Social Media Proves a Powerful Measure of Voter Sentiment and Accurate Predictor of Gubernatorial Race

Bryan Merica
Co-Founder of Unearth, a digital-first agency and Co-Founder of Fox & Hounds

My company, Activate Direct, teamed up with Tulchin Research and PWSMC Social Media consulting, to release a detailed study of social media content related to the 2010 California governor’s race between candidates Meg Whitman and Jerry Brown. The study demonstrates how campaigns can use techniques of "social listening" as both a real-time poll and an ongoing focus group, augmenting traditional public opinion research methods and identifying potential crises early.

Study Highlights
The analysis unlocked several key findings:

  • Social and polling data were closely correlated.
  • The ratio of positive to negative social sentiment was very much in line with the ratio of favorable to unfavorable ratings shown by traditional polling.
  • Social chatter was driven by key campaign events.

By analyzing the daily volume of social media chatter over the campaign timeline, it is clearly evident that peaks in social conversation volume coincided with major campaign events. There were three major peaks the team observed: 

  • Brown’s announcement that he would run for Governor  
  • The primary election
  • The largest peak of all, the Nannygate scandal

The impact of Nannygate was significant. While Brown also had a negative spike during this time because of a related scandal ("Whoregate"), the overall gap between his positives and negatives is not nearly as far apart as Whitman’s during this time, nor did Brown’s negatives spike nearly to the same degree as Whitman’s did over the same period.

Study Methodology

  • The study compared social media and traditional polling data throughout the gubernatorial election cycle to determine the effectiveness of social listening at measuring voter sentiment and its correlation with election outcomes.
  • Social listening involves the use of advanced monitoring tools to identify and analyze relevant conversations on social media. Working with social media monitoring and analysis tools from Sysomos, the team captured hundreds of thousands of relevant comments.
  • The study analyzed comments from January 1, 2010 through Election Day, November 2, 2010. The volume of relevant social media comments was measured during this period along with the overall sentiment (positive, neutral, negative) of those comments.
  • More than 200,000 posts and comments were analyzed from blogs, forums and news outlets over the course of the study.
  • The study reviewed the overall favorability rating of the candidates in social media and compared these numbers to those of traditional polling to better understand how social listening and polling can be effectively utilized together for political and public affairs initiatives.

The full report is available here.

Comment on this article


Please note, statements and opinions expressed on the Fox&Hounds Blog are solely those of their respective authors and may not represent the views of Fox&Hounds Daily or its employees thereof. Fox&Hounds Daily is not responsible for the accuracy of any of the information supplied by the site's bloggers.