Commentary

Current Viewability Standards Are Useless

by Paolo Gaudiano , Op-Ed Contributor, May 7, 2015

As publishers and advertisers scramble to improve the performance of digital advertising, much discussion is centering around the problem of viewability. It is my opinion that a major problem with impression-based performance metrics lies in the way that viewability is defined. I can see that for anyone steeped in the digital advertising ecosystem – from demand to supply and everyone in between – a viewability criterion of 50% of the ad visible for one second might make sense.

But for anyone who has ever visited a Web page, this viewability standard makes absolutely no sense. Even for takeover ads, I have become able to ignore the content completely, as my eyes jump straight to the X or whatever part of the page I have to click to get around the obstacle. Given typical human reaction times, by the time your DMP has captured my “impression,” I have not even had time to let out a puff of annoyance. When you consider more traditional display ads, especially in the page rails, this kind of viewability standard makes even less sense.

Recently, similar metrics have been proposed for videos, which in my opinion are equally laughable. Unless I am actively seeking a longer video, I find pre-rolls extremely annoying, and if it’s a common piece of news, I will frequently close out the offending window and go look for the content elsewhere. Even on my best, most alert day, it will take me at least two to three seconds before I realize what is happening and close that window. Ka-ching! Someone just made a buck on a “viewed” video.

What I find most perplexing about the apparent, collective mental freeze in the advertising ecosystem, is that the meaningfulness of viewability metrics can easily be tested. Cadres of graduate students have spent years in dark academic labs strapped to antiquated electronics equipment to measure their responses to stimuli even more mind-numbing than digital ads. In fact, there are several institutions that carry out all sorts of psychophysical tests specifically related to marketing and consumer behavior.

Why hasn’t the ANA or the IAB or the 4A’s or the WFA or some other illustrious body tried to do at least some sensible experimentation? For example, what is the typical reaction time of someone in response to a banner ad? A takeover ad? A video? Does our brain even react to an ad in one second? And what is the cognitive impact of only showing half of a banner ad? If I were back in academia, I could keep 100 grad students busy with these ideas.

From a superficial business perspective, I can see how the combination of fear and groupthink must quickly squash any such rebellious concept. But in the meantime, as the industry bemoans problems like click fraud, viewability and pricing transparency, we continue to dump tons of toxic waste upon the very readers that constitute the lifeblood of advertisers and publishers alike. The emperor has no clothes, and the empire is going up in flames.

programmatic, viewability

12 comments about "Current Viewability Standards Are Useless".

Check to receive email when comments are posted.

David Mountain from Marketing and Advertising Direction, May 7, 2015 at 3:47 p.m.
At some level, viewability on mobile should be judged differently than desk or laptop, just due to the nature of the small screen. But Paolo's central point of market research is fine, though the direct marketer in me quails at the thought that all viewability numbers are going to be dictated by a skewed demographic.

My knee-jerk reaction as to why the industry doesn't want to address this is that after a decade of providing data to the clients, then getting beat over the head and shoulders by it, while offline CPMs mostly hold their value... well, it gets tiresome. Someone come up with a viewability number for highway billboards or real-time TV, please. (crickets noise on)
Reply

Howard Zoss from Zig Marketing, May 7, 2015 at 4:46 p.m.

agree completely .. this a joke and everyone in the industry is culpable for this crap ... plenty of ways to test it and create better styandards ... but then nothing would be deemed viewable

Paolo Gaudiano from Infomous, Inc., May 7, 2015 at 5:53 p.m.

@David - I am not sure I would recommend going as far as fine-tuning viewability metrics to specific devices or demographics. An 80/20 approach would be fine: the current standards are a joke, it should be possible to come up with a fixed rule that is a bit more realistic.

And I completely agree that dumping the blame on agencies / trading desks is wrong, and is part of the reason why the industry has painted itself into this corner.

Paolo Gaudiano from Infomous, Inc., May 7, 2015 at 5:58 p.m.

@Howard, it is in fact a problem that between non-human traffic and viewability, the tru numbers of how many ads are actually seen is probably a small fraction of ads served - much smaller than the already low figures being presented today. As a consumer of information, I doubt that more than a handful of ads ever manage to grab my attention. I guess the question is whether the industry is ready to face a major change in which per-unit costs will go up dramatically, and so will engagement levels and accountability.

Ed Papazian from Media Dynamics Inc, May 7, 2015 at 6:23 p.m.

Paolo, you are perfectly right, but were a meaningful and validated standard developed, the result would be a huge discounting of digital "ad impressions", and the effect of accepting these realities would be catastrophic on ad revenues. If a branding advertiser discovers that the $23 CPM paid for his video ads is actually $95, while a typical TV CPM is in the vicinity of $10-12, even with digital's targeting superiority, the margins are too great to support much more than dabbling in digital.

Unfortunately, all of the blather that is being thrown about to the effect that TV has its "viewability" problems too, is utter nonsense---- only the most gullable advertiser will buy the notion that every viewer runs to the bathroom every time a commercial is on the TV screen.

I'm afraid the the digital ad selling community is going to try to bull it through, promoting highly unrealistic---and inflated----viewability standards as a way to protect its ad revenues. Perhaps direct response and other forms of promotional/sales service advertisers will accept much higher ad rates tied to realistic viewability metrics, but I don't think this will have much traction on the branding side.

Ted Mcconnell from Independent Consultant, May 7, 2015 at 7:53 p.m.

I think its time to temper this pile-on with some facts. The in-view standard has never pretended to measure whether someone actually saw an ad. There are other measures for that.

The job of the in-view measure is to validate that the ad could have been seen. Why such a low bar? In fact, its the same concept other media use called "Opportunity to see". The media has done its job if they give viewers an opportunity to see an ad. Simple. The creative has done its job if the creative unit sitting in the media causes a consumer to be reminded or persuaded. You can't make the media responsible for bad creative, as would happen in the alternative standard you seem to be suggesting.

Clear enough? The measure isn't trying to do what you seem to think it is. Never was. And yes, brains react to ads in less than a second. And yes, cognative science was consulted. Anyway, most of your complaints are about intrusive ad formats, so actually, while I feel your pain, I think that by attacking viewability you might be barking up the wrong tree.

Paolo Gaudiano from Infomous, Inc., May 7, 2015 at 9:28 p.m.

@Ted, you make a valid point, but you are hanging on to a definition of viewability that, while historical correct, is functionally not very useful, and increasingly so given the rapid technological advances we are witnessing. I can understand that back when digital advertising was in its infancy and we had no technology to track more nuanced aspects of ad delivery, a minimum guarantee of viewability might have made sense. But today it maeks no sense, especially when you consider the money and energy poured into things like targeting and programmatic. What is the point of delivering the "right" ad to the "right" audience if that audience is probably never going to look at the "viewable" ad? Hanging on to anachronistic performance measures encourages intellectual laziness, which results in a flood of crappy ads that annoy much more than they engage. If the advertising ecosystem could start to think about consumers - how they think, what they like, what is meaningful - instead of focusing on increasingly opaque technologies and vacuous metrics, I believe that we could push toward a much better world where advertisers engage more sparingly and more meaningfully with consumers.

Paolo Gaudiano from Infomous, Inc., May 7, 2015 at 9:36 p.m.

@Ed: your math make sense but only relative to the current situation. Yes, if someone were asked to pay $95 for a digital video view, they should absolutely go back to TV advertising -- if TV advertising can deliver the sales or brand affinity they seek. But if that $95 digital video ad delivers conversions that are 5x as high as TV conversions, then the video ad is still better. The more meaningful the metrics, the more accurate our ability to measure the results and determine true ROI. Market forces could then push pricing where it belongs.

The current proliferation of digital media across multiple channels and devices creates a unique opportunity to gain an unprecedented understanding of the consumer. In my opinion, the majority of efforts to leverage technological advances focuses on increasing the loundess of the message rather than fine-tuning it. As long as advertisers think of CPMs, CPCs, CPAs, CTRs and foget the most important C - the Consumer - we will continue to witness an escalation of the intrusiveness and ubiquity of advertising, which in the end will benefit nobody.

Ed Papazian from Media Dynamics Inc, May 8, 2015 at 6:55 a.m.

@Paolo, I appreciate the point you are making, however if we take an average branding campaign, with the same ad message and demo targeting and compare the likely effects between a TV "exposure" and a digital "exposure", it is difficult to make the numbers work for digital. With TV giving the brand a 100% visibility factor----even if many viewers leave the room or cease paying attention----- the ad will be recalled by 30-40% of the "audience". Of these, around half ---15-20%---will link the brand's name to the comercial and roughly 5-7% will change their mind about the brand and be more inclined to use or buy it than before they saw the ad. These are typical normative findings from numerous ad recall studies, using a variety of methodologies. All things being equal---same viewer or user, same brand, same ad message, how can digital expect to match TV's average exposure recall/motivation norms, when only 30% of its audience can even see the comercial in its entirety?

Of course, one can say, let's count only those times when the ad is fully visible on both screens. Fair enough, but even then, we have to consider the issue of screen size, other visual distractions that are on the page, etc. What's more, the contention is that digital ad sellers should charge more than before per "delivered" impression, to make up for the "views" lost due to fraud, truncated messages, invisible ads, etc. So, now, instead of charging $23 per thousand "viewers", the figure rises to $50 or $75 or who knows what-----while TV's CPM remains the same. To make that work, a digital ad exposure to the same person, using the same message for the same brand, must generate double or triple ----or more--- in recall, brand linkage and buying motivation. That's a tall order.

I realize that there are other factors to be considered and an ad campaign's effects are more than the sum of its individual exposures relative to costs ( CPMs ). The use of digital can be rationalized in many positive ways----finer targeting opportunities, timimg advantages, reaching a consumer in a different way, the ability to control frequencies, interactive response capabilities, etc. but the time is long past when digital can defend the indefensible situation that prevails today. If digital wants to be an effective ad medium and garner a fair share of branding dollars, it needs to get organized like an ad medium and stop hoping and praying that TV will lose its audience, so advertisers will simply have to come to digital----and take it the way it is. That's not realistic.

Paolo Gaudiano from Infomous, Inc., May 8, 2015 at 1:04 p.m.

@Ed - you are right. My response was a bit glib, and the likelihood that anyone would pay $95 for a digital ad and get 5x impact is non-existent. You said it much more eloquently when you suggested there are ways to rationalize the positive aspects of digital. Part of my frustration is when advertisers seem more intent on justifying questionable metrics rather than trying to extract value in new and creative (pun not intended) ways.

George Ivie from Media Rating Council, May 8, 2015 at 2:49 p.m.

Paolo, we appreciate you continuing the very important discussion on viewability. To help clarify, the Viewable Impression Guidelines simply establish a minimum opportunity to see threshold. The metrics you mentioned (50% of pixels in view for 1 second for display ads and 2 seconds for video ads) are not a measure of engagement or effectiveness. They indicate a moment of qualification: they represent the first instance where an audience had an opportunity to see a digital ad and therefore be included in an audience based currency metric we are also developing.

Crucially, these guidelines establish counting mechanisms and should be used as just one ingredient in the ad campaign measurement mix. Again, they are a foundational metric for the forthcoming digital audience-based currency metrics. The intention of viewability is to define whether a consumer even had the chance to see the ad, not whether they engaged with it or took action as a result of seeing it.

The viewable impression’s time requirements were determined after extensive study of hundreds of millions of data points in various environments and contexts. Based on this research, we arrived at two continuous seconds as the minimum time required to qualify as a viewable video impression precisely because it proved to be just prior to the time when people began to recognize and act upon the ads: the opportunity to see. For more information, please visit: www.measurementnow.net or www.mediaratingcouncil.org

Paolo Gaudiano from Infomous, Inc., May 8, 2015 at 4:13 p.m.

@George - Tank you, I appreciate the clarification as I am sure will other MediaPost readers. I understand the value of this metric, and that its purpose is meant to ensure that an ad could have been seen. Your organization's recent definitions for mobile of loaded ads and served impressions helps further to clarify the meaning of these metrics.

But when we consider industry-wide CTR levels, even when you account for viewability, there is clearly much that could/should be done "downstream" of the initial possible view. Perhaps that's where bodies like the IAB and 4A's could look for opportunities to define additional metrics that take into account what happens after an ad has the chance to become viewable.