Social Media footprint : Developing a Key performance indicator based on word - of - mouse A CASE STUDY OF NOKIA Supervised by: Prof. Dr. Annouk Lievens Prepared by: Osama Dukhan
Social media footprint: Developing KPI based on Word-of-Mouse
May 12, 2015
This study uses content analysis to develop a key performance indicator based on customers’ word-of-mouse, collected randomly from three different social networking and user review sites. Using LIWC text-based analysis software, we conducted a content analysis of over 420 customers’ reviews and comments containing a reference to one of Nokia smartphones models. Our results show that each of Nokia smartphones models has a unique sentiment multi-dimensional profile. Thus, there are differences in customers’ attitudes toward Nokia products. Later, we based on this fact to build consumers’ attitudes index to measure Nokia’s performance from customers’ perspectives over time. Moreover, we go further by using this index to build an early warning system, which could lead Nokia to identify any potential improvements or even to predict potential future developments.
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Social Media footprint:Developing a Key performance indicator based on word-of-mouse
A CASE STUDY OF NOKIA
Supervised by: Prof. Dr. Annouk Lievens
Prepared by: Osama Dukhan
Outlines:
- Aim of study
- Research questions and answers
- Limitations
- Suggestions for future researches
- Conclusions
Aim of study:
The main goal in this research is finding a tool could help firms to see more through the eyes of customer, thereby act faster.
Aim of study:
“See More Act Faster”
Automated Sentiment Analysis
Performance management
Social MediaField of
Study
Figure (1) Field of study
Research Questions:
How we can use word-of-mouse to generate an accurate quantitative indicator?
Is it possible to measure the performance of the firm based on this KPI?
To what extent can we use this KPI to drive the Firm’s performance?
Research Questions:
How we can use word-of-mouse to generate an accurate quantitative indicator?
Is it possible to measure the performance of the firm based on this KPI?
To what extent can we use this KPI to drive the Firm’s performance?
Using WOM to generate an accurate quantitative indicator.
• Amazon.com
• C|net
Data Collection
Data Cleansing
Data Testing
KPI Developing
Using WOM to generate an accurate quantitative indicator.
Data Collection
• Misspelling.
• The quality of reviews/comments
Data Cleansing
Data Testing
KPI Developing
Using WOM to generate an accurate qualitative indicator.
Data Collection
Data Cleansing
• Consumer’s feelings and attitudes toward different products
Data Testing
KPI Developing
Data Testing
Multi-dimensional profiles of Nokia Smartphones
Anger
Using WOM to generate an accurate qualitative indicator.
Data Collection
Data Cleansing
Data Testing
• \PA\ Positive attitudes index
• \NA\ Negative attitudes Index
KPI Developing
Data Testing
Moving from Qualitative data to Quantitative data
• Amazon.com
• C|net
Data Collection
• Misspelling.
• The quality of reviews/comments
Data Cleansing
• Consumer’s feelings and attitudes toward different products
Data Testing
• Positive attitudes \PA\index
• Negative attitudes \NA\Index
KPI Developing
Research Questions:
How we can use word-of-mouse to generate an accurate quantitative indicator?
Is it possible to measure the performance of the firm based on this KPI?
To what extent can we use this KPI to drive the Firm’s performance?
• Amazon.com
• C|net
Data Collection
• Misspelling.
• The quality of reviews/comments
Data Cleansing• Consumer’s feelings
and attitudes toward
different products
Data Testing
• \PA\ Index
• \NA\ Index
KPI Developing
• Time variable as a moderator
Performance Measurement
# 1st Step # 2nd Step
Nokia – Smartphones released over one year
# Name of product (Model) Announce date Release date
1 Lumia 925 May, 2013 June, 2013
2 Lumia 928 April, 2013 May, 2013
3 Lumia 720 Feb, 2013 April, 2013
4 Lumia 520 Feb, 2013 April, 2013
5 Lumia 505 Dec, 2012 Jan,2013
6 Lumia 620 Dec, 2012 Jan,2013
7 Lumia 822 Oct,2012 Nov,2012
8 Lumia 510 Oct,2012 Nov,2012
9 Lumia810 Oct,2012 Nov,2012
10 Lumia 920 Sep,2012 Nov,2012
11 Lumia 820 Sep,2012 Nov,2012
12 Lumia 610 NFC April, 2012 Sep,2012
13 800C Mar, 2012 Sep,2012
14 808 Pure view Feb, 2012 June, 2012
15 Lumia 900 Feb, 2012 May,2012
Nokia - Smartphone
Nokia – Smartphones released over one year
# Name of product (Model) Announce date Release date
1 Lumia 925 May, 2013 June, 2013
2 Lumia 928 April, 2013 May, 2013
3 Lumia 720 Feb, 2013 April, 2013
4 Lumia 520 Feb, 2013 April, 2013
5 Lumia 505 Dec, 2012 Jan,2013
6 Lumia 620 Dec, 2012 Jan,2013
7 Lumia 822 Oct,2012 Nov,2012
8 Lumia 510 Oct,2012 Nov,2012
9 Lumia810 Oct,2012 Nov,2012
10 Lumia 920 Sep,2012 Nov,2012
11 Lumia 820 Sep,2012 Nov,2012
12 Lumia 610 NFC April, 2012 Sep,2012
13 800C Mar, 2012 Sep,2012
14 808 Pure view Feb, 2012 June, 2012
15 Lumia 900 Feb, 2012 May,2012
Nokia - Smartphone
Nokia – Smartphones released over one year
# Name of product (Model) Announce date Release date
1 Lumia 925 May, 2013 June, 2013
2 Lumia 928 April, 2013 May, 2013
3 Lumia 720 Feb, 2013 April, 2013
4 Lumia 520 Feb, 2013 April, 2013
5 Lumia 505 Dec, 2012 Jan,2013
6 Lumia 620 Dec, 2012 Jan,2013
7 Lumia 822 Oct,2012 Nov,2012
8 Lumia 510 Oct,2012 Nov,2012
9 Lumia810 Oct,2012 Nov,2012
10 Lumia 920 Sep,2012 Nov,2012
11 Lumia 820 Sep,2012 Nov,2012
12 Lumia 610 NFC April, 2012 Sep,2012
13 800C Mar, 2012 Sep,2012
14 808 Pure view Feb, 2012 June, 2012
15 Lumia 900 Feb, 2012 May,2012
Nokia - Smartphone
Figure 9: The volatility of PA index and NA index over one year
0
1
2
3
4
5
6
May-12 Jun-12 Jul-12 Aug-12 Sep-12 Oct-12 Nov-12 Dec-12 Jan-13 Feb-13 Mar-13 Apr-13 May-13 Jun-13
Positive Attitudes or \PA\ index Negative Attitudes or \NA\ Index
Figure 7: NOKIA's Customers attitudes over one year
0.00
1.00
2.00
3.00
4.00
5.00
6.00
May-12 June-12 July-12 August-12 September-12
October-12 November-12
December-12
January-13 February-13 March-13 April-13 May-13 June-13
Positive attitudes index Negeative Attitudes Index
Research Questions:
How we can use word-of-mouse to generate an accurate quantitative indicator?
Is it possible to measure the performance of the firm based on this KPI?
To what extent can we use this KPI to drive the Firm’s performance?
• Time variable as a moderator
Performance Measurement
• Developing early warning system
Performance Drivers
# 2nd Step # 3rd Step
Warning System:
We identified many alarm signals during the study period.
We conducted a simple comparison between Nokia’ stock performance
(NOK), as a one of financial KPI, and consumers’ attitudes index as a non-
financial KPI.
We found that it could reflect to what extent both financial and non-
financial indicators can work together harmoniously.
Figure 7: NOKIA's Customers attitudes over one year
0.00
1.00
2.00
3.00
4.00
5.00
6.00
May-12 June-12 July-12 August-12 September-12
October-12 November-12
December-12
January-13 February-13 March-13 April-13 May-13 June-13
Positive attitudes index Negeative Attitudes Index
Figure 7: NOKIA's Customers attitudes over one year
0.00
1.00
2.00
3.00
4.00
5.00
6.00
May-12 June-12 July-12 August-12 September-12
October-12 November-12
December-12
January-13 February-13 March-13 April-13 May-13 June-13
Positive attitudes index Negeative Attitudes Index
NOKIA’s Stock performance during the period (1st of June, 2012 till 15th of August, 2013)
Early warning system
0.00
1.00
2.00
3.00
4.00
5.00
6.00
Positive attitudes index Negeative Attitudes Index
Early warning system
0.00
1.00
2.00
3.00
4.00
5.00
6.00
Positive attitudes index Negeative Attitudes Index
Pros and Cons:
We used word-of-mouse to generate an accurate quantitative indicator.
We measured the performance of Nokia based on this KPI.
We used this KPI to build a warming system.
Limitations:
We can classify these limitations into Three different categories :
Limitations related to data.
Limitations related to the selected sample.
Limitations related to the methodology.
Limitations:
We can classify these limitations into three different categories :
Limitations related to data.
Limitations related to the selected sample.
Limitations related to the methodology.
Limitations related to data:
Example :
“Dropped it from a two feet high and the screen broke, too sad that I have been using it for only one month :( ”
Reason of ignore:
Based on LIWC2007 dictionaries, this text will be considered as anegative review, but in fact, the implicit feeling refers to high loyalty andpositive attitudes toward this Smart phone.
Limitations:
We can classify these limitations into three different categories :
Limitations related to data.
Limitations related to selected sample.
Limitations related to the methodology.
Limitations related to the selected sample.
Not all online discussion groups or user review sites have reviews
regard all set of products (models) which were selected in our
sample.
Furthermore, this research was conducted only on a small size of
population. Therefore, to generalized the result, the bigger size of
sample, the better.
Limitations:
We can classify these limitations into three different categories :
Limitations related to data.
Limitations related to selected sample.
Limitations related to the methodology.
Limitations related to the methodology.
we used the release date of products as a milestone over the study
period. However, these milestones were not always divided by
equal time lags because launching products to the market was not
conducted based on a regular time base. This point could generate
a bit of confusion, especially when we conducted a comparison
between Nokia’ stock performance (NOK) and consumers’
attitudes index.
Figure 9: The volatility of PA index and NA index over one year
0
1
2
3
4
5
6
May-12 Jun-12 Jul-12 Aug-12 Sep-12 Oct-12 Nov-12 Dec-12 Jan-13 Feb-13 Mar-13 Apr-13 May-13 Jun-13
Positive Attitudes or \PA\ index Negative Attitudes or \NA\ Index
Suggestion for the future researches:
it becomes to invest these techniques to assess customer opinions and toidentify what topics customers are talking about, and that will be useful tocharacterize the opinions that they express about those topics.
It is a due time to identify a wide range of opinion expressions, includingmotivations, recommendations and speculations.
It is a due time to use word-of-mouse as a tool to driving firms’performance not just use it as a performance measurement.
Conclusion:
• In contrast with many studies, this research mainly focused on the‘second way’ of social media as a communication channel, by trying toanswer the question; to how extent we can use the data collectedthrough customers’ review and comments to develop a KPI.
• this indicator could be an easy, timely, and inexpensive source ofinformation for many stakeholders, like the management, customers,competitors, shareholders, consumer protection funds and so on.
Conclusion:
• Finally, we have to admit that developing KPI through the eyes ofcustomer could be a more promising investment for:
• performance-focused companies which have Business-to-Customer(B2C) business model and their products have a plenty amount ofcustomers’ reviews and comments over various online discussiongroups and user review sites.
Thanks for your attention !
Social Media footprint:Developing a Key performance indicator based on word-of-mouse
A CASE STUDY OF NOKIA
For any question; Please don’t Hesitate to contact me
Related Documents
