Sizing Up PVRs

The consensus among advertising research executives is that digital video recorders (DVRs) could revolutionize the methods of collecting TV-viewership data, refining the industry’s understanding about who’s watching what program and what commercial when and for how long.

With 70 percent or more of TV-viewing households relying on the boxes to receive their programming, and more and more signing on to manipulate programming using DVRs and other time-shifting devices, buyers could get viewing data that is close to a census, rather than relying on the ratings based on much smaller samples from Nielsen Media Research.

“There’s no question we see the future of television measurement clearly being focused—not on the sampling method, but on data-oriented measurement,” says Tim Hanlon, svp of Starcom Worldwide.

Hanlon notes that audience measurement “is already behind, given where television and technology are today. DVR, on-demand, all these new wrinkles in TV viewing clearly suggest a more sensitive and comprehensive source of measurement both for programmers and advertisers.”

So far, Nielsen, owned by Adweek Magazines parent VNU, has not been able to include DVR households in its sample, since only an estimated 3 percent of TV households have the devices. But with the rapid growth of DVR sales, that is projected to change rapidly, thanks to an aggressive push by cable operators. By the time Nielsen’s plan to deploy a more advanced meter—which has been in development for several years—actually occurs, DVRs could be in one out of five households.

To its credit, Nielsen has made moves to keep up with TV technology. Last March, the company said it plans to include time-shifted viewing from DVRs into its ratings beginning next July. A month earlier Nielsen said it was teaming with TiVo to build a panel of 10,000 TiVo homes to produce data about time-shifting and other video-recorder uses. And it was reported last month—with few details emerging—that Nielsen was in early talks with major cable operators like Time Warner Cable, Comcast and Cox Communications about creating an audience-measurement system based on digital-cable boxes.

Selling data from set-top boxes is not new. Highly “granular” data culled through set-top boxes has been available—and sold to marketers, researchers and network executives—for years, notes Kimber Sterling, director of advertising and research sales at TiVo. Sterling, a former media planner on the Dell account with Goldberg Moser O’Neill and founder of the media department at Butler, Shine & Stern, recalls: “The feedback we would always hear from clients is how do I know my spending is justified? Even if the data is reliable—which it’s not because it’s within the margin of error—what does it mean in terms of sales, intent to purchase? When you blend our ability to measure with the interactive ad platform we are creating, you see this technology taking off.” TiVo counts as clients BMW, Charles Schwab and Walt Disney Co.

Still, aggregating data out of millions of set-top boxes may not be as easy as it sounds. Even Hanlon—a longtime critic of Nielsen’s methodology who refers to the company’s “chokehold on the TV measurement of today”—concedes, like others, that the sheer expense bars the creation of other forms of measurement. He also gives credit to Nielsen for starting measurement on media such as out-of-home and gaming and for its TiVo relationship. Hanlon looks for technology to present “a better story than what we see today, to get a much cleaner and closer view of what people are really watching.”

Then there’s the privacy issue. “You have to be careful about how detailed such data can get,” says Hanlon. “But it is fair to say we can get much more granular about viewing habits without violating privacy.” He points out that the ability to measure in aggregate by ZIP code, for example, is a far more detailed measurement than a simple sampling of males ages 18 to 49. “The sad reality is that marketers, when wanting to use TV, always had to dumb down into media metrics that are broad and may hit the target audience, but which may have excess audience.”

David Poltrak, evp/research and planning at CBS Television, says that the small number and nonrepresentative demographics of DVR households limit the research potential. DVR owners, he explains, “are early adopters. We know from our research that TiVo owners are not even representative of DVR owners.” He says that data from those sources “has to go through the same scrutiny as a Nielsen or an Arbitron.” Still, he adds, the technology “does represent a lot of new potential in ways we can measure the medium.”

Poltrak sees the future as “multi-operational, where you measure the same thing in different ways.” He says, “We need to see [not only] the participation rate increase but also the census-type data for the percentage of the overall population that will allow us to do modeling and more sophisticated research. More and more we will rely on qualitative research to study the environments of programs and which messages work in those programs. This has been done on a proprietary basis but will have to be done on a more ongoing basis in syndicated form. There’s a great opportunity going forward.”

The question remains when. “At the end of the day, we all work for the advertisers,” says research guru Tony Jarvis, late of MediaCom, who sits on the board of the Advertising Research Foundation and is a member of the media research committee of the American Association of Advertising Agencies. “If things don’t get better a lot faster [regarding audience measurement], America will remain so far behind the rest of the world. This is the country where advertisers spend the largest amount of dollars in TV. We really need something a lot more sophisticated.”

Tony Case is a contributing writer to Mediaweek.