DVR Ratings Prove Time-Sensitive

Television has always played around with the concept of time, what with different regional “standard” times, East and West Coast feeds, and transmission delays during “live” events. But the measurement of time-shifted viewing via digital video recorders (DVRs)—and ultimately video-on-demand and broadband servers—is further stretching the definitions of TV time and the meaning of “live” TV viewing. And deciding exactly what constitutes live viewing is creating tension between broadcasters and the ad industry.

“Advertisers and agencies see it differently,” acknowledges Shari Anne Brill, director of broadcast research at New York-based media shop Carat USA. “We are delivering viewers to our advertising messages, and we want to use the currency that best reflects that. At this point, we have no information about fast-forwarding or time-shifting time-sensitive messages, so for our clients to pay for that would be crazy.”

It’s a debate that has been raging since Nielsen announced plans last year to provide its new DVR-included ratings in three streams: one including only “live” viewing, another including live plus same day of playback on DVRs, and a third including live viewing and seven days of playback. Large agencies such as Magna Global USA and Mediaedge:cia immediately took the position in favor of live-only ratings as the basis for advertising guarantees, while the major broadcast networks argued that all of the measured playbacks should count.

At press time, Carat was expected to release a position paper on the subject, but, in general, both sides appear to be moderating their stances. In its analysis of ratings reports last week, Fox referred to “live-plus-same-day” playback as the Nielsen ratings “standard,” softening its earlier position that ad deals should be based on “live plus seven days.”

Meanwhile, the cable industry and some big ad shops like Carat have taken a wait-and-see approach. “It’s way too early to make any decisions where you come out on this at this point,” says Ira Sussman, director of research at the Cabletelevision Advertising Bureau, which has been advising the bureau’s members not to rush out with any marketplace positions on DVR ratings. “There is new data, but there really isn’t any new knowledge yet,” he adds, noting that relatively few households in Nielsen’s national sample have been included with DVRs.

Ultimately, Sussman says, it will come down to individual situations among advertisers, agencies and networks based on their unique needs and the types of programming they buy. So-called day-and-date advertisers that run time-sensitive commercials related to a specific offer may pay a premium, he believes, to guarantee live-only ratings, while less time-sensitive advertisers may opt for lower ad costs and include more time-shifted viewing.

The problem, says Carat’s Brill, is that “live isn’t really live” when it comes to Nielsen’s new data, because it still includes a significant number of shows recorded via videocassette recorders, which may never be played back.

That has been a problem ever since Nielsen first began including VCR recordings in its official ratings data decades ago. Initially, Nielsen said it was unable to measure VCR playback and would add the data back in a couple of years when it developed the capability. The ad industry mostly forgot about it, until recently. “A couple of years turned into a couple of decades, and pretty soon, we all forgot about it,” says Brill. “The shift to DVRs has reminded us again.”

Agencies have, in fact, jumped on the issue, calling on Nielsen to exclude VCR recordings from its “live” ratings data stream. “They should at least stop calling it live,” says Magna Global USA Director of Audience Analysis Steve Sternberg. “It’s not live.”

Agencies are doubtful that Nielsen will make those changes or that it will ever come up with a means of measuring VCR playback. But they are jumping on the issue at a time when the concept of TV time is being re­defined to ensure that advertisers know what they’re getting when they buy advertising.

But both the advertising and the TV industries are learning that, when it comes to the measurement of time-shifted viewing, time doesn’t stand still. Last week, for example, Nielsen disclosed that it had encountered a new generation of DVRs that automatically delays TV transmissions as much as five seconds without a viewer’s knowledge or consent. The DVRs are part of digital set-top devices being introduced by cable and satellite operators, not standalone systems like TiVo’s.

The delays are caused when the DVRs buffer downloads of programming on their hard drives. In early models, that took only a second or two, but even that was enough to cause Nielsen unexpected problems when the time used by its ratings meters to measure viewing in DVR households was out of sync with the “machine” time of DVRs.

Nielsen executives said they were surprised by those DVRs, which caused far longer delays than any they had encountered, and warned that they expect to come across other ones that delay playback even more.

Nielsen spokesman Jack Loftus says that’s what caused the ratings giant to delay its initial release of DVR ratings early this year and that its engineers are still sorting through the mess. One solution was to change the rule Nielsen uses to classify “live” viewing—extending it from anything played back within four seconds of being transmitted to eight seconds.

That seems OK with Madison Avenue. “I don’t have a problem with that, because they have no other choice,” says Magna’s Sternberg. “If the machine delays something, Nielsen has to accommodate that.”

Sternberg says the key is that Nielsen is measuring what viewers “perceive as live,” regardless of whether their DVR is delaying the actual broadcast without their knowledge. However, when Nielsen altered its edit rule for “live” to anything played back within eight seconds, it meant it also would include programming that viewers actively pause. That’s fine with Sternberg—as long as it’s only five seconds or thereabouts.

“But if it gets up to 15 seconds,” he says, “I think I might start to worry about it.

Meanwhile, Madison Avenue consumer gurus are beginning to wonder about another issue: a slight delay’s effect on how consumers view live television. “It shouldn’t matter if you’re watching TV that’s delayed five seconds or eight seconds after most other people are watching,” believes Paul Parton, strategic planner at New York ad agency The Brooklyn Brothers.

In fact, he says, it probably wouldn’t matter for most forms of programming. However, for shows intended to be viewed live—sports events, news, concerts—the temporal shift could have a profound impact on the way people think about “live” television programming.