Commentary

TiVO Doesn't Impact Advertising Effectiveness?

by Joe Marchese , May 18, 2010

I have a DVR. I record every program I view regularly, so I can watch uninterrupted by commercials. If I want to see a show the same night it is on, I wait until at least 20 minutes of the program have passed before starting to watch, so that I can view it without interruption. This has been the case for a couple years for me, and most of my peers. So you can imagine my surprise when a story hit my news feed titled "Duke study: TiVo doesn't hurt TV advertising" My first thought was, "Maybe I'm the only one who uses a DVR to skip advertising?" But once I came to my senses, I began to think about other explanations for the study's strange results, and they aren't pretty.

First of all, the title of the article is a little misleading. Duke Professor Carl Mela is quoted as saying that TiVo usages doesn't seem to impact people's "shopping patterns," which is very different from saying that having a TiVo doesn't hurt advertising's impact. The distinction is very important (more on this in a second). I know that one person does not make up a sample, but I know for a fact that because I have a DVR, over the past couple of years I have used my DVR to avoid consuming a massive number of commercials. So even if I am the only person on the planet who does this, there is at least some impact ;-) So why didn't the study show more impact on people's shopping behavior?

Professor Mela cites some potential examples, ranging from the realistic -- that people without DVRs who wish to avoid commercials do so simply by changing the channel or leaving the room -- to the seemingly ludicrous: that when people use the fast-forward feature, they have to pay close attention to the commercials whizzing by so they don't miss the start of their program (and this does what? What's really funny is that I have heard this one before).

While I buy that people who don't want to consume advertising but don't have a DVR device can avoid it in other ways, or simply turn their brains "off" during advertising, I find it hard to swallow that those without DVRs can avoid commercials at the same rate and efficiency as those with DVRs. This study may simply show that television advertising today has less impact than originally thought. If people in the homes with DVRs did skip any more advertising then those without, AND the study showed no relative difference in shopping patterns, it sounds more like an indictment of the current interruptive model of television advertising than an acquittal of TiVo's role in murdering television's revenue model.

But there's another option that I lean toward, which could explain why owning a DVR shows no impact on purchasing behavior: there is no viable substitute to television advertising. This is why I said earlier that it is a very important distinction to note that the study is not isolating advertising effectiveness, but rather "shopping patterns." Sure, maybe people who own a DVRs are watching less of Brand X's television advertising, but as long as Brand X's competitors don't find a better way to reach those people with their message efficiently, television is still the best way to reach the broadest audience in the most engaging manner, so Brand X holds market share. Let's just not kid ourselves that the current model of television advertising isn't less effective for those with DVRs, or that the current, interruptive model of advertising isn't in serious peril.

What this study says to me is that there is opportunity. The opportunity is for content, cable and Internet companies to create new, highly engaging, mass-reach vehicles for advertisers. Because once viable alternatives for advertising are available, those marketers who are early adopters will be able to gain market share on their competitors -- if their products and advertising messaging are good, of course.

Side note. While I admit I have watched more ads on Hulu in the past year than I have watched on my television thanks to my DVR, is adopting television's interruptive model of advertising and throwing it on the Web really the best we can do? I would offer that there are other solutions that will lead to people WILLINGLY spending more time with adverting on their own schedule. Stay tuned.

19 comments about "TiVO Doesn't Impact Advertising Effectiveness?".

Check to receive email when comments are posted.

Mike Bloxham from Magid, May 18, 2010 at 11:37 a.m.
Excellent post Joe. Salvaged some really great insights from what would otherwise probably have been consigned to obscurity as a result of lazy reporting and presentation.
I agree that in the absence of other truly breakthrough communication from competitors then the effect of DVRs (whatever it really is) will be the same for all. But when a brand breaks the mold consistently, then it will be intriguing to see what case studies emerge and how the brands they reference fare against the rest.
Reply

James McDonald from SuperString Theory, May 18, 2010 at 11:41 a.m.

Joe, you are correct in your habits and my family does uses a DVR in exactly the same way...as do most people I know who have one. A very prominent television research executive walked into my office in 1999 and pointed to an article on a new device called TiVo and said "It's over". He was referring to the idea that once consumer are empowered to "opt out" of the interruptive ad process then the familiar model would fall. What we both discovered is that advertisers need to believe in the "myth" as much as the programmers. The ratings agencies perpetuate the myth, the nets spin the impact and the whole thing keeps moving through another upfront with more money being spent...am i nuts or is this whole thing set for a catastrophic fall. e.g. among the most expensive spot buys...American Idol, most skipped ads by DVR's...American Idol. You do the math.

Douglas Ferguson from College of Charleston, May 18, 2010 at 12:12 p.m.

Amen, brother. I share your incredulity at these conveniently-appearing studies that essentially say, "Pay no attention to the man behind the curtain." TiVo is changing the way people watch TV. Maybe the imminent collapse of the broadcast model is 7 years away instead of 2 years away, but it's getting harder to root for the status quo.

Warren Lee from SEO-CUBED.COM, May 18, 2010 at 12:20 p.m.

My wife also likes to skip through the commercials, which is kind of annoying because I actually like them! But I am training myself to avoid clicking on annoying ads like ad mobs ads on my iPhone. I guess its kind of the same thing as ignoring commercials on TV, except for mobile.
I think understanding why and how people put up barriers to advertising is very interesting. Ive noticed that some ads these days are almost impossible to avoid, such as embedded ads in mobile applications. Also I wonder that considering there are more ways to avoid advertising when you don't want it, such as while watching an engaging tv show, then do people turn to other advertising sources to make up for this. For example I noticed that since we got the DVR, and I have less exposure to tv ads, I then make up for this by reading magazines and enjoying ads there instead. But maybe that's just me? I actually like advertising, especially when its delivered on my own terms.... oh yeah and targeted ads on open social... don't get me started!

Jonathan Hutter from Northern Light Health, May 18, 2010 at 1:46 p.m.

There is one fatal flaw in your argument, and it comes in the first paragraph. "This has been the case for me...and most of my peers." This point of view is why Twitter is not as big as people in media think, and why Hulu has not replaced network television. On that last point, Joe is certainly correct. TV still hasn't been replaced by other forms of media that do its job better.

However, there are forms of media that people do seek out for the advertising. Two examples - NASCAR (how boring is a naked car with no ads anyway) and an old, and now dying example, the Sunday newspaper. I know people who read the sales flyers first and the articles second.

Jim Garrity from BridgeTwoWorlds, LLC, May 18, 2010 at 1:49 p.m.

Joe,

I couldn't agree with you more. My family consumes tv the same way you do. Duke is a an institution with an impeccable reputation. I find it hard to believe they published such a flawed piece of research. Their assumptions imply that every tv commercial has as its primary objective to drive traffic to retail.

Greg Alvarez from iMeil, May 18, 2010 at 2:39 p.m.

Shouldn't the question must be "Do I have to see ads in a TV service I pay for?"?

IMHO, anyone pay to see programs, not advertising.

Doug Frechtling from George Washington University, May 18, 2010 at 3:25 p.m.

Thanks for confirming how my wife and I have used TiVo for 5 years or more. My wife gets impatient if I don't FF through commercial messages fast enough.

In addition to your ideas, I expect we will see a stronger push for product placements in the top drama/adventure shows. But then, I have product placement in my own kitchen and can't tell you more than half of the brands there. But when I see a product in a show that I buy, I notice it. Perhaps TV is all about confirming brand preferences rather than building new ones.

Anne Peterson from Idaho Public Televsion, May 18, 2010 at 5:59 p.m.

I haven't had television reception, cable or satellite for more than 10 years, but when I did, and I recorded on tape, we either left the room or fast forwarded through the ads -- and that goes back to the latter half of the 1980s. Nothing new here; go to the next curtain.

George McLam, May 18, 2010 at 6:21 p.m.

Joe, you are not alone. When I read the original article you speak of I thought it must have been planted by someone in the (advertising) industry to make advertisers 'feel better' about the TV time they buy. There have been numerous articles on this topic over the years, and I don't even agree with those that state less than 90% of DVR owners skip commercials.

I will say there is a huge difference between "fast forwarding" through the commercials, and "skipping" over them. When FFing, you can see them whiz by.

My company actually has patented technology that identifies commercials in a TV broadcast so they can be (even automatically) skipped. This technology was implemented in a large number of VCRs up until about 2001. When used on a DVR, the commercials are not even seen at a higher speed.

I like the idea of product placement or companies literally underwriting shows; but don't insult me by also running commercials. Hour-long programs should be closer to 58 minutes instead of the current ~42.

Fraser E from Opinions expressed herein are solely my own, May 18, 2010 at 7:16 p.m.

I see. So even among resarchers, the only explanation for results that challenge our ever-so-enlightened worldview, derived from our own imperfect perceptions of our own desired behavior, projected on the rest of the population, is that the research must be flawed? *I* skip ads, my wife skips ads, therefore ad recall is impossible among DVR users! Anybody dare say otherwise in the name of research? Circle the wagons and light the torches!

Joe Marchese, May 18, 2010 at 8:08 p.m.

@fraser well, to be fair the study was sponsored by TiVo... oh and I didn't question the results, just the conclusions. It would seem to be a matter of fact that if only 10 people in world use a DVR to skip ads, then ads are at least that much (however small it may be) less effective. Right? Am I missing something else? So with that, I can find more than ten people by walking down the street that use DVR in just such a manner. Am I just running into those people?

Also, the argument that not everyone has a DVR is fine, but this study compared DVR owners with non-DVR owners. so % of the population with DVR should not be a consideration.

Carl Mela, May 19, 2010 at 8:40 a.m.

As one of the authors of this study, I would like to thank Joe for posting the work and his thoughtful comments. I thought I would offer a couple of clarifications:

1) The paper (available on my website - http://www.duke.edu/~mela) does not address whether TiVo hurts advertising. It addresses whether DVRs affect sales, specifically in the context of consumer packaged goods. We tracked about 10,000 households over three years, some of which were given a DVR and some of which were not. Contrasting the purchase behaviors of the two groups we found no evidence that DVR ownership affected the sales of advertised or private label brands. Given the sample size, the statistical variation around this null effect is quite small and the result is statistically reliable.

2) Several astute readers noted TiVo sponsored the research. To be clear, neither Duke nor I received any funds nor would we accept them. Nor was the study funded by the advertising industry. Rather, the experiment was sponsored by the advertisers themselves who wanted to know if DVRs were making their ads less effective. Like many readers and the companies who sponsored this work, we believed DVRs had a deleterious effect on sales and set about to measure precisely how large it was. As we did not believe the null result at first, we spent over 3 years trying to find an effect. The paper reports a number of these alternative analyses but there were many more we did not. Personally, I would be happy if other researchers find otherwise; but as a researcher my goal is to let the data speak louder than I.

3) The study does not indicate less than 90% of owners skip commercials as some have suggested (actually, all owners use the fast forward feature). Rather, the data indicate about 95% of viewing occasions (closer to 90% of viewing time) is live and thus can not be skipped. When views are from recorded content, then the likelihood of skipping an ad is 2/3. As indicated in the paper, these statistics are consistent with data collected by other researchers using different DVRs and households. Given 10% of views are recorded and 2/3 of ads are skipped in recorded mode, the average percentage of ads that are skipped is 10%*2/3, or about 6-7%. This is far smaller than I would have surmised before these studies and I believe it is the most plausible explanation for the lack of a notable DVR effect on sales.

4) To be clear, the "ludicrous" explanation of noticing ads that are whizzed by is not one that can be validated with our data. Instead our paper (on page 31) cites 5 separate studies (both experimental and field studies) that have documented this fact. These studies suggest awareness and recall of ads are higher in fast forward mode than no exposure mode.

5) Fortunately, we had access to information for nine advertising campaigns. For the households who own DVRs, we could therefore track advertising exposure for those nine campaigns to sales. We find the advertising effect on sales to be consistent with extensive previous literature on ads. This result provides weak evidence against the notion that advertising simply does not work at all hence DVRs can have no effect.

6) Most important, I fundamentally agree with Joe's thesis there has got to be a better way. Hulu holding viewers captive to ads is not a positive step forward. DVRs coupled with sophisticated analytics can facilitate better matching of advertising content to viewer preferences at the household level, making at least some exposures more relevant to viewers and obviating the need for advertisers to buy impressions that are of little value. With targeting, I suspect everyone can win and I intend to follow this line of inquiry next.

Mike Einstein from the Brothers Einstein, May 19, 2010 at 9:46 a.m.

If DVRs had been in existence 30 years ago, we wouldn't be having this discussion today, because the TV model would already be dead and buried.

The bottom line is that we keep pushing a product (ads) on folks who not only don't want it, but who are willing to pay extra to avoid it.

If it weren't for those big red cups on the table, I wouldn't have any idea who sponsors American Idol.

Joe Marchese, May 19, 2010 at 11:48 a.m.

@Carl thanks for getting involved in the discussion. You points are very valid and appreciated. I was a little tongue in cheek when pointing out that the study was sponsored by TiVo, but as I said, I have no doubts in the validity of the research. I am instead offering an alternative assessment of the "why".

Now, back to the "ludicrous" - Can you please post the link to one the studies that show that people who fast-forward ads are still measurably impacted by those ads? I need a column for next week ;-)

Carl Mela, May 19, 2010 at 3:04 p.m.

Joe -- happy to help. Sadly, most of the material does not have links because the academic journals are still using a "stone age" media model; however, Google Scholar sometimes has them. The fast forward papers are:

[1] Brasel, S. Adam and James Gips, “Breaking Through Fast-forwarding: Brand Information and Visual Attention,” Journal of Marketing, 72, 4 (November), 31-48.

[2] Goode, Alister (2007), “Duckfoot: 'What Happens at x30 Fast Forward,” in PVRs and Advertising Exposure: LBS Conference Report and Update, Sarah Pearson and Patrick Barwise, Editors: London Business School.

[3] Mandese, Joe (2004), “Equitable's No Longer Questionable: Data Reveals Nets Position Some Advertisers Better than Others, Media Daily News, (September 17).

[4] du Plessis, Erik (2007), “DVRs, Fast Forwarding and Advertising Attention,, Admap, September, 48-51.

[5] Siefert, Caleb, Janet Gallent, Devra Jacobs, Brian Levine, Horst Stipp and Carl Marci (2008), International Journal of Advertising, 27 (3), 425-446.

Peter Kloprogge from Pointlogic USA, May 19, 2010 at 5:59 p.m.

Thanks Joe, really like your article and as is apparent with all the comments it does make address a subject close to many.

I've seen some studies from Europe that show that people overestimate the amount of fast-fowarding that they do as most viewing is still live. The amount of fast-forwarding first increases when people buy a DVR but then falls back to lower levels.

But another perspective is the following. Most sales models that I've seen contribute a large part of the effect to television advertising. And let's say, for argument sake, that in the end only 20% of the purchased GRPs actually deliver exposures. The sales models than contribute the effect to the 100% of GRPs which means that the actual delivered exposures are 5x as effective as we think. I believe this to be true without fully understanding why this is happening. Perhaps skipping ads puts the ads that we do see in a less cluttered environment making them more effective. Perhaps we tend to skip less if we do see the first five seconds of an ad and these seconds are relevant. Perhaps there is something else going on.

But what I believe this does show is that television's potential is probably as big as it ever was. If exposures are 5x as effective (and remember this number is for argument sake) than we currently measure than the challenge - which outweighs any argument on CPP - is to be creative in how we use television in order to have a campaign that outperforms the average on the number of exposures delivered.

Bruce May from Bizperity, May 19, 2010 at 8:48 p.m.

Thousands of scientists working for Big Tobacco produced study after study that proved smoking is harmless. This went on for over 50 years. People have a tremendous capacity to believe in what they want to believe. The study in question did not answer the question, "Did these campaigns include print, outdoor, online or other media channels to make up for the lost impact of TV ads?" For my part, I can live without commercials but I can't live without quality content that is produced by the revenue generated by commercials. Producers do not get enough revenue share from monthly cable bills to produce anything more than a few good stupid cat videos... but I can watch those already online. Broadband TV or Digital TV (whatever we end up calling it) will not work without significant revenue for producers. Run the numbers. It just won't work.

Paula Lynn from Who Else Unlimited, May 20, 2010 at 2:51 p.m.

Listen to Bruce.