With social media gaining a strong foothold in the day-to-day lives of ordinary people, the private sector and other research groups have toyed around with the idea of using social media as a device for predictive analytics. While the 2008 U.S. Presidential race witnessed Barack Obama’s campaign embrace the power of sites such as Facebook and YouTube, it was only in the recently concluded Congressional elections where the body of research received the attention it needed.
According to Facebook’s own team of political analysts, the fans of political candidates on the site have accurately predicted over 70% of the election’s most observed races with just their “likes” alone. In a study that involved 98 key races for Congress, Facebook found that 74% of candidates who have the most fans on the site won their respective House seats. In addition, the social network’s results also showed that 82% of the Senatorial candidates with more Facebook fans won their races.
However, that’s not to say that the “predictions” of Facebook’s campaign fans are right on the money all the time. One of the Democratic Party’s most significant losses, incumbent Congressman Alay Grayson of Florida, had more than 30,000 fans on Facebook. But he still lost the seat to Republican challenger Daniel Webster, who only had a following of 4600 people on the site.
Nevertheless, these misses have not discouraged social media researchers from studying how to make use of the piles of data gleaned by from social networks. Analysts Sitaram Asur and Bernardo Huberman of HP Labs even went so far as to use Tweets from Twitter to predict the success of movies. While their tests were only applied to big Hollywood films, the results were still significant—so significant in fact, that it just might put the Hollywood Stock Exchange out of business.
Enform believes that with the way things are going, it’s just a matter of time before social media really changes the landscape of market research.