The Neuroscience of Learning: A New Paradigm for - Maritz Institute
PROPRIETARY AND CONFIDENTIAL, MARITZ COPYRIGHT 2009 1 Integrating Survey and Social Media Research...
-
Upload
shamar-douberly -
Category
Documents
-
view
214 -
download
1
Transcript of PROPRIETARY AND CONFIDENTIAL, MARITZ COPYRIGHT 2009 1 Integrating Survey and Social Media Research...
PROPRIETARY AND CONFIDENTIAL, MARITZ COPYRIGHT 2009 1
Integrating Survey and Social Media Research
Understanding the Voice of the Customer (VOC) in mobile devices using
Maritz Product Launch Monitor study.
Michael House
Vice President, Text Analytics
Maritz Research
Sentiment Symposium
April 12, 2011
PROPRIETARY AND CONFIDENTIAL, MARITZ COPYRIGHT 2009 2
Today’s Session
1. Briefly review how organizations struggle when it comes to “putting the voice of the customer (VOC) to work”
2. Discuss a best practice in VOC integration, a Uniform Set of Customer Experience Categories, that enables comparing and contrasting multiple VOC sources
3. Examine preliminary results from the PLM study
A. Sentiment of Sources
B. A Look at Two Brands: Blackberry and HTC
4. Wrap Up
PROPRIETARY AND CONFIDENTIAL, MARITZ COPYRIGHT 2009 3
1The struggle to “Put the Voice of the Customer (VOC) to Work”
PROPRIETARY AND CONFIDENTIAL, MARITZ COPYRIGHT 2009 4
Assessing Organizations Success in Using VOC
• Maritz conducted the Maritz Voice of the Customer Practices & Challenges Survey* in 2009 with managers– ‘Blue Chip’ companies
– 131 managers responded
– Multiple industry sectors
* Presented by Dr. D. Randall BrandtVice President, Customer Experience & LoyaltyMaritz Research
American Marketing AssociationMay 12 2009
PROPRIETARY AND CONFIDENTIAL, MARITZ COPYRIGHT 2009 5
Where Organizations Still Struggle
• Integrating multiple sources of VOC data to define priorities for improvement
• Demonstrating the link between customer and financial metrics
• Linking the VOC to internal operational and service metrics
• Integrating the VOC and the Voice of Employees
• Clarifying survey-based action items so that their owners know what to do or fix
• Pinpoint practices or business processes that must be improved
• Communicating action plans to address VOC-driven issues
• Getting managers/partners to act on VOC-driven issues
PROPRIETARY AND CONFIDENTIAL, MARITZ COPYRIGHT 2009 6
In a Nutshell...
• Managers desire to know the positive and negative aspects of a product or service to drive improvements and marketing activities
• But most organizations are not where they want to be when it comes to integrating and deploying insights drawn from the Voice of the Customer
Integrating
VOC Data
Link VOC toBusiness Results
Analyze
VOC Data
Customer survey metrics
Capture VOC Data
PROPRIETARY AND CONFIDENTIAL, MARITZ COPYRIGHT 2009 7
2Ideas for Putting the Voice of the Customer to Work
PROPRIETARY AND CONFIDENTIAL, MARITZ COPYRIGHT 2009 8
Our Vision for an Integrated CEM/CGM Offering
• Customer experience sources today can broadly be grouped into:
– Primary customer research and other ‘internal’ sources– Public media, including social media
• Each source has its biases and limitations. None have an exclusive license to the “truth”. The best, most robust picture emerges from the integration of all sources.
• This integration should be designed into a customer experience measurement and management process (Integrated Architecture):
– Built upon a holistic understanding of the business process and how those processes produce measureable business results (Blueprinting and Linkage)
– Supported by a common framework for combining findings across category – a Uniform Set of Customer Experience Categories
PROPRIETARY AND CONFIDENTIAL, MARITZ COPYRIGHT 2009 9
Considerations Important in Using VOC Information Sources
Social Media Survey Data
Cannot be sure where the author fits – a customer, an intender, awareness?
We know the respondent: they can be associated with specific locations, employees, products, transactions…
Data is ‘opt in’ and needs to be interpreted in that light.
Sample is ‘representative’ and provides a balanced snapshot.
Market “coverage” is very high. Surveys generally have far fewer observations.
Data is available continuously with dynamic updating of information, charts…
Data tends to be historical.
Documents contents are highly variable.
Documents, or comments, have focused content
Text data can be analyzed using a common structure to identify important insights!
PROPRIETARY AND CONFIDENTIAL, MARITZ COPYRIGHT 2009 10
Bottom Line…
Maritz perspective is social media monitoring coupled with survey research can help managers and marketers maximize success and make well-founded, informed decisions.
PROPRIETARY AND CONFIDENTIAL, MARITZ COPYRIGHT 2009 11
3
A Look at VOC Sources Using a Uniform Set of Customer Experience Categories
PROPRIETARY AND CONFIDENTIAL, MARITZ COPYRIGHT 2009 12
The Smartphone Product Launch Monitor
• Maritz monitored online media and surveyed over 3,000 US consumers in the evolving Smartphone market from November and December 2010
• Maritz and evolve24, a Maritz Research company, used a combination of:– Social media identification (looking for mentions of
interest in or buying Smartphone's on Twitter or Facebook)
– Surveys done among consumer panels• Topics include awareness, consideration, and
product likes/dislikes of Smartphone's among those who recently purchased or plan to in the near future
• Devices included those currently available, as well as those announced, but not yet released
• We developed a list of common categories using social media and our expertise in the market to compare text results
PROPRIETARY AND CONFIDENTIAL, MARITZ COPYRIGHT 2009 13
The Survey Open End Questions – Driving Sentiment into the Comments
• “You said that you eliminated [Smartphone] from those that you had been thinking about.” (Model Considered)– “What did you like about it that made you consider it in
the first place?”
– “And what didn’t you like about it that made you eliminate it from your list?”
• “What model did you purchase/would you buy?” (Model Purchased)– “What do you like about it that makes you choose this
model?”
– “Are there things that you don’t like about this model?”
For these comments and social media, we examine the volume and sentiment of responses by categories.
PROPRIETARY AND CONFIDENTIAL, MARITZ COPYRIGHT 2009 14
We Also Asked Which Categories were Most Important in the Consideration Phase
Smartphone price Price of voice plan Price of data service plan Wireless coverage quality Phone call sound quality Picture-taking capabilities/camera Dropped Calls Video-taking capabilities Music-playing services and capabilities Video-playing services and capabilities Availability of video games Look and feel of the device The quality of the applications loaded The keypad
Brand name Operating system Ease of use The number of available applications
loaded Technical performance Ability to synchronize Office applications Quality of the available technical support After sales service Battery life The screen resolution GPS capabilities The size of the screen Ability to access e-mail The app store that comes with the device
“Among the list below, please indicate the item you think is most important to your choice of Smartphone.”
We look at the percentage of respondents.
PROPRIETARY AND CONFIDENTIAL, MARITZ COPYRIGHT 2009 15
3A A Look at Sentiment…
PROPRIETARY AND CONFIDENTIAL, MARITZ COPYRIGHT 2009 16
Comment Sentiment
• For survey questions, focusing on dislikes and likes of Smartphone's yields expected sentiment
• Overall average across the four questions is 0.63– For Model Purchased, the average
is slightly higher at 0.66.
• This approach can be valuable in online, mobile, and mail survey modalities to identify both positive and negative information
Survey Sentiment (0.0 to 1.0)
Likes Dislikes
Model Purchased .94 .37
Model Considered .81 .41
Average Sentiment
% Topics Above
Average
% Topics Below
Average
.55 70% 30%
• For social media, we extract sentiment at the topic level
• Average is moderately lower than survey likes/dislikes average
PROPRIETARY AND CONFIDENTIAL, MARITZ COPYRIGHT 2009 17
3B
Brands:BlackberryHTC
PROPRIETARY AND CONFIDENTIAL, MARITZ COPYRIGHT 2009 18
Approach
• For the ‘likes/dislikes’ question, we identified the top 5 categories assigned by brand– In all cases, the 5 categories added up to more than
70% of the total number of assignments.
• For the survey quantitative approach, 10 or more categories are needed to cover that number of respondents
• We also look at the top 5 categories by brand for social media
PROPRIETARY AND CONFIDENTIAL, MARITZ COPYRIGHT 2009 19
3BBrands:Blackberry
PROPRIETARY AND CONFIDENTIAL, MARITZ COPYRIGHT 2009 20
BlackBerry Model Purchased LikesPercentage of Respondents Assigned
37%
22%
19%
15%
14%
The keypad
Ease of use
Price of device
Look & feel of device
Ability to access e-mail
Ca
teg
ori
es
BlackBerry Model Considered LikesPercentage of Respondents Assigned
40%
27%
19%
13%
10%
The keypad
Look & feel ofdevice
Ability to access e-mail
Ease of use
Brand name
Ca
teg
ori
es
• 4 out of 5 categories are consistent
• User interface stands out:• Ease of use• Keypad• Look & feel
• E-mail also consistent
PROPRIETARY AND CONFIDENTIAL, MARITZ COPYRIGHT 2009 21
BlackBerry Model Considered DislikesPercentage of Respondents Assigned
33%
29%
15%
15%
13%
Look & feel of device
Price of device
Ease of use
Number available appsloaded
Technical performance
Ca
teg
ori
es
BlackBerry Model Purchased DislikesPercentage of Respondents Assigned
44%
29%
16%
15%
10%
Look & feel of device
The keypad
The size of the screen
Technicalperformance
Price of device
Ca
teg
ori
es
• Fewer categories are consistent
• The user interface also stands out:• Look & feel• Screen size• Keypad• Ease of use
• Some concern about price and performance
PROPRIETARY AND CONFIDENTIAL, MARITZ COPYRIGHT 2009 22
From the Quantitative Perspective,
BlackBerry Most Important Purchase Reason Percentages
10%
9%
9%
9%
8%
7%
6%
6%
5%
5%
5%
Price of Device
Coverage quality
Ease of use
The keypad
Price of data service plan
Ability to access e-mail
Ability to synchronize Office apps
Technical performance
Call sound quality
Operating system
Price of voice plan
Cat
ego
ries
• Also see the user interface may be an issue, but doesn’t stand out:• Keypad• Ease of use
• Price shows up more
• E-mail, while not on top, does show up
PROPRIETARY AND CONFIDENTIAL, MARITZ COPYRIGHT 2009 23
BlackBerry Social Media Top Five Categories
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Oct Nov-Dec Jan
Time Period
Ease of use
Picture-taking capabilities
Battery life
App store with device
BTW: This increase demonstrates the need for monitoring social media to catch escalating topics…
• For Nov/Dec, the survey timeframe, we see ‘ease of use’ is the only user interface category – it may be getting lost in the mix
• E-mail is consistent with what customers ‘like’ about Blackberry• The discussion of other features point to a more ‘functional’ focus
PROPRIETARY AND CONFIDENTIAL, MARITZ COPYRIGHT 2009 24
Large number of tweets referring to getting ‘free’ products…
PROPRIETARY AND CONFIDENTIAL, MARITZ COPYRIGHT 2009 25
For Blackberry…
• User interface shows up as an important area in what consumers like/dislike in the survey
• UI does show up in the most important consideration, but not a top issue
• In social media, less emphasis on UI and more focus on features – may be related to less ‘hands on’ experience
PROPRIETARY AND CONFIDENTIAL, MARITZ COPYRIGHT 2009 26
3BBrands:HTC
PROPRIETARY AND CONFIDENTIAL, MARITZ COPYRIGHT 2009 27
HTC Model Considered LikesPercentage of Respondents Assigned
21%
20%
18%
16%
16%
Look & feel ofdevice
Operating system
Price of Device
Technicalperformance
The size of thescreen
Ca
teg
ori
es
HTC Model Purchased LikesPercentage of Respondents Assigned
41%
23%
18%
16%
15%
Look & feel of device
Technical performance
The size of the screen
Price of Device
Picture-takingcapabilities
Ca
teg
ori
es
• 4 of 5 categories consistent
• The user interface shows up:• Look & feel• Screen size
• Performance is important• Technical• OS
• Price of device is also important
PROPRIETARY AND CONFIDENTIAL, MARITZ COPYRIGHT 2009 28
HTC Model Considered DislikesPercentage of Respondents Assigned
39%
25%
14%
10%
8%
Price of Device
Look & feel of device
The keypad
Battery life
Technical performance
Ca
teg
ori
es
HTC Model Purchased DislikesPercentage of Respondents Assigned
33%
33%
17%
14%
5%
Look & feel of device
Battery life
The keypad
Price of Device
The size of the screen
Ca
teg
ori
es
• Very consistent results
• UI is still a concern:• Look & feel• Keypad• Size of screen
• Battery life surfaces
• Price of device is a concern
PROPRIETARY AND CONFIDENTIAL, MARITZ COPYRIGHT 2009 29
HTC Most Important Purchase Reason Percentages
12%
12%
10%
9%
8%
8%
7%
5%
5%
3%
Technical performance
Operating system
Price of Device
Coverage quality
Ease of use
Call sound quality
Price of data service plan
Price of voice plan
Ability to access e-mail
Dropped Calls
Ca
teg
ori
es
• Performance surfaces as a top issue (technical, OS)• Price does shows up more• UI and battery life do not show up at all
PROPRIETARY AND CONFIDENTIAL, MARITZ COPYRIGHT 2009 30
HTC Social Media Top Five Categories
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Oct Nov-Dec Jan
Time Period
Operating system
Picture-taking capabilities
App store with device
Battery life
Same issue with HTC as Blackberry…
• UI does not surface• Performance shows up in OS• Battery life is consistent with comments in survey,
but not the checklist
PROPRIETARY AND CONFIDENTIAL, MARITZ COPYRIGHT 2009 31
For HTC…
• User interface shows up as an important area in what consumers like/dislike only in the survey comments
• OS may be a core area for HTC – shows up in all three sources
• Battery life surfaces in comments and social media, and may be an area to look into
PROPRIETARY AND CONFIDENTIAL, MARITZ COPYRIGHT 2009 32
4 Wrap UP
PROPRIETARY AND CONFIDENTIAL, MARITZ COPYRIGHT 2009 33
Summary…
• Uniform Common Categories provides a basis for comparing information from multiple text sources
• Survey data is a balanced approach for discerning topic areas impacting products and services
• Asking likes/dislikes helps drive sentiment into comments to get a balanced view– A consideration for ‘rating websites’?
• Social media measured dynamically is critical for identifying escalating topics and measuring effects of actions taken
• All three sources add value and insight into the ‘picture’ for a given Smartphone
PROPRIETARY AND CONFIDENTIAL, MARITZ COPYRIGHT 2009 34
Questions
??