity left in the system (empty taxi per hour) (Wong, 2008: 998).

This indicator also represents the client’s average waiting time which is an important factor to assess the quality of service. Indeed, this factor influences the decision of choosing a taxi instead of other mode of transport and therefore, it is crucial for achieving the price and the equilibrium of the market (De Vany, 1975; Abe and Brush, 1976; Foerster and Gorman, 1979).

In most of the markets, where quality is a variable, each company can decide on the quality of the product. In the taxi market, the average waiting time is not a variable nor can easily be differentiated among competitors since is dependent on the number of taxi per hour available. Although larger companies may be able to impact the waiting time, most of companies will not be able to offer to customers a lower waiting time than the one offered by most companies in the market (Frankena and Pautler, 1986: 4).

3

2.2.2. Equilibrium models

Although insightful and interesting, the studies above could not simulate with precision the taxi behavior.

Taxis are a comfortable mode of transportation, fast and reliable, but they impact the traffic in a city, especially when moving without clients (Wong et al., 2008: 985).

Since the demand for this service is limited to a certain area, this service should be evaluated through an equilibrium model that uses traffic assignment to analyse origin and destination of passengers. This is an important step, as the models started taking into account that taxis operate in urban networks, sharing the streets with daily traffic and other public transport modes (Salanova et al. 2011: 159).

Based on this, Yang and Wong (1998) presented a network model describing how vacant and occupied taxis move in an urban network searching for customers and providing transportation services to them. The model relates the taxi fleet size, the uncertainty of drivers in finding clients and service quality (waiting time). They concluded that the average utilization of a taxi decreases with the number of taxis available and the higher the taxi utilization rate, the higher will be the costumer waiting time. Thus, the size of the taxi fleet and taxi utilization data must be regulated in order to achieve a better utilization while maintaining a certain level of service (K.I. Wong, 2008: 988). Similar conclusions were also reached by Yang and Wong (1998) through aggregated models.

Manski and Wright (1976), Schroetter (1983) and Arnott (1996) concluded that the increase in the number of taxis would be benefic both for costumers and taxi companies. However, this intriguing result may result in an «unstable situation» and is «highly unlikely to happen» in «a real taxi market» (K.I. Wong, 2001: 821).

All the models commented above use a linear taxi fare structure, making long-distance (from/to the airport) trips more profitable and creating over-supply in airports, wasting many taxi service hours in the airport queue (Salanova, 2011: 152). This problem was studied by different researchers as Loo (2007), who included non-linear taxi fares in its Hong Kong model.

The models have become more sophisticated. For example, Hyunmyung´s model (2005) simulates drivers learning processes, implementing the «day-to-day evolution approach» introduced by Horowitz (1984), Vythoulkas (1990) and Cascetta and Cantarella (1991).

It is also important to refer «the effectiveness of taxi information systems in reducing unnecessary travels. In fact, using information systems is equivalent to a 20% increase in the number of taxis when looking at service quality (as pointed out by Yang and Wong) (Salanova et al., 2011, 154). This result is also aligned with the results from the IMTT 2006 study about the conditions of taxi services in Lisbon.

A combination of these results and new Intelligent Transport Systems (ITS) allowed the development of other studies. As an example, Martinez et al. (2014) proposed a new concept of urban shared-taxi services through an agent-based simulation model. He concluded that the proposed system may lead to significant travel time saving and a fare reduction to passengers close to 20%. Taxis will also benefit in an increase in operation efficiency, measured by revenues/km.

3. Urban freight flows

Around 80% of European citizens live in urban areas. According to the European Union (2007) and the United Nations (2006) this number is expected to increase, which will generate higher demand for express delivery services in urban areas (United Nations, 2006).

Despite its importance, urban fright transport is far from being «efficient as it could» (Lindholm, 2012: 134).To overcome these limitations, and because urban areas are considered «particular obstacle to freight transport, both in logistic performance and in its environmental impact», legislators have been preparing a «wide range of regulatory, technologic and logistical measures». (Balm, 2014: 394).

3.1. Urban logistics

«Urban logistics is the process of distributing goods (including not only retail, but all economic sectors within cities and urban regions, including reserve flows of waste or not used goods. The process of distribution entails three types of flows: physical, logical and contractual. Finally, urban logistics agents are all those (private and public) agents that have any sort of interest, either commercial, social or other». (Macário and Reis, 2009: 150). The constraints and problems faced by this activity worldwide are generally known (Anand et al., 2012a and 2012b; Browne et al., 2012, Balm et al., 2014).

The growth in urban population and the consequent increase in the activity will have an impact on gas emissions, traffic congestion in and around cities, noise pollution and number of accidents. (Browne et al., 2012: 23).

In this context, it is essential to create foundations for the sustainability of urban freight (Baindur and Macário, 2012: 110; Macário and Marques, 2008: 146).

Among others, Anand et al. (2012: 113), associates responsibilities of the impacts of the urban logistic activity to the different stakeholders and to the variety and unequal interest that end up making them taking the wrong decision, increasing the problem. Indeed, there are many stakeholders involved in the process and each one has its own interests (Macário, 2008: 91).

To achieve the sustainability of the activity it is really necessary to improve the regulations. However, measures implemented by local authorities in order to mitigate the negative impact of the activity have proved to be, sometimes, inefficient (Baindur and Macário, 2012: 110). Thus, «there is a need to better understand the interactions and the decision-making process among different stakeholders in a more systematic and analytic way», in order to better understand urban freight movements (Anand, 2012:114; Cherrett, 2012: 31).

Nonetheless, and as referred by Ogden (1977), the diversity of stakeholders involved raises some questions. The approaches to the problems will need to be different and will need to take into account each one of the stakeholders.

Odgen (1978) defends two types of approaches. The first, a «commodity-based» approach, oriented towards an economic perspective, consists on classifying the models in relation to the reference unit of quantity or delivery model. Researchers such as Hutchinson (1974), Ogden (1978) and Russo and Comi (2004) also worked in developing models such as this.

The second proposed approach, vehicle-based, targets traffic planning and authority regulation. It aims to understand the freight delivery movements, both at the origin and destination

4

points. Research on these models can be found in Routhier and Aubert (1998) or Holguin-Veras and Patil (2008) works.

More recently, other developments have been made in this area of research. Muñuzuri (2010), based on an agent-based approach, sought to «model the complex freight network», among others.

As previously mentioned, it is important, when making a decision, to not forget to consider all the different stakeholders involved in the logistic (Macário, Galelo and Martins, 2008: 94).

As Ibeas (2012: 1068) addresses «the issue of data collection, models and methods for urban freight transport demand investigation and the difficulties related to costs and the fact that few data are often available». The author considers «important to investigate the transferability of results in order to improve their use and to assess whether the obtained results is dependent on any particular condition, and whether the lessons learnt in one city can be transferred to other cities». Moreover, Leonardi (2014: 97) defends that despite the negative impacts related to the activity, an efficient operation of this logistic leg «generates economic profitability to the business» and simultaneous advantages. Thus, in order to assess the advantages of each one of the different solutions, there have been efforts in defining Key Performance Indicators (KPI). In this way, each one of the agents involved in the logistic chain can be evaluated in the success of implementing of an urban logistic system.

4. Business model

Osterwalder and Pigneur (2010: 14) defend that «the starting point for any good discussion, meeting or workshop on business model innovation should be a shared understanding of what a business model actually is». They also defend that it is necessary to define a «business model concept that everyone understands», one that is «simple, relevant and intuitively understandable».

The study developed by Osterwalder in this topic starts in 2004 with the preparation of his PhD thesis, the Business Model Ontology. During this work, the author tries to specify and conceptualize business models (Osterwalder, 2004: 1). In 2005, Osterwalder, Tucci and Pigneurs, try to reference all the concepts associated to business model, especially the ones related to information systems. The analyse of these results together with the ones obtained during the PhD, allowed Osterwalder to conclude that the business model consists in the «translation of strategic issues, such as strategic positioning and strategic goals into a conceptual model that explicitly states the business functions» (Osterwalder et al., 2005: 1).

This topic was further developed with the publication of the book Business Model Generation (2010), where Osterwalder and Pigneur develop a different methodology. In this book, the authors use schemes and tables to explain examples. Although this technique is not very common in scientific work, it is very practical and clear, which makes it more accessible to the readers.

Osterwalder and Pigneur (2010) develop a method to describe the business model through «nine basic building blocks that show the logic of how a company intends yo make money. The nine blocks cover the four main areas of a business: customers, offer, infrastructure and financial viability».

1 – Customer segments – all the customers for whom the company creates value («For whom are we creating value?», «Who are our most important customers?»).

2 – Value propositions – seeks to solve customers’ problems and satisfy their needs («What value do we deliver to the customer?», «Which one of our customer’s problems are we helping to solve?», «Which customer needs are we satisfying?», «What bundles of products and services are we offering to each customer segment?»)

3 – Channels – value propositions are delivered to customers through communication, distribution and sales channels («Through which channels do our customer segments want to be reached?», «How are we reaching them now?», «How are our channels integrated?», «Which ones work best?», «Which ones are most cost efficient?», «How are we integrating them with customer routines?»)

4 – Customer relationships – the types of relationships a company establishes with specific customer segments («What type of relationship does each customer segments expect us to establish and maintain with them?», «Which ones have we established? «How costly are they?», «How are they integrated with the rest of the business model?»)

5 – Revenue streams – the cash generated from each customer segment («For what value are our customers really willing to pay?», «For what do they currently pay?», «How are they currently paying?», «How would they prefer to pay?», «How much each revenue stream contribute to overall revenues?»)

6 – Key resources – resources required to produce and deliver a value propositions to the customers («What key resources do our value propositions require?», «Our distribution channels?», «Customer relationships? «Revenue streams?»)

7 – Key activities – the most important activities a company must do to make its business model work («What key activities do our value propositions require?», «Our distribution channels?», «Customer relationships?» «Revenue streams?»)

8 – Key partnerships – the network of suppliers and partners that make the business model work («Who are our key partners?», «Who are our key suppliers?», «Which resources are we acquiring from partners?», «Which key activities do partners perform?»)

9 – Cost structure – the business model elements result in cost structure («What are the most important costs?», «Which key resources are most expensive?», «Which key activities are most expensive?»)

The nine blocks are grouped into a tool called «Business Model Canvas».

Image 3 – Business Model Canvas; Source: Business Model

Generation, 2011

5

5. Case study: Lisbon

In this chapter, the points previously studied, in particular, taxi operations, and business models, will be evaluated in the context of Lisbon.

5.1. Taxi services in Lisbon

Although the few operational studies available in Lisbon, it is worth mentioning the Mach 2006 IMTT study where «3,460 taxi vehicles were identified in the city of Lisbon». The study compared the operational results between taxis operating with access to a dispatching center and taxis operating without access. In 2006, 52% of taxis in service in Lisbon belonged to the first group while 48% made part of the second (the study considered a daily average of 3,100 taxis in operations).

The study concluded that taxis working with a dispatching center achieved, among others: A number of services per day «30% above competitors

without dispatching center», that represents 4 more services per day;

Additional 4 hours per day of an average work; Less «empty» kilometers per service completed (5 «empty

km» for those with access to a dispatching center and 7 «empty km» for the remaining);

An average revenue per day 18% higher than competitors (86€ with access to dispatching center and 73€ for the remaining);

Results per day 3 times higher.

Because there is no public information about the number of taxi services and their distribution in Lisbon, the booking data of May and August 2011 and January 2012, from the taxi dispatching center Retalis was considered and analysed.

The available booking data indicates that this dispatching center has booked on average around 2,200 services per day. The data also indicates that on average there are more services during working days (2,340) than weekends (1,870). It is also interesting to note that Fridays are the day of the week with more services booked.

Image 4 – Average number of services per day; Source: Author

Image 4 – Average number of services per type of day; Source: Author

The distribution of the demand during the day is identical over the working days and the weekend.

As it can be seen in the previous chart, the peak for booking orders on working days happens at 9h45 in the morning, with the peak hour happens between 9h00 and 10h00, at the start of a working day. The highest number of new orders happens, usually, during the morning (between 07h30 and 10h30) and is, on average, 178 per 15 minutes interval. During the day, the number of orders decrease, reaching the lowest value during the night, between 00h00 and 5h00, with 37 order per 15 minutes intervals.

Although the daily profile during the weekend is more homogenous, there is a clear increase in the number of services during the night while compared to the night of a working day.

Image 5 – Average number of services in 15 minutes interval; Source: Author

Regarding the average time of service during a day, the similarity between working days and weekends is closer. The highest average time of service is observed both in the morning and in the evening, during the hours that correspond to the traffic peak in the city. The average length of service is superior to 19 minutes (19:01 minutes) between 7h00 and 10h30 and close to 20 minutes (19:28 minutes) between 14h30 and 18h00. It should be noted that during weekends, at these same hours, the length of a service is lower, perhaps related to the presence of a smaller number of vehicles on the road (see chart below). During the remaining hours of the day, the service average length is close to 17 minutes.

6

Image 6 – Average length of service (minutes) during the day; Source: Author

Through the data provided by Retalis, it was also possible to identify the locations from where the services start. During weekdays, the three locations with highest number of calls are Rua Castilho, Rua Rodrigo da Fonseca (Avenida da Liberdade) and Avenida Jose Malhoa, representing 1,58%, 1,52% and 1,35% of the total number of booking orders, respectively. During the weekend, Rua Rodrigo da Fonseca is the location with the highest number bookings (1,97%) followed by Rua Castilho (1,85%) and Rua Jose Malhoa (1,53%). The image below shows the top 20 locations with the highest number of booking orders in Lisbon during working days.

There is an expected concentration of booking orders in the Lisbon city center. The top 20 locations represent close to 17,2% total orders during working days and 21,55% during weekends.

Image 7 – Top 20 locations in number of booking orders during weekdays; Source: Author

5.2. Urban freight transport in Lisbon

Although considered «a major opportunity for promotion of the economic activity», Lisbon’s logistic system is considered undeveloped by Viegas et al. (2005).

Viegas et al. also point the «inadequate availability of logistic support infrastructures, high concentration of business and services, inadequate access regulations for freight and delivery vehicles, insufficient parking capacity for loading and unloading operations and inexistent monitoring of this parking spaces, usually occupied by other vehicles» (CML, 2005).

It should also be noted the existing «insufficient articulation among intermodal infrastructures and the delay in developing a network of logistic platforms» resulting in a logistic scene in Lisbon dominated by «road transport only occasionally replaced by intermodal services» Viegas et al. (2005).

ANACOM conducts biennial surveys on the use of postal services by the population in order to assess the level of use and quality of postal services by «households».

According to the December 2014 publication, the service «continued to experience an overall decrease, statistically significant in the use of postal services». The main services used by the interviewed users were mail (27%), small parcels (13%) and express mail (2%). The service, ranked on a scale of 1 to 10, was voted between 7.5 and 8.2 for «average satisfaction», depending on the considered service. It should also be noted that overall «client satisfaction» has decreased between 2012 and 2014.

«Confidentiality» continued to be the service «that greater satisfaction provides to the clients of the various postal services», followed by «delivery times». Finally, with the lowest satisfaction rate, «prices». With regards to small parcels, «undamaged deliveries» and «confidentiality» were the factors presenting higher satisfaction rate.

In addition, ANACOM results show that «most of the mail and small parcels sent by households in 2014 were collected by other private consumers. Only registered mail presented a significant percentage of business and service collectors».

Lisbon has a diverse offer of express delivery services. These transport services constitute «a differentiated element of the logistic chain and in the relation with the customer, either B2B (business to business) or B2c (business to customer) models.

To study these services, four transport companies have been analised based on the Osterwalder and Pigneur (2010) model previously referred. The companies were selected based on their operations (only companies that operate in Lisbon were selected=. The four companies selected were: CTT Expresso, postal service of Portugal; DHL, an international logistics company; Camisola Amarela, specialized in eco friendly delivery services of small parcels in Lisbon; and Timeplus, other Portuguese express delivery company. The result of this analysis is presented in Table 1.

6. Proposed business model

The analysis of the previous studies and data provided background and support for the proposal of creating a new business model, in particular, an alternative transportation service that uses taxis to provide express delivery services in Lisbon.

This alternative business model will be studied and analysed using Osterwalder and Pigner’s methodology – the business model canvas (2010). The business model canvas has four main blocks: a) customers which includes customer segments, channels and customer relationships; b) infrastructure which includes key resources, key activities and key partners; c) financial viability which includes revenue streams and cost structure d) offer, value propositions.

7

Image 8 – Alternative business model; Source: Author

6.1. Customer segments, channels and customer

relationships

The target customer segments of this new business model will be the companies, organisations and persons that are interested in a express delivery service in Lisbon. In particular, the organisations and persons interested in a seven day, 24 hours service or / and in a low cost service.

The proposed service will have four channels to communicate with customers: website, phone, app and social networks. These channels are fundamental to build good relationships and trust with the customers. The phone and the website will be the main communication channels, through these customers will be able to acquire the service (acquisition channel) and find out more about the service (information channel). It will also be developed an app which will permit customers to acquire the delivery services through a self-service portal. Finally, the social networks will permit a closer contact and will be important to promote the service among new potential customers.

6.2. Key activities, resources and partners

The proposed transport service is a door-to-door express delivery service for packages under 15Kg, 24 hours per day and seven days a week.

Key resources of this service include the vehicles and the drivers. As previously referred, these vehicles are the taxis operating in Lisbon, connected to a dispatching centre and equipped with a GPS system. The dispatching centre and website are essential to the communication with the customers, and so also considered as key resources of this service.

The partners, i.e. the taxi companies and the dispatching centre, will be essential to the implementation of the proposed service, as these will provide the key resources referred previously (vehicles, drivers, equipments and website). Note that the current partners of the taxi companies (local authorities, insurance companies, petroleum companies) have not be considered as partners of this new service.

6.3. Value proposition

The proposed service provides the opportunity to operate express delivery services using taxis in Lisbon, and create value to the customers. Taxis provide a reliable service, and have a strong brand and reputation. Also, this service will satisfy the need of customers of having a door-to-door service available seven days per week and 24 hours per day.

Additionally, the proposed service offers real time tracking (online), quick response to new orders, and an eco-friendly service (as it does not increase the number of vehicles in the city). Finally, the proposed service has a lower price compared to other express delivery services (see chapter 6.6).

6.4. Revenue streams

The revenues will be generated by the delivery services and will depend on the travel distance and on the number of available taxis at the time of the service. The price of the service will depend on the operating costs and opportunity costs (analysed in chapter 6.5). Other revenue sources such as publicity in the website, app or vehicles are not considered in this study as they are already part of the revenue stream of the current taxi service.

6.5. Cost structure

The proposed business’s cost structure will be based in four categories: a) initial investment, b) sunk costs, c) operational costs, and d) opportunity costs.

6.5.1. Initial investment

The initial investment includes the costs related to the app, website and software development as well as the costs related to the promotion of the proposed service.

6.5.2. Sunk costs

One of the biggest opportunities of the proposed business model is the possibility to operate a delivery service where the costs related to staff, acquisition of the vehicle, insurance, pneumatics and administrative services are absorbed almost entirely by the costs of the taxi service. These are called “sunk costs”, costs that occurred in the past, are independent of future events and cannot be recovered. Many economists share the opinion that “sunk costs” should not be considered in a financial decision or, in other words, in a financial decision only future costs should be considered (Krugman et al., 2008).

6.5.3. Operational costs

The third category of costs, the operational costs, compromise the costs related to the exploitation of the service, specifically the fuel, maintenance and repair of the vehicle, and the call supplement (currently 0,80€ per service) paid directly by the customer to the dispatching centre. As referred previously, other operational costs such as staff and insurance are not considered in the cost structure of the proposed business.

According to Federação Portuguesa do Táxi (2014: 12), costs related to the fuel and maintenance and repair of the vehicle represent 27,8% and 16,2%, respectively, of the total operational costs of taxi services. To determine the fuel costs of the proposed service, it was calculated the average fuel consumption, fuel price and average travel distance. Based on the most popular cars in Portugal this year, the average fuel consumption was calculated as 6L/100 Km, and the fuel price as 1,254 €/L (average price on 26 June 2015) (www.precoscombustiveis.dgeg.pt).

The average travel distance is determined based on the average time of travel of the taxi service (assumed to be the same for express delivery services) and average velocity in Lisbon. Using Google Maps, an application that provides historical information and traffic previsions in different days,

8

times and locations, it was studied the velocity of six different routes in Lisbon (on 8 July 2015) during the day.

The average velocity obtained was 21,79 Km/h. To validate this analysis, the result was compared to two studies. The first study indicates14,6Km as the average velocity of Carris vehicles (buses, trams and elevators) in 2014. The second study suggests 25,1 Km/h as the average velocity in the main roads in Lisbon. Taking into account these two studies and the fact that the velocity of taxis would probably be in the middle of these two parameters, it can be concluded that the result obtained is a good estimate of the average velocity of taxis in Lisbon.

The average travel distance was calculated multiplying the average velocity and average travel time during each 15 minutes interval, during the day. The value obtained was 6,38 Km, which is not very distant from the average distance per service referred in Revista Táxi, 5,27 Km (FPT, 2014).

The last step to calculate the fuel cost was multiplying the fuel price, average consumption and average travel distance, achieving a fuel cost of 0,48€ per service. Taking into consideration the relative weight of the fuel costs (27,8%) and the maintenance and repair costs (16,2%), the maintenance and repair costs are 0,28€ per service.

The total operational costs are the sum of the fuel (0,48€), maintenance and repair (0,28€) and call supplement (0,80€). The total cost is 1,56€ per service.

6.5.4. Opportunity costs

The opportunity cost is defined as the value of the best alternative that must be forgoten to pursue a certain action. In other words, the value that would have been obtained if a different action would have been taken (Krugman et al. 2008).

In fact, when the taxi driver accepts to deliver a parcel, he/she rejects the possibility to do a taxi service. Therefore, the opportunity cost of not doing a taxi service must be considered when calculating the economic costs of an express delivery. In this case, the opportunity cost should consider the probability of doing a taxi service and the expectable profit of that service.

Based on the data provided by Retalis distaptching centre (analysed in chapter 5.1), during the period analysed (May 2011, Ausgust 2011 and January 2012), an average of 1108 taxis was identified. However, the number of the taxis available varies during the day. According to the study from IMTT (2006) 60% of taxis work on 23 hours shifts and 40% on 12 hours shifts.

After determining the number of taxis available, the next step is to estimate the number of taxi services and their distribution during the day. It is assumed that the distribution of all taxi services follows the same distribution of the services of the Retalis dispatching centre. According to the data provided by Retalis, each taxi does 9,5 services per day. However, if the number of services ordered on the street is considered, the average number of services is 22 for taxis operating on 23 hours shifts, and 11 for taxis operating on 12 hours shift (IMTT, 2006).

After estimating the number of taxi available and the number of provided services during the day, it is possible to determine the occupancy rate of the taxis throughout the day.

Image 9 – Taxi occupancy rate in Lisbon; Source: Author

Assuming that the average travel time of express delivery services is the same of the taxi services, and that the time of loading and unloading varies between 10 and 15 minutes, the average time of each express delivery service is around 45 minutes. Therefore, the probability of ignoring a taxi service is the probability of having a taxi service in the next 45 minutes. The probability of the taxi service (P.Si) is calculated below:

Equation 1: P. Si =STi+ STi+1+ STi+2

S.Taxisi

STi is the number of services in the 15 minutes interval i and S.Taxis is the number of taxis in the same interval i. The results obtained are presented in the graphic below.

Image 9 – Probability if a taxi passenger service during the next 45 minutes; Source: Author

Moreover, to determine the opportunity the cost, it is also necessary to calculate the profit that the driver could generate from the potential taxi service. This profit is computed as the revenues (price of the service) minus the operational costs (fuel and, maintenance and repair costs). The operational costs of delivering a taxi service are equal to the costs of delivering a parcel, i.e. 1,56€ per service (calculated in chapter 6.5.3). The price of a taxi service is dependent on the time of the day (day or night) and was calculated based on tariff 1. The average price of service at day and night are 5,40€ and 6,66€, respectively (waiting time and supplements were not considered in the calculation).

Lastly, the opportunity cost is obtained by multiplying the probability of taxi service occurring in the next 45 minutes and the profit from the potential taxi service:

Equation 2: OCi = P. Si × (PServicemi

VAT 6%−

FC−MC

VAT 23%)

Where OCi is the opportunity cost in the 15 minutes interval i, P.Si is the probability of a taxi service occurring in the next 45 minutes, Pservicei is the price of a taxi service in i, FC is the

9

average fuel cost per service, and MC is the maintenance and repair costs per service. The opportunity cost is calculated without VAT, i.e. the VAT is deducted from the price of the taxi service (6%) and the fuel and maintenance and repairs (23%).

6.5.5. Economic cost

The economic cost of the express delivery service is estimated by summing the opportunity cost and the operational costs (without VAT).

Image 9 – Economic cost of purposed service; Source: Author

6.6. Comparison between the different business

models

In this chapter, the proposed service is compared against the four express delivery services described in chapter 5.2 – CTT Expresso, DHL, Timeplus and Camisola Amarela.

The price of the proposed service is compared with the price of the current express delivery services in Lisbon. In this analysis, only delivery services of 10-15 Kg packages in Lisbon were considered. Using these parameters, the proposed taxi service is more competitive than the majority of the services of CTT Expresso, DHL, Timeplus and Camisola Amarela, with Camisola Amatrla service «Bicycle – normal (until four hours)» as the only exception.

Osterwalder and Pigneur (2010) business model canvas framework is used to compare the proposed business model with CTT Expresso, DHL, Timeplus and Camisola Amarela business models, and summarized in the table 1.

The comparison between the different services, in particular their value propositions, can also be analysed using the «Strategy Canvas» framework (Kim and Mauborgne, 2004). In this framework, «the horizontal axis on the strategy canvas captures the range of factors that an industry competes on and invests in, while the vertical axis captures the offering level that buyers receive across all of these key competing factors». This framework not only serves to identify the main factors that the industry competes but also to reoriented the focus from competitors to alternatives and from customers to non customers of the industry.

The framework illustrated in image 10 clearly identifies an opportunity in terms of speed, availability and price of the service, represented by the factors «service schedule», «speed of collection», «speed of delivery», and «price».

Image 10 – Strategy Canvas; Source: Author

The advantages of the proposed service include low price, high availability, speed of collection (no need to pre-booking) and delivery, and eco friendly. Indeed, the proposed taxi delivery service targets customers that need an affordable express delivery available at any time of the day.

However, it is important to note that the proposed service has some limitations when compared to the existing services, in particular, its area of activity limited to Lisbon, absence of collection points, limited size and weight of the load, and impossibility of offering personalized solutions.

Existing services New service

Key Partners

Government (CTT) Airlines (CTT) Freight forwarders (TP, CA) Franchise business (TP)

Taxi companies Taxi dispatching centers

Key Activities

24 hours per day, mail and parcel deliveries, up to 30kg, in Portugal (CTT) Global service of express deliveries using company owned equipment and

vehicles (DHL) Distribution, delivery, collection and transportation of documents and parcels

through a network of couriers with their own vehicles (TP) Eco friendly distribution of parcels up to 40 kg (TP)

24 hours per day, 7 days per week, express deliveries up to 15 kg in Lisbon.

Key Resources

Post office/Delivery and Collection (CTT, DHL) Large national distribution network (CTT) Owned fleet (boats, airplanes, etc.) and equipment (DHL) Warehouses (DHL, CTT, TP) Global network (DHL) Vehicles (TP) Messengers (TP) Bicycles (CA) Location of distribution warehouses in Lisbon (CA)

Vehicles – taxis (existing) Communication tools

(existing)

10

Value proposition

Home delivery in the Iberian Peninsula through a company owned fleet(CTT) National distribution network and stores(post offices) (CTT) Reliability (CTT) History/Brand (CTT, DHL) Customized solution (CTT, DHL, TP) Online tracking of vehicles and mail/parcels (CTT, DHL) Delivery and collection points (CTT) 24hours, 7days per week service (CTT) Integrated logistic on an international network (DHL) Door to door collection and delivery following costumer directions (CTT, DHL,

TP, CA) 7 days per week delivery and collection of parcels in DHL collection points

(DHL) Deliveries up to 90 minutes after collection call (TP) Internal mail service for companies (TP) Eco friendly deliveries (CA) Express service up to 4 hours in Lisbon city center (CA)

Express deliveries in Lisbon based on taxi service

Reliability History/Brand Online tracking 24 hours per day, 7 days a

week, door to door service Low cost service Quick response to delivery

requests

Costumer relationship

Post office (CTT, DHL) Website (CTT, DHL, TP, CA) Telephone (CTT, DHL, TP, CA) Publications (CTT, DHL) Events (CA)

Website Telephone E-mail Taxi App

Customer segments

Householders and companies for mail, express mail and parcel distribution in the Iberian Peninsula (CTT)

Householders and companies for express parcel deliveries around the world (DHL)

Householders and companies for mail, express mail and parcel deliveries (TP, CA)

Companies with need for customized service (CTT, DHL,CA,TP)

Householders and companies for mail, express mail and parcel distribution

Channels

Post office/ delivery and collection(CTT, DHL) Website (CTT, DHL, TP, CA) Telephone (CTT, DHL, TP, CA) E-mail(CTT, DHL, TP, CA) Cash (CTT, DHL, TP, CA) Transaction (CTT, DHL, TP, CA)

Website Telephone E-mail App Taxi Cash Transaction

Cost structure

Operational costs – staff costs, contracted services, stocks (CTT) Operational costs (delivery, maintenance, exchange rates, materials,

insurance, etc.) (DHL) Staff costs (DHL, TP,CA) Fleet cost–owned and leasing (CTT, DHL, CA, TP) Warehouse, delivery points and office costs (CTT, DHL,TP,CA)

Operational cost Opportunity cost

Revenue streams

Provision of services (CTT, DHL, TP, CA) Provision of services

Table 1 – Business model comparison; Source: Author Note: CTT:Correio de Portugal; TP: Timeplus; CA: Camisola Amarela

7. Conclusion

With new technical developments entering the market nowadays, there is a need for traditional industries and services to rethink themselves through innovation/ diversification/ adaptation in order to keep the businesses profitable. This challenge is similar to the one the taxi industry is facing, especially with the rise of new alternative services (e.g. Uber) that can jeopardize the existing operating service.

Simultaneously, this study provides a better knowledge and understanding of the Lisbon taxi system, in particular, the average length of a service, the distribution of the demand during the day, main locations in the city and the overall occupancy rate of this passenger transport system.

This work also evaluates and compares the current conditions of express delivery services in Lisbon. Through this assessment it was possible to observe the potential for an alternative service to the existent ones, with the advantage of being cheaper. Moreover, a new service that can also optimize the taxi service in the city.

Also important to refer the existence of numerous studies about taxi-sharing models (Ma, Zheng and Wolfson (2013), d’Orey, Fernandes and Ferreira (2012), among others), which could be combined with the express delivery service here proposed: a traditional service of passenger transportation, integrated with a delivery service, in which the collection or delivery of parcels/documents will be only possible when the vehicle is not carrying passengers although, the movement of passengers and goods may be done simultaneously, if necessary.

Image 11 – Proposed model; Source: Author

11

An eventual implementation of this project will necessarily require further research, such as, O/D location for collection and delivery of good in Lisbon or even the maximum volume of goods/parcels/documents that could be carried. This could give a better understating of the potential turnover generated by this purposed new service.

Also, this study has also some limitations that should be taken into account. In particular, limitation on the lack of data for taxi services in Lisbon. Only 3 months of data were used when it could be considered ideal to use the 12 months of the year. Also, the results obtained were based on pre-booked services provided by telephone calls and did not consider neither «hail» nor «rank» services. This will naturally impact indicators, such as the number of services in 15 minutes intervals. Finally, an assumption was made for the average number of taxis in service during the day in Lisbon.

Despite the limitations, clearly assumed, I think, it can be said that this study is a solid and grounded work and is able to answer two important questions: the express collection/ delivery of goods in the city of Lisbon, presenting a true alternative for the existing express delivery services; and, simultaneously, supporting the «clash» that the taxi service will face in short/ medium term, given the challenges that have been posed to it. Additionally, the proposed service presents an economically viable alternative, to bring real value to the costumer and that will not harm the traffic network in the city as it takes the most from an existing element of the network.