The proof that tweeting boosts TV ratings

Nielsen has just confirmed what we have known for some time now, that tweeting does boost TV watching. If there is a surprise, it is that the correlation does not appear to be universal among TV shows, although that may reflect lack of certainty about the level of statistical significance in more borderline cases. Nielsen used time series analysis techniques developed by Nobel Prize-winning economist Clive Granger, already well proven in a variety of fields such as econometrics, meteorology and neuroscience. This takes measured twitter activity around the time a program is aired and allocates diminishing weights to relevant tweets the longer they are after the end of the show. Nielsen analyzed minute-by-minute trends in its Live TV Ratings and Tweets in this way for 221 broadcast primetime program episodes, using its SocialGuide, for this Twitter Causation Study. This revealed that Live TV ratings had a statistically significant impact on related Tweets among 48 percent of the episodes sampled. Correspondingly, the volume of Tweets caused statistically significant changes in Live TV Ratings among 29 percent of the episodes.

Unfortunately Nielsen does not give a breakdown between different types of show, given that it is likely audiences will be much more susceptible to social buzz for some categories than others. It is possible that viewing of games shows and drama series is more likely to correlate with twitter activity than, say, in the case of serious documentaries, which generate less buzz, or premium sports, which attract a peak audience anyway. This is certainly borne out by anecdotal evidence, as in the case of the screening in the U.S. by BBC America of the iconic UK sci-fi series "Doctor Who," which has been running there on and off for half a century. Apart from sports, the show was the number one telecast across social media on the Sunday it showed and attracted 895,000 total linear viewers, according to Nielsen. That is a high number for a non-sports program in early August, when many viewers are on vacation, especially for a cult British series many Americans had never heard of before. In that sense, the unexpected success can be attributed to the power of social media.

Yet even if Nielsen’s results were expected, they do provide that official confirmation that will persuade both programmers and advertisers to take Twitter more seriously. It will also help Twitter itself promote its TV advertising platform launched in May 2013, based on technology from its recently acquired subsidiary Bluefin Labs. This enables advertisers to extend TV commercials interactively in the twittersphere by targeting people who have just watched the show in which the spot ad appeared. Twitter detects viewers of the particular show, assumes they watched the ad, and sends them a tweet relating to the spot, which could, for example, invite the recipient to book a test drive, make a purchase or view a longer form of the ad.

We can note that some brands are already coordinating campaigns across TV and social media in the belief that the whole will be greater than the sum of its parts. In the U.K., for example, fast food group Subway is currently running a promotion featuring a spot ad introducing a story that is also played out in a Facebook game incorporating Google Street View to inject some realism. This trend can only grow.

Of course, all this still begs the question for broadcasters of how they can harness the power of social media directly and in near real time. There is all this social media interaction going on about shows while people are watching them, but it occurs on Twitter and Facebook outside their domain. The question has been how to bring it inside their walled gardens. U.K. broadcaster Channel 4 believes at least part of the answer lies in bringing the mountain to Mohammed if Mohammed cannot be enticed to the mountain. It is planning to launch an app that will incorporate a twitter feed inside its second-screen platform. The hope is that by providing a short cut to Twitter, it will persuade viewers to tweet that way, thereby bringing the activity into an environment it can control and measure. It will be interesting see how this develops.