What is Social Media Intelligence? 2 ... âSentiment and Venue Choice in Social Media Posting .... At a minimum, dashbo
Social Media Intelligence Wendy W. Moe
[email protected] www.wendymoe.com @wendymoe David A. Schweidel
[email protected] www.dschweidel.com @dschweidel
What is Social Media Intelligence? • The challenge facing organizations
• Social Media Intelligence – Understand the science of opinion and the implications for social media data and metrics – Measure what matters (not what’s easy to measure) 2
INTEGRATION
Social Media Intelligence by Wendy W. Moe and David A. Schweidel, Cambridge University Press, coming Fall 2013
Building Social Media Intelligence WHAT? WHERE? WHO?
“Online Product Opinions: Incidence, Evaluation and Evolution,” Marketing Science, 2012, with David A. Schweidel “The Value of Social Dynamics in Online Product Ratings Forums,” Journal of Marketing Research, 2011, with Michael Trusov
“Sentiment and Venue Choice in Social Media Posting Behavior,” under review at JMR, with David A. Schweidel
“Drivers of Social Media Rebroadcasting: Investigating the Role of Message Content and Influencers,” under review at JMR, with Yuchi Zhang and David A. Schweidel 3
WHAT do we say?
4
Social Media Posting Decisions Opinion / Brand Evaluation Do I post?
What do I post?
Where do I post?
Sentiment Product Attribute
Venue Format Domain
5
Rating Trends Average Variance
Activists Post frequently Attracted by lack of consensus More negative Variance and volume make them more negative
Low Involvements Post infrequently Deterred by lack of consensus More positive Variance and volume make them more positive 6
Implications
7
What should we do? • If you’re listening, don’t over-react. The ratings environment does not necessarily reflect the opinions of the entire customer base • “Listen to ALL the American people” – Encourage diversity of opinion – Monitor different venues / data sources
8
WHERE do we post?
9
Social Media Posting Decisions Opinion / Brand Evaluation Do I post?
What do I post?
Where do I post?
Sentiment Product Attribute
Venue Format Domain
10
What affects posted sentiment?
General Brand Impression (GBI)
Venue
Venue-specific dynamics
Message topic 11
Does it Matter Where We Listen? Proportion of Positive Comments
From a sample of social media comments pertaining to an enterprise software brand (courtesy of Converseon) 100% 80% 60%
Blog Forum
40%
Microblog
20%
Aggregate
0% 1
2
3
4
5
6
7
8
9
10 11 12 13 14 15
Observation Month
Distribution of Comments
100% 80% 60%
Blog Forum
40%
Microblog Other
20% 0% 1
2
3
4
5
6
7
8
9
10 11 12 13 14 15
Observation Month
12
Plus, sentiment metrics vary depending on topic
13
GBI vs. Offline Survey 0.4
9.1
0.3
9.05
0.2
9
0.1
8.95
0 -0.1
8.9 1
2
3
4
5
6
7
8
9
10
-0.2 -0.3
8.85 8.8
Month of Overlap Period (t) GBI in month t-1
Survey in month t
Average Survey Response
GBI
• Potential for GBI as a lead indicator • Correlation with survey (t) – – – – –
8.75
GBI = .376 Avg sentiment =.008 Blogs = .197 Forums = -.231 Microblogs = .394
• Correlation with survey (t-1) – – – – –
GBI = .881 Avg sentiment = .169 Blogs = .529 Forums = .213 Microblogs = .722 14
GBI vs. Monthly Close of Stock Price Regressed brand’s stock price against GBI (controlling for market index). Across iterations… Mean Coeff
% p-val