Commentary

Who Really Skips Ads Via DVRs?

by Steve Sternberg , Featured Columnist, June 6, 2011

I went to the movies recently with my wife, my 12-year-old son, and my mother-in-law, who is in her early 70s.

Prior to the movie, the theater aired several commercials, none of which my wife or I had seen before. My son, however, seemed quite familiar with them, and commented that the last ad was one of his favorites. My mother-in-law also seemed to recognize the commercials.

When I said I never saw any of those commercials, my son said "Of course not, Dad, you never watch commercials. You fast-forward through everything. I love watching commercials."

It struck me that my son, as usual, is correct. We DVR and time-shift much of our prime-time viewing. We seldom watch commercials anymore. Maybe only during the 20% or so of the time we are watching something live - usually sports, news, award shows, or off-network series on cable or syndication (such as "NCIS" or "Everybody Loves Raymond").

DVR penetration among adults 18-49 is about 45%. Nationally, 27% of adult 18-49 prime-time viewing of the broadcast networks and 11% of viewing of ad-supported cable networks season-to-date is through DVR playback.

Simple math reveals that 60% of adult 18-49 broadcast viewing and 24% of ad-supported cable viewing in DVR homes is time-shifted. Cable is so much lower because of the high percentage of off-network series and movies, as well as reality shows. If we just looked at original scripted dramas on cable, it would actually exceed the 60% level of the broadcast average.

Is my little convenience sample even remotely accurate? In DVR homes, are the under-18 and over-70 viewers the ones watching most of the commercials on TV?

This is one of the major quandaries facing our industry right now.

How do you deal with a medium where the programming is more popular than ever, but commercial avoidance is also greater than ever -- and growing? Where a high degree of DVR playback might be enough to get a moderately performing show renewed, even though most of those additional viewers are not watching the commercials?

There have been a number of studies over the past few years indicating that commercials on TV are more effective than commercials online. The findings are usually carefully worded to compare people who actually were exposed to the commercials on different media. There have also been one or two studies showing just the opposite - although sample sizes were generally too small to have confidence in the conclusions.

The Council for Research Excellence's groundbreaking Video Consumer Mapping Study, which I was privileged to work on, FOX's recent work with Innerscope (and NBC's previous work with the same company), plus several other studies have indicated that people who watch commercials on television are more engaged than those who watch commercials online, and those who actually watch the ads during DVR playback are at least as attentive as those watching live TV.

But here's the rub. Roughly 80% of time-shifted viewing via DVRs includes skipping the commercials. How can I say this? It's just as valid as the 60% often quoted by the Big Four broadcast networks and Nielsen (although they generally say 40% watch the commercials, which makes it sound better). Nielsen simply cannot measure fast-forwarding, or chooses not to (I'm not sure which). C3 does not eliminate fast-forwarding as many think it does. Fundamental flaws in the way Nielsen calculates both minute ratings and C3 result in a substantial amount of fast-forwarding being included. My guess is that the oft-cited 60% commercial skipping via DVRs is really closer to 80%, perhaps even higher. I know it is in my household. How about yours?

When Nielsen first started reporting DVR data, they made a point of telling people that you couldn't just subtract Live viewing from Live + 7 to get the time-shifting viewing data. They quickly stopped saying that. I wonder why. It couldn't be because their biggest clients jumped on the calculation to demonstrate that DVRs were not the severe threat to commercial viewing that advertisers feared they were.

The real question is why neither Nielsen nor the organizations committed to doing excellent research are trying to find out how much fast-forwarding via DVRs is actually taking place (or at least how much Nielsen is missing). It seems to me to be far more important than defining set-top terminology or measuring cross-platform viewing. Both are important, but pale in comparison to understanding the dynamics of time shifting.

It's actually quite simple to do this type of analysis.

Take 10 or 20 of the industry's leading television researchers, hook up our television sets to a peoplemeter. Let us record all of our viewing for one night and play back the programs over the next three days. We will write down our second-by-second viewing, including any fast-forwarding, commercial viewing, pausing, etc. Then calculate the C3 rating based on our actual viewing behavior and compare it to what the Nielsen meter calculates as the average C3 rating. Media researchers are accustomed to doing extremely detailed analyses. Keeping track of one night's worth of viewing should not present any problem. This would provide a realistic look at how much fast-forwarding activity is not being accounted for in Nielsen's metrics. <

metrics, television, tv

16 comments about "Who Really Skips Ads Via DVRs?".

Check to receive email when comments are posted.

Art Salisch from SMRC, June 6, 2011 at 4:35 p.m.
Great Idea, Steve-
Count me in...
Reply

Douglas Ferguson from College of Charleston, June 6, 2011 at 4:38 p.m.

I thought TiVo collected the very information you seek. Nielsen is so 20th century. Don't use a convenience sample. Ask any of the people using STBs (liek TiVo) to report their data on ad-skipping. They can measure their respective populations without resorting to a sample.

Chuck Lantz from 2007ac.com, 2017ac.com network, June 6, 2011 at 4:53 p.m.

Rather than writing down your viewing record, couldn't the actual elapsed playback time of each program be recorded automatically? That would indicate if the entire program, including commercials, was actually viewed, or if the viewer fast-forwarded through the ads.

In other words, if the total elapsed time recorded on the DVR was 22 minutes when a 30 minute episode was being watched, that would most likely mean that 8 minutes of ads were skipped.

Chuck Lantz from 2007ac.com, 2017ac.com network, June 6, 2011 at 4:55 p.m.

Sorry, I should have said "if the actual time viewed on the DVR", and not "recorded on the DVR"

John Driscoll from School Family Media, June 6, 2011 at 5:42 p.m.

This is purely anecdotal but I have two teenage girls and a tween boy in my house and they use our DVR all the time and watch their favorite shows in between homework and other activities and they never watch the commercials. I think like alot of technology these days the younger generation takes to it, and as such advertisers trying to reach that audience will need to take a hard look at the numbers and their ROI.

Jason Krebs from Tenor/Google, June 6, 2011 at 5:53 p.m.

Most people stopped watching commercials when the remote control was invented. The industry didn't care then, they don't care now.

Tim Orr from Barnett Orr Marketing Group, Inc., June 6, 2011 at 8:01 p.m.

I frequently find that when I am time-shifting, I forget that I am doing so, and let the commercials run. Probably happens to a lot of people.

The notion that nobody who time-shifts watches commercials is an argument from logic, not from data. It's the same kind of thing that led pre-Galileans to believe that a 10-pound ball would fall 10 times faster than a 1-pound ball. Stands to reason, but is untrue.

Doug Garnett from Protonik, LLC, June 7, 2011 at 5:22 p.m.

The mythology of the DVR as "ad killer" took hold early. And, feeds off the belief that humans are simple.

"I never watch an advertisement" is a simple believe that the ad business took to. It's also false. We all notice ads far more than we know and we watch them (even time shifted) far more than most would admit.

SO, I'm surprised to find on Media Post a column that hasn't even read what's easily found on Media Post: DVR's not only haven't decreased measured TV ad effectiveness, they are measured to have improved it.

Here's my blog post that discusses the Media Post article with Nielsen's research. http://dsgarnett.wordpress.com/2011/03/02/more-research-shows-dvrs-e-g-tivo-increase-advertising-impact/

And now, let's admit that we're all complicated and the consumer interacts with advertising in complicated ways that defy the ad businesses attempts to make everything black and white.

In the meantime, I have to go watch John Steward time shifted and I sure hope there's an ad for iCloud because I'd love to see it.

John Grono from GAP Research, June 8, 2011 at 1:15 a.m.

Dear, oh dear. Supposition being passed off as fact. For starters, a DVR knows what is tuned - not what is viewed. It 'knows' that at least one person is there if it is live tuning, but it has no idea whether it is more than one, age, gender etc. It also has no idea if it is a recording. And much of what is recorded is not viewed.

Here in Australia, from our TV panels we know every second of content on around 10,000 TV sets in over 5,000 homes with over 13,000 people. We know whether that content is 'live', 'as live', or 'time-shifted' (up to 7 days cut-off). We also know for shifted viewing whether it is in real-time or trick-play (FFWD, RWD, Pause etc). ANYTHING that is played back that is NOT in real time is not credited. So, if someone doesnn't watch the ad-break ... no contribution to the rating. If they do watch the ad-break ... they do contribute to the rating. We know if they payed back within 1 hour, 1 day or 1 week (and lots more time intervals in-between).

The thing is you would surprised just how many people do NOT skip the ads. Sure they may get upo and leave the room or be distracted (as happens with any and every medium), but our compliance checks indicate that in 90+% of all instances the 'presence in the room' matches the validation call. Some call this antiquated and 'last-century' - I call it responsible high quality research.

George McLam, June 8, 2011 at 1:37 a.m.

Just because someone is watching live does not mean they are watching the commercials. Just because they don't "skip" with their recorder also does not mean they are watching the commercials. Commercial avoidance (going to the kitchen/etc) has been around as long as commercials.

If there was not so much advertising I might have an interest in it. It floods my mail box, my phone, radio waves, bill boards and even bills I am sent from companies I am already doing business with. If advertisers really want their work to be more EFFECTIVE, they need to stop diluting it with so much of it.

Steve Sternberg from The Sternberg Report, June 8, 2011 at 1:06 p.m.

I think some people misread my article, or missed the point. I was merely commenting on the need for better research regarding DVR activity. Nielsen clearly understates fast forwarding, so the 60% skipping they report is too low. Very few seem to want to do the research needed, because they are afraid to find out the answer. It frankly will not benefit the sellers or the buyers to know the full extent of commercial avoidance via DVRs. It will help advertisers, however. And just because you can't really measure live viewing commercial avoidance on an ongoing basis, doesn't mean you shouldn't measure fast-forwarding. In reality, people do not leave the room that often or switch channels frequently when watching TV, so I think it is clear that there is significantly more commercial avaoidance in delayed viewing than live viewing.

Doug Garnett from Protonik, LLC, June 8, 2011 at 6:42 p.m.

Steve -

Perhaps we need additional data. But since the advent of the DVR hasn't decreased the effectiveness of TV ads, I really don't see what we'd learn that would make any difference.

And I think it could be very mis-leading. I've worked with 1/2 hour infomercials. A lot of producers/researchers love dial groups because they give us "so much information".

In truth, "information" is exactly what they give us - not anything useful. What we've learned is that the issues that affect an infomercial's effectiveness won't show up on dials.

Just so, the important information about an advertisement's impact seems unlikely to show up on the DVR rewind statistics.

So, why do we want them? Not clear to me.

Steve Sternberg from The Sternberg Report, June 9, 2011 at 2:03 p.m.

Doug - not sure what you mean by saying the advent of DVRs has not decreased effectiveness of TV ads. Yes, TV ads that ware seen, are just as effective as ever. But you relaly don't understand why we want to know how much fast-forwarding is taking place? Because commercial avoidance does decrease ad effectiveness. You don't need to prove the sky is blue. This is as self-evident a concept as there is. An ad can only impact people who have seen it.

Doug Garnett from Protonik, LLC, June 9, 2011 at 11:36 p.m.

Steve - Of course people need to see ads to be influenced by them.

My point is that the solid research shows that the DVR didn't decrease ad effectiveness meaning it didn't overall change what was going on for consumers.

We also know, therefore that FFing may not really reveal much to us (dial research in our business proves quite misleading).

The reason people pay attention to ads in a commercially important way is that the ads tell them something meaningful. But techniques like dial and FF analysis focus us on microscopic issues that are rarely connected to success.

This IS theory in the case of FF research (although it's shown true in dial research which seeks to analyze the same thing). But this begins to explain why, given TVs superb continued and growing effectiveness, I don't share your enthusiasm for spending the time you suggest.

John Grono from GAP Research, June 10, 2011 at 3:10 a.m.

Point well made Doug. Indeed there are cases where ads in FF are MORE effective. Back in the '70s or '80s when VCRs still roamed the earth we had a shock absorber company in Adelaide make a 30-second ad. I must say it was far from ground-breaking creative, but up in the top corner was a little animated graphic showing how a shock-absorber worked.

So how was this effective? Because when the ad was in a programme that you were fast-fowarding through, that little shock absorber was going sixty-to-the-dozen and you simply couldn't help but notice it and chuckle.

Darrin Stephens from McMann & Tate, June 16, 2011 at 2:15 p.m.

"Take 10 or 20 of the industry's leading television researchers, hook up our television sets to a peoplemeter...This would provide a realistic look at how much fast-forwarding activity is not being accounted for in Nielsen's metrics."

Actually, it would only provide a realistic view of how 10 or 20 TV research geeks fast-forward. That's not real research and you know it.