For months now we’ve been compiling data on upcoming films as they have been discussed within social media, looking at conversational volume, sentiment, and viewing interest so that we may have a good idea as to what’s going to be hot and what’s going to be a flop. Now it seems that someone else is on the scent as well. Word came to us last week via Mashable, the most obsessive social media news site on the web, that a study had been published demonstrating how Twitter has a high predictive power when it comes to box office performance. Suffice it to say, we weren’t too surprised.

It turns out that studying Twitter volume and sentiment around films allows for better predictability than the current industry favorite. By applying the methods we’ve been using, they found that box office performance predictability was as much as 1-2% better, a couple of percent, as Mashable points out, does represent millions at the box office. But that’s not all.

We know from experience that integrating the broader spectrum of social media data (i.e. cinephile sites, Facebook, blogs and forums) with this Twitter data provides a more comprehensive view. As an example, below is a type of graph we’ve been building and sharing with our clients in the Media & Entertainment space:

This chart displays social media conversational volume over the 9 weeks prior to the release of “Pandorum”, paired with both the overall norm for horror flicks, as well as a norm for films that opened under $10 million in box office revenue the first weekend. You can see that “Pandorum” followed the trend of other Horror films, but experienced less unaided awareness discussion 4 weeks before release. The box office norm came from a group of films whose opening weekend generated an average of about $7 million in domestic totals and which received, in the week prior to opening weekend, an average of about a thousand posts in social media. “Pandorum” only saw about 800 conversations. So, what was its opening weekend total? $4.4 million. Bingo.

Box Office indication is one of myriad ways to wield the power and volume of social media conversation, and it’s a great way to demonstrate the effectiveness of the kind of work we do here at CI. As we continue to integrate new resources and improve the robustness of our data, we look forward to a future of knowing, not guessing; of listening, and not (like traditional survey-based film trackers) asking.