Nielsen isn’t abandoning set-top-box (STB) data, but it is substantially scaling back plans to use it for local measurement. It boils down to the need for speed.
In the 55 largest markets, Nielsen has determined -- at least for now -- that it’s not possible to collect and process the data in a timely enough fashion to continue with overnight ratings, a major client need. So, Nielsen has altered the new “hybrid” system it laid out last summer.
In the 25 markets with local people meters (LPMs) and 30 with set meters, plans had called for ratings to come from a mixture of the current panel method; new code readers; and STB data (technically return-path data). Now, ratings in those markets will be generated by the traditional methods with code readers augmenting them.
Nielsen will use STB data, however, in all 155 diary markets if available. Some will have the three-source system -- diaries, code readers and STB information. An undetermined subset will have just diaries and STB data. Diary markets generally get ratings only four times a year, allowing ample time to process the unwieldy STB data that Nielsen collects from Charter and DirecTV.
There is one other arena where Nielsen is altering plans: sample size. In LPM and set-meter markets, Nielsen had planned to effectively quadruple the size. Now, it’s going to triple it initially. So, for example, LPM markets with 600-1,000 homes in the sample will go to 1,800-3,000.
(Diary markets were to experience an effective doubling, but Nielsen is still evaluating how the sample size may be altered.)
As for the rollout, Nielsen has finished installing the code readers (which pick up audio signals) in LPM markets Dallas and Charlotte, while it should finish in St. Louis next month. Then, in August, installation will begin in five set-meter markets: Albuquerque, Birmingham, Ala., New Orleans, Nashville and Greenville, S.C. It will also start in four diary markets: Flint, Mich., Madison, Wis., Reno, Nev. and Santa Barbara, Calif. All markets offer data from Charter and DirecTV.
In August, initial preview data from Dallas, Charlotte and St. Louis will become available for advertisers and station groups to evaluate. After six months, there will be another six-month parallel phase, where clients can integrate data in workflow systems for evaluation. The "hybrid" data won’t be ready to serve as a currency, though, until at least 2014.
Some media buyers and station executives have touted the value of STB data with hopes it would improve the current Nielsen methods. But Nielsen executive vice president Matt O’Grady said industry executives have been understanding about the change in plans necessitated by the ungainly STB data.
“There was not a groundswell of concern," he said.
He added the call was to “deliver the best quality data” possible. Nielsen has been working with the Media Rating Council (MRC), which accredits measurement processes, during development.
The “hybrid” system has two aims with its larger sample size: bring more stability to ratings and eliminate “zero cells,” which could indicate no one in a particular demo was watching a program.
Pat Liguori, a top researcher for the ABC-owned stations, said Nielsen made a “prudent decision” to drop STB data in the metered markets in order to allow for overnight numbers. There was some talk about releasing overnight data without STB contributions and then another batch later incorporating it, but Liguori said there wasn’t much hunger for the complications that could bring.
“Most of us felt that (would) muddy the waters,” she said. “We have too many data streams as it is.”
As for disappointment with the lack of STB data, Liguori said there remains an opportunity to purchase it from Rentrak, which ABC does in several markets.
Rentrak, whose measurement system is propelled by STB data from Charter, AT&T and Dish, also doesn’t offer overnight ratings (its final numbers come in nine days after air). But, it has plans to produce overnight numbers – likely derived from a subset of its total STB pool.
“I’m not promising that it’s going to be tomorrow,” said Rentrak Chief Research Officer Bruce Goerlich. “It’s on our to-do list. We have thoughts on how to do it. It is doable.”
Charles Chunn, who heads research for the Post-Newsweek group, wrote in an email that the company is encouraged by Nielsen’s efforts to boost sample size, but it is “still too early to tell what the real-world impact” will be. Post-Newsweek has stations in metered markets, so without STB data, the code readers will be critical to its ultimate opinion.
For his part, O’Grady says there is a lot of work to convince the likes of Chunn that Nielsen can upgrade long-criticized local measurement with its new system. “We have to prove this to the clients,” he said.