NYU Stern
Share / Print

Beyond Likes and Tweets: Marketing, Social Media Content and Store Performance

By Craig Stacey

Craig Stacey Headshot Article

... despite the popularity of Likes and tweets metrics, these findings suggest that social media impact is not just about volume and valence: knowing “what” consumers say matters.

Social media is often tracked in volume and valence, such as Facebook Likes or positive versus negative Tweets. New monitoring tools also allow marketers to track the content of social media conversations, for example, consumers’ responses to the brand’s marketing, their emotional attachment to the brand or product or their reports of purchase behavior related to the brand. Marketers are greatly interested in how marketing activities can affect these social media metrics and to what extent each drives business performance.

Here, Koen Pauwels, Craig Stacey and Andrew Lackman address the following questions:
  1. To what extent does social media conversational content, versus quantity (volume) and sentiment (valence), explain business performance in the short term and the long term?
  2. How do paid marketing actions stimulate specific word-of-mouth (WOM) conversational content, and how do these indirect performance effects compare to direct marketing impact?
Their dataset includes marketing, store traffic and website traffic data from a major US retail brand; WOM data from Crimson Hexagon, a company that specializes in social media monitoring and analysis; and natural search data from Google.com. This yielded complete data for 55 weeks (July 2010 - July 2011). Their vector autoregression (VAR) model estimates how offline and online marketing induce online search and social media conversations and how each of these endogenous variables has direct and indirect effects on business performance.

Their results indicated that different social media content has substantially different performance implications and that marketing actions with small direct effects can have large total effects by stimulating social media conversations. TV, print and online marketing showed a substantial indirect effect through online search and/or the WOM content metrics of “love for the brand” and “love the ad.” The effects of radio were mostly direct on store sales, but then reinforced through “went there/purchased” WOM content.

Content-specific WOM performed better than Facebook and Twitter volume and valence metrics in explaining store traffic and online traffic. Among the conversation topics, “love for the brand” had larger long-term traffic effects, but neutral conversations on “went there/purchased” drove traffic in the short run.

Thus, despite the popularity of Likes and tweets metrics, these findings suggest that social media impact is not just about volume and valence: knowing “what” consumers say matters. Managers who understand how conversations affect their business, and how their marketing strategies influence online conversations, may leverage paid marketing to spark WOM in a way that can positively influence their brands and bottom line.

___
E. Craig Stacey is Director of Research, New York University Stern Center for Measurable Marketing and Founding Partner, Marketing Productivity Group. Koen Pauwels is Professor of Marketing, Ozyegin University. Andrew Lackman is Business Analyst, Marketing Productivity Group.