Study: Online Video Ads Beat TV Ads In Viewer Recall

  • by May 24, 2011


Viewers pay more attention to online video ads than to traditional TV commercials and also recall them better, according to new research that utilized Affectiva's facial tracking algorithms and second-by-second biometric modeling of cognition, excitement and stress levels.

The research measured the reactions of 48 viewers watching one hour of programming in Interpublic Group's West Coast IPG Media Lab.

Conducted by the Media Lab during March in conjunction with video ad network YuMe, the study determined that on average, online viewers pay more attention to the screen than do traditional TV viewers -- and the greater attention levels carry over to advertising.

Online video ads received 18.3% more viewer attention in the study than TV commercials -- a much higher disparity than the 8.5% greater viewer attention garnered by online video content over TV content.



This was largely due to the finding that when transitioning from program content to ads, the attention of TV viewers dropped off three times faster than that of online viewers.

While fast-forwarded ads, such as those recorded on a DVR, were partly to blame, IPG and YuMe found a much larger cause to be "the familiar cadence of TV content." Conversely, the study found that "systemic disruption" caused by online video's "unpredictable ad cadence and forced, short bursts of video ads (rather than predictable ad pods) appears to help decrease ad avoidance behaviors without causing consumer backlash..."

In fact, DVR users were found to pay higher-than-normal attention to commercials because they were concentrating on skipping them -- but their later recall was actually lower. "DVR users were 38% less likely to correctly recall the brand for any TV ad they saw, aided or unaided, compared to non-DVR users," the study found.

For TV viewers overall, paying attention during commercials had no effect on whether they could or could not remember them or recall them unaided. And those who recalled ads when aided actually had below-average attention scores. But online viewers who paid attention to ads later recalled them, with online video ad recall twice as high as TV ad recall -- "offering proof that recall and attention correlate," according to IPG and YuMe.

Both TV and online video content had plenty of competition from other media during the study, most notably smartphones. The 48 viewers studied had been surveyed a week earlier to determine how they normally watch TV, then told to bring with them what they needed to recreate their "normal" TV viewing experience.

They showed up at the Media Lab not only with phones, but with laptops, games, magazines, food, makeup and even a guitar, IPG and YuMe said. Other predetermined distractions were also available -- from DVRs stocked with participants' favorite shows to their own email and IM programs.

All these off-screen distractions hurt attention levels more during advertising than during program content. As the study summary put it: "If given the chance to avoid ads, most people will."

Yet while distractions affected both online and regular TV viewing, with over half of video viewers said to have had a second screen active during commercials, the distractions hurt TV to a larger extent -- with 62% of TV ads avoided while they played, compared with 45% of online video ads. Also, 17.5% of TV ads were fast-forwarded.

"DVRs hurt TV advertising in a way that has not yet come to online video," according to the research summary. "Compared to TV, online video is measurably better at delivering ads that are impossible to skip technologically and impractical to skip behaviorally."

A final conclusion: "normal ratings data is not nearly complex enough to measure how chaotic and highly individualized media consumption is."

7 comments about "Study: Online Video Ads Beat TV Ads In Viewer Recall".
Check to receive email when comments are posted.
  1. Raoul Marinescu from Pluto TV, May 25, 2011 at 7:27 p.m.

    "The research measured the reactions of 48 viewers". Really? 48 viewers, that's all it takes to call something "research". I guess!!!

  2. Kevin Horne from Verizon, May 26, 2011 at 3:53 a.m.

    Strapping someone into a straight jacket (see pic complementing article) is probably the only hope online video ads have...

  3. John Grono from GAP Research, May 26, 2011 at 4:34 a.m.

    Well that gave me a giggle Kevin.

    And Raoul, I have had first-hand experience with eye-tracking research in the OOH sphere, and biometric tracking in the television sphere.

    As the metrics that are being recorded are basically limbic functions (viewing, breathing, sweating etc) they are generally (i) very homogenous across the population - e.g. men use their eyes in the same way as women etc (ii) are not subject to respondent error - e.g. such as in recall diaries, button-pushing, answering questions etc, (iii) can't be easily gamed. Therefore, what is traditionally thought of as a low sample size is generally more than adequate when working in this sphere.

  4. Doug Garnett from Protonik, LLC, May 26, 2011 at 3:16 p.m.

    The problem with this purely factual observational research is that we're jumping to a massive set of assumptions based on limbic functions.

    I equate it with paleontology. Paleontologists are masters at reading the fossil record and creating theories about what those facts they see imply about bigger things. Except, these paleontologists told us in the 60's that dinosaurs were colorless, loner, reptiles. But today they tell us that many dinosaurs are related to birds, had feathers, lived in flocks, etc...

    But with Dino's, it didn't really affect a manufacturer's future whether they were right at the time or wrong.

    WIth this research, the researchers look at (so we're told) limbic system and only observe what they're able to observe.

    Connecting these observations with meaning is a massive jump. Is it any more reliable than any other research? I don't think so. Does it reveal more truth? Probably not - it probably clutters more than it reveals.

    Besides, note that these researchers aren't even up to date on TV. DVR's are now shown to have IMPROVED TV ad effectiveness - not decrease it.

    But they're just selling a service and have apparently been so distracted by limbic's that they missed this truth.

  5. Doug Garnett from Protonik, LLC, May 26, 2011 at 3:24 p.m.

    Oh, yes. Beside the method being seriously questionable (a mere development of the dial method), what's with this headline?

    Let me re-state it:

    "If you can hijack someone's web browsing experience and get them to choose to watch your video, they'll tend to remember it."

    "By contrast, when programming is interrupted by advertising, a lot of people won't remember it. Some will and that's how TV has always worked. Oh, by the way, TV reaches millions more than you'll ever reach with your web video."

    Having noted this truth, did we really need to torture 48 people with diodes in order to learn what was already obvious?

  6. Aaron Hendon from Proceed Media Group, May 26, 2011 at 6:02 p.m.

    Let me get this straight - people will skip ads if they can? Really? And if they can't skip them they don't? Good to know. I think more research here is needed just to be sure.

  7. Frank Nichols from Media360inc, May 26, 2011 at 7:18 p.m.

    Dr. Ernest Dichter proved back in the 1950's that advertising recall was influenced by attentiveness to the medium in which it was presented. Comparing Television and Internet streaming video using the methodologies employed in this study almost made me laugh. This sounds like somebody's agenda based research in which the outcome is not only logical, but predictable. But then to claim that one medium is more effective than the other without taking into consideration the inherent variables aligned with each is a rather sophomoric approach to claiming any sort of comparable bottomline effectiveness of the advertising.

Next story loading loading..