Let's Really Look At DVR Viewing Before Deciding the Next Metric

MediaWorks Viewpoint: OMD's Dr. Sandra K. Eubank

Published on .

Most Popular
What are the facts about DVRs? Will they have a devastating impact on TV advertising, or are they just another tempest in a teapot?
Sandra K. Eubank
Sandra K. Eubank

In the last six months, I have read article after blog after commentary predicting the demise of TVadvertising due to DVRs. The fact that DVRs enable viewers to skip the commercials has been "fast forwarded" to its -- seemingly -- logical conclusion: Consumers will no longer watch any TV commercials.

Well-known industry blogger Douglas Rushkoff wrote in an April 13th post: "The emergence of cable TV, pay TV, VCRs and now DVRs, has introduced viewers to television without advertising. Whether we fast forward, delete or simply pay for the ad-free 'Sopranos', we're watching TV on our own terms. And we don't want to go back."

Pretty scary stuff if you're in the advertising business.

According to a survey of 133 national advertisers conducted by the Association of National Advertisers and Forrester Research, 70% of marketers believe that digital video recorders (DVRs) and video-on-demand (VOD) would "reduce or destroy" the effectiveness of 30-second TV spots.

Wait a minute.
Of course I agree that new technology enables change and that it may even drive change eventually. But I believe we should look at hard facts collected from consumers to really understand the impact these new technologies will have on TV viewing. Consumers can surprise even the most savvy marketer. One thing my career in marketing has taught me is that it's pretty tough to change consumer behavior. Consumers are like cats: They do what they want to do, not what we want them to do. Our job is to understand what they want and make a business out of it. Survey data is how we figure out consumers.

What does hard data tell us about DVRs and their impact on commercial viewing? The first study to be broadly published was conducted by Millward Brown, a well-respected advertising-research firm, and paid for by the largest TV networks. The study showed that advertising recall was the same among households that owned DVRs as it was among those that didn't. It also estimated that 60% of viewing in homes with DVRs is live rather than prerecorded.

In research, it sometimes happens that the results do not agree with widely held expectations. It might be a fluke, or it might be that researchers' expectations were wrong. This dichotomy calls for more research. The idea is that "fluke" results won't be replicated, so if several studies find the same result, it must be "truth" even though unexpected.

Since the Millward Brown study was published, several other studies have been conducted and produced similar results. OMD conducted a study that confirmed the finding of similar levels of ad recall among households with DVRs vs. those without, and also explored the critical reasons why. In the OMD study, households with DVRs were asked whether they viewed a specific program live or recorded it on the DVR and replayed it later. For two specific programs evaluated, 40% of the DVR-owning viewers (math check: about 5% of the total population) recorded the program and played it back; the other 60% watched it live.

Not surprisingly, commercial recall was significantly lower among DVR households who recorded the program than among non-DVR households. In support of this recall finding, a large proportion of DVR households reported that they regularly skip commercials when they play back prerecorded programs. However, very surprisingly, commercial recall was significantly higher among DVR households that watched live than among non-DVR households.

Now, admittedly, we still don't know exactly why recall is comparable among DVR owners and non-owners. The OMD study traced this result to very different behavior among live vs. prerecorded viewers. OMD hypothesizes that live viewers are more engaged in programs they "choose" to watch live. Choosing to watch now when they could just as easily record may mean they really do pay more attention to what's on the screen. Other people may suggest alternative hypotheses.

However we explain this result, the fact remains that there is a growing body of evidence that the impact of DVR ownership may not be as devastating as has been feared. While one study might be a fluke; several studies with similar results are not. Marketers should use facts like these to try to understand how DVRs really are going to impact commercial viewing. Once we understand consumers' behavior, we can figure out how to make the new behavior work for us.

I believe we should continue to assess the impact of DVRs over the next few years. More studies likely will confirm that recall declines only when recording and playing back, and will help us better explain why. Moreover, DVR owners' behavior should be tracked, as it could change over time. The penetration of DVRs is projected to increase rapidly over the next few years, and new owners may well behave differently than today's early adopters. If new owners record more programs than today's owners, DVRs could have a bigger impact in the future than they do today. Furthermore, today's owners may change their behavior over time as they get more comfortable with the device.

At the same time, smart marketers may want to hedge their bets by exploring new and creative ways to make sure their ads get noticed. For example, one marketer developed an ad with an imbedded coupon to try to get fast-forwarders to slow down and pay attention. Other ads have been developed with still images, such as text or logos that are visible when fast-forwarding or viewing. Goodyear developed an animated version of its blimp that floated across the screen during the Olympics. And let's not forget product placements and program sponsorships, which are booming. As marketers explore these options, they should test the response to see if they are working or not.

Finally, there is the hot issue of how to measure viewing in the world of DVRs. As you probably know, Nielsen originally excluded DVR owners from its sample, assuming they would skip all the ads. In doing so, they unwittingly prevented themselves (and us) from seeing some of the very data we so desperately want now in order to understand viewing. Having realized this, Nielsen is racing to build the penetration of DVR viewers in its panel up to representative levels by January 2007.

This change means that, as of January, Nielsen's Live ratings will reflect all live viewing in a sample that includes a representative number of DVR households -- which is ideal from a research perspective. Nielsen is also offering a different type of data called "live plus same day" and even "live plus seven day." This new data set adds viewers to a program's rating if they prerecorded and played back a program within the specified time period.

OOOPS! Rewind!
I don't think marketers should adopt this new data stream as the gold standard yet. While existing research shows that DVR households have comparable ad recall to non-owners (so they should be in the sample after all), those same studies also show that recall does decline when viewers record and play back. To me, this clearly means advertisers should pay for DVR live viewers but not for any "plus" viewing.

This thinking seems to have prevailed in the current TV upfront, in which networks wanted to do deals based on "live-plus" ratings, and advertisers refused to accept this new data standard. This could become an even bigger issue in the future, if and when people start recording more. I imagine that as penetration of DVR owners goes up, recording is bound to increase. We'll (of course!) need more research to figure that out, but until data show incremental benefits from delayed viewing, advertisers should continue to hold their ground.

Dr. Sandra K. Eubank is OMD's U.S. director-communication insights and research. Dr. Eubank has more than 20 years of market-research experience, and joined OMD in 2005 from Ipsos-Novaction, North America.
In this article: