Nielsen turns ratings eye on ads
This fall, the research company will focus in on how many actually watch commercials, not just TV shows.
By ASSOCIATED PRESS
Published July 12, 2006
NEW YORK - Television advertisers have long wanted to know how many people actually watch their commercials, but until now ads have been sold based on the average viewership of a whole program.
That could be on the verge of changing, however, now that Nielsen Media Research is preparing to offer data for the first time on how many people are tuned in during the commercial breaks.
Jack Loftus, a spokesman for Nielsen, a unit of the privately held company VNU, says the major networks asked the research company several months ago to provide the new data, and Nielsen, whose TV rating business employs 1,800 in Oldsmar, plans to begin doing so this fall on a test basis.
However, much negotiation remains to be done among advertisers and the networks over how the data will be used and in what form, Loftus said, adding that the earliest it could be adopted as an arbiter of value would be the advance ad-selling season for the fall 2007 season, called the "upfront."
"Obviously this is going to have a huge impact on next year's upfront," Loftus said, "but we don't know what."
The change came about as networks and advertisers tussled this year over whether to count shows that are played back on digital video recorders, or DVRs, as part of the viewership that advertisers get charged for.
Walt Disney Co.'s ABC network pressed particularly hard for those viewers to be counted, but yielded in the end after advertisers resisted.
The new data showing the viewership of commercials will also include those played back on a DVR device within seven days, something that is likely to favor highly rated shows that are more likely to be viewed on recorded playback such as American Idol.
Advertisers had resisted paying for viewership that was played back, partly because they're concerned that many people are skipping through the ads, and also because Nielsen's sample included fewer households with DVRs than the national average.
Nielsen spokesman Gary Holmes said Nielsen's sample is now about 7 percent made up of homes that have DVRs, vs. the national average of about 12 percent. Holmes says Nielsen expects its sample to catch up to the national average of DVR penetration by the end of the year.
Networks have long acknowledged that viewership dips during commercials. David Poltrack, the head of research for the CBS network, says that studies have found commercials consistently rate about 5 percent lower than regular programming.
What networks are eager to show, Poltrack says, is that not everyone actually skips through the ads when shows are being played back on a DVR, say if the remote is out of reach or the viewer is walking around the room or doesn't feel like skipping through the ads.
Poltrack says research suggests that somewhere between 40 and 50 percent of audiences do watch commercials on playback. Nielsen's Holmes said the company wasn't releasing any data on commercial viewership yet.
This could favor big shows like CBS's CSI, Poltrack says, which go up against other popular shows like Grey's Anatomy, which CSI viewers couldn't otherwise see since they're scheduled in the same time slot. "We know that a lot of people are going to watch one live and record the other one and watch it later," Poltrack says.
For advertisers, the new data represent a welcome means to measure the effectiveness of their ads, but also more work in figuring out how to agree on ways to value a new audience measure.
Advertisers are increasingly asking for clearer breakdowns of how many people see and act on ads, especially now that Internet advertising can provide great detail on how many people made a purchase after seeing and clicking through an ad online.
"This is a step in the right direction," said Brad Adgate, director of research for Horizon Media Inc., an advertising buying and planning agency. "This is one of the important things that buyers want."
Adgate added, however, that the data won't be a perfect solution: Not every ad will be rated, just the average ratings for all ads in a given time period will be measured.