Social Media Intelligence: Measuring Brand Sentiment from Online Conversations
David A. Schweidel, Wendy W. Moe, and Chris Boudreaux, 2012, 12-100
Increasingly, businesses are turning to social media as a source of market research. Comments posted on social networking sites, blogs and microblogs, and discussion forums have provided a wealth of data from which marketers have been trying to extract metrics pertaining to the health of their brand. Traditionally, carefully designed surveys have been employed for this purpose. Today, with the proliferation of social media, marketers have turned to social media to “listen in” on the conversations surrounding their brand.
However, key differences exist between traditional research methods and social media listening. A well-designed survey identifies and targets the relevant respondent population; and questions are structured to focus on specific topics of interest, with care taken to avoid any potential response biases. In contrast, social media provides an unstructured and open forum allowing anyone to comment on any topic of interest to them.
In this environment, researchers have identified several factors that influence posted opinions, ranging from venue effects, where the choice of where to post is related to what you post, to social dynamics, where the social interactions in the venue alter the opinions subsequently expressed. As a consequence, metrics based on opinions expressed in social media are often not comparable to those expressed through a well-designed survey.
In this study, David Schweidel, Wendy Moe, and Chris Boudreaux investigate the potential to infer brand sentiment from social media conversations. Their analysis employs data collected from a variety of social media domains. Controlling for various factors that can influence the posted opinion, the authors propose a hierarchical Bayesian regression model and derive a measure of online brand sentiment.
The authors apply their model to data pertaining to a leading enterprise software brand and show how their proposed approach provides an adjusted brand sentiment metric that is correlated with the results of an offline brand tracking survey (correlation = .604). In contrast, a simple average of sentiment across all social media comments is uncorrelated with the same offline tracking survey (correlation = -.002). Their findings show systematic differences in sentiment expressed across different social media venues and across different posters. Additionally, their method provides a tool with which to decompose overall sentiment into an underlying brand sentiment versus sentiment focused on specific products in the brand portfolio or attributes of the brand. The authors further apply their model to a number of brands across different industries and demonstrate potential pitfalls associated with simple average sentiment measures.
Their findings demonstrate the potential for social media to be incorporated into the brand’s research activities; however, these activities must be undertaken with care. Monitoring a single type of venue would not allow managers to distinguish venue-specific factors from the general impressions of the brand. However, firms may be able to infer overall brand sentiment from a broader sample of comments drawn from multiple venues, with consideration given to differences in the comments’ focal attributes and products, posting venue, and customer experiences.
David A. Schweidel is Assistant Professor of Marketing, Wisconsin School of Business, University of Wisconsin, Madison. Wendy W. Moe is Associate Professor of Marketing, Robert H. Smith School of Business, University of Maryland. Chris Boudreaux is Sr. Vice President of Business Integration, Converseon, Inc.
Social Media Intelligence
Wendy Moe (2013) [Video]
Getting More-Reliable “Intelligence” from Social Media Data (2013) [Article]
Measuring Online Brand Sentiment (2012) [Article]
Online Product Opinions: Incidence, Evaluation and Evolution
Wendy W. Moe and David A. Schweidel (2010) [Report]
- Corporate: FREE
- Academic: FREE
- Subscribers: FREE
- Public: $0.00
3 WAYS to GET CONNECTED
Employees of MSI Member Companies enjoy the benefits of complete online access to content, member conferences and networking with the MSI community.
Qualified academics benefit from a relationship with MSI through access to MSI.org, conferences and research opportunities.
The public is invited to enjoy partial access to msi.org content, a free e-newsletter, selected reports and more.