Our team has been talking a lot about how we measure sentiment, conversation, velocity, etc. We use various tools to paint the picture for our clients of who is talking about them, what they are talking about and if the conversation is positive or negative.
I found this article from Mashable interesting (and not surprising) on how Twitter played a role in predicting the closest Iowa Caucus in history. Globalpoint conducted a Twitter sentiment analysis that combined candidate tweet volume with the sentiment surrounding each candidate. The data they saw matched up closely with pre-Iowa polling data from NBC. The biggest difference in the data was that unlike the NBC pre-polling data, the Twitter sentiment analysis predicted that Rick Santorum (who lost by only 8 votes) would be the front runner. Michael Urban, President of Globalpoint said that “many polls showed Santorum moving to the top tier of candidates, but didn’t have numbers even close to the 25% he gained in Iowa.” “Twitter is indeed a leading indicator of public opinion,” stated Urban, “but it leads polling data and elections by about two weeks.”
We believe that Twitter is an excellent resource to learn from and we continue to spend time investing in ways to measure how conversation on Twitter affects our brands. It’ll be interesting to follow Mashable as they continue to report on Twitter’s role in predicting elections throughout 2012.