SAN FRANCISCO (Reuters) - A frenzy of Twitter chatter about a broadcast program generally means more viewers are tuning in, while more popular TV shows generate more Twitter activity, a new study by Nielsen has found.
The media measurement firm says that for the first time it has proved with statistical rigor what many in the media business have suspected in recent years: that a show's ratings and the magnitude of its Twitter buzz are correlated.
The findings will be announced on Tuesday by Nielsen, who claims the results back up the premise of a new ratings system it will soon unveil.
Nielsen and Twitter are set to begin publishing this fall a new "Nielsen Twitter TV Rating" that measures the volume of Twitter conversation about every program.
The privately-held microblogging company has long sought to convince large brand marketers that it is a potent, real-time ad-delivery platform that complements and augments the live television-viewing experience. Marketers on TV can simultaneously reach their audience through a "second screen" because Twitter users often watch TV while tweeting away on their phones, Twitter has previously argued.
In the new study, Nielsen researchers analyzed a total of 221 episodes. The volume of tweets caused statistically significant changes to live TV ratings for 29 percent of the episodes, Nielsen said, without quantifying the degree of Twitter's influence on ratings.
The companies said the analysis methodology used for their study was developed by Nobel Prize-winning economist Clive Granger.
(This story deletes comments from Nielsen executive Mike Hess; the comments were taken from an earlier press release and relate to a different study; also corrects 3rd paragraph to make clear only Nielsen would announce the study results, not both Nielsen and Twitter)
Reporting by Gerry Shih; editing by David Evans