Social chatter is a finger on the pulse of society, reflecting what we’re watching, reading, buying, and discussing. In some cases, analyzing the data reveals more than a reflection: a premonition, accurately correlating with and guessing at some future event.

Here are a few interesting examples of social data forecasting and/or mirroring events in the “real world.” We’ll dive deeper into the intersection of social and offline events in our upcoming webinar, What Twitter data is telling you about real consumers

Predicting the stock market

In 2010, academic researchers at Indiana University and the University of Manchester found that tweets can accurately predict the “up and down changes in closing values” of the Dow Jones Industrial average. Certain dimensions of “Twitter mood” were revealed to be 86.7% accurate in predetermining stock market value at the end of the trading day.

Our own research found that Twitter mention volume has a .91 positive correlation with brands’ stock prices. Things like positive news that make stocks move upward also tend to make Twitter chatter spike, while the things that sink stocks don’t incite the same conversation levels.

Projecting film box office performance

Twitter chatter can also predict box office performance for movie releases – more Twitter mentions means a higher grossing opening weekend. A recent study found a 0.90 correlation between film mentions on Twitter and ticket revenue. And predictions based on Twitter are consistently better than those produced by an information market such as the Hollywood Stock Exchange, the gold standard in the film industry.

And social has even had some success in foretelling the Oscar winners. In 2010, social chatter accurately anointed The King’s Speech as Best Picture and Natalie Portman as Best Actress for her performance in Black Swan. The analysis is still imperfect – missing on the big three categories last year – but advances in sentiment analysis might improve social Oscar conjectures in the future.

Foreseeing unemployment spikes and reflecting macroeconomics

A recent study found that analyzing social media conversation can predict spikes in the unemployment rate. Researchers looks at online conversation over a two-year period in both the US and Ireland, and found that increased talk around things like taking more mass transit, postponing vacations, etc. foreshadowed unemployment spikes in both countries.

Our own research also revealed that brand-related conversation reflects macroeconomic trends. Mentions of “price” in reviews show a -.66 correlation with the US Consumer Confidence Index – the less confidence consumers have in the economy, the more they talk about price in reviews. When comparing review references to price to the Dow closing index, we found a slightly stronger negative correlation of -.68.

Influencing voter turnout in elections

Analysis of millions of Facebook conversations on Election Day 2010 found that posts on the social network can have a measurable (if limited) effect on voter turnout. A special “get out the vote” message run on election day showed users pictures of their friends how had voted. The study showed that the message generated an additional 340,000 votes nationwide, but could not determine whether the votes were for Democrat or Republican candidates.

Echoing the contagiousness of positive emotion

Another study proves that Twitter conversation can identify society’s collective emotional state, based on “large-scale expressions of moods shared via social media.” The study also revealed a relationship between social ties formed by individuals and their shared moods – positive people flocked to other positive people, and negative to negative. Additionally, highly active social networkers are more likely to share predominately positive emotions. And when information is shared via links in Tweets, they’re more often associated with positive moods – possibly indicating that positive emotions are more contagious and share-worthy.

Register for our upcoming webinar to learn what Twitter data can predict about your business.

Leave a Reply

XHTML: You can use these tags: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>