Nielsen initiative faces hurdle

Statistics glitch goes unnoticed for nearly eight months

Nielsen’s Extended Screen research initiative, which aims to broaden viewing measurements to include both online and TV stats, is off to a rough start.

The company issued a startling mea culpa to some of its subscribers on Wednesday, saying that since the end of January, some of the statistics measured by new service Npower have been inflated due to a software glitch.

The affected figures include average frequency numbers (indicating how often a show was watched over the measurement period) and qualified AA GRP numbers, which tell subscribers how many viewers watched the show for a significant period of time. Those stats are “likely overstated in most cases,” Nielsen admitted in an email to clients.

That may sound dire, but industryites say they’re annoyed rather than horrified. “It has nothing to do with C3 impressions,” said one source. C3 impressions are a major currency for advertisers, since they track live average commercial delivery plus three days of commercial playback on DVR systems. The glitch is unlikely to force many networks to offer make-goods.

Still, the mistake — and the length of time during which it went unnoticed — is unlikely to boost confidence in the new system.

Nielsen remains the industry standard in television ratings, but the digital market (which the Extended Screen program tries to quantify) is facing competition from digital measuring agencies ComScore and Quantcast.

For now, the company is apologizing and offering to rework the data. “Once we have determined the root cause, a software update will need to be implemented to correct the AVG Frequency calculation,” the company said. “Once the software update has been implemented, Nielsen will re-run all impacted reports.”

The Media Ratings Council, which audits ratings services, said it is investigating the errors through accountancy firm Ernst & Young.