Commentary
How To Solve The Viewable Ads Debacle
- by Cory Treffiletti , Featured Contributor, March 28, 2012
If an ad shows up on a Web page, and it’s below the fold where no one can see it, did it make an impression?
The buzz this month is viewable ads, and the issue of whether or not your ads are being seen on the Web. It’s a big issue because, as we all know, click-through is a horrid metric and actual click-through rates are abysmal, to say the least. The only metrics that we’ve had left to champion have been exposure and engagement, and you can’t engage with something you never actually see.
The comScore findings that 31% of all online ads are never seen are alarming, but also somewhat comforting. If click-through, engagement and interaction rates are so low, then at least we can discount those poor metrics by 31% and assume that if 100% of the ads were seen, then maybe the numbers would at least be a little bit better!
However, the fact that this high a percentage of ads are wasted is an issue because one of the strongest elements of online advertising has been the opportunity to reduce waste. Online provides targeting, which means you can expose your message to only the most prime audience members -- but now we find that a third of those prime audience members are not seeing the ads. That is wasted value.
advertisement
advertisement
Back in the old days there was the concept of OTS, or opportunity to see. This concept was applied to the counting of online ad impressions, and was used by ad servers to come to a standard definition for an impression. That definition is in place, and has been in place for years. But it’s clear the spirit of that definition has not been enforced by the long tail or the short tail of media. comScore also surfaced that the issue of viewable ads doesn’t get fixed by paying a higher CPM. It’s an issue across the Web.
What does this mean for the industry?
Well, one possible fix might mean that you tier the pricing for ads. My suggestion would be that only ads above the fold are able to be priced on a CPM basis, and any ads below the fold of a standard website should be priced on a CPC or CPE (cost per engagement) basis. Additionally, I would recommend that ads on mobile browsers be sniffed and priced in a similar fashion, realizing that the fold is much higher. In the case of mobile, only ads at the top of the page would be CPM based and the rest would be performance priced or some other model.
This is not a difficult issue to fix, in theory, but in practice it’s a doozy. The bad news is the vast majority of “impressions” on the Web are going to become devalued rather quickly, and the overall opportunity for revenue from the long tail of websites is going to drop dramatically. The good news is that the best inventory, the truly premium above-the-fold, quality inventory, is going to be priced higher. It’s a simple supply-and-demand discussion. There is more supply of below-the-fold, lower-quality inventory and a finite supply of premium placements. Think of it as a market correction for the ad industry. These kinds of corrections are good in a maturing industry, and necessary if we are to take our rightful place as the number-one advertising and marketing medium in the eyes of big brands.
Of course, we could just piddle around on this issue for a couple of years, and have “committees” and “initiatives” put in place to deal with it. That would be helpful and productive (not so much).
How do you want to see this issue addressed?
YES with one caveat Cory -- below the fold imps can also be used as "added value."
Regardless this is a very simple fix to a complex problem -- the only argument you will get back I imagine is that there are some impressions "viewed" below the fold so you are throwing out some value there -- nonetheless nice job proposing an easy solution -- easy should be the new black...
Once again a non-scientific report causes a stir. Let's face it, We see few complaints about a billboard or subway skinning that goes unseen, and what about the DVR? But suddenly, in spite of being able to access great inventory at impossibly low CPMs in many cases, 'viewable ads' has become an issue.
Every campaign, based on KPI and or ROI/ROAS is different and should be judged that way. Sometimes, cheap below-the-fold impressions on sites with great content where people are still reading at the bottom of the third page work better than pricey, ego driven premium.
I agree and believe that market forces can and will address this. I'd rather see more time put into improving creative so that when an impression is seen it's not a blight on the digital landscape.
"Half the money I spend on advertising is wasted; the trouble is I don't know which half."
Fact is there will always be some portion of advertising that's unseen or wasted. Ultimately the market decides a rate that is acceptable for both parties to benefit or the deal doesn't get done.
I'm with Stephen, the "problem" is more than unviewed ads, let's see more energy into creating great advertising.
Technically speaking, the price of viewability is already baked into the cost of advertising. If pricing is efficient (a bold assumption, I know), then the cost of ad impressions should be deflated by the percent of ads that are not viewed (because that is how much the conversion rate should be deflated). Once ad views are measured and confirmed, the price of those ads will just inflate and the marketer running the campaign will get the same CPA or eCPA as before. I think market forces as Stephen suggests will fix this. Smart targeting companies will quickly realize the arbitrage opportunity in viewed ads (they are currently under priced but deliver more) and smart pubs will soon see the value in making ads more viewable (i.e., more revenue for them). I think forcing any strict pricing policy and/or ad view standards may end up causing a debacle. Influential industry policy makers and those with deep pockets will end up skewing policy and standards to their favor, leaving the smart and smaller players less able to steer this marketplace towards its potential efficiency.
Cori, I think you may be slightly misinterpreting Comscore's 31%. I believe it was 31% fraud...the percentage of ads that are not seen extend way above the 31% figure. Banner blindness probably accounts for more unseen ads than fraud...the two combined make a deadly duo.
Some worry that online ads are never seen by browsers with AdBlock installed.
Good stuff Cory...while Publishers like us have lamented over the years about the fact that "our job is to bring the audience to the table and the advertiser and the creative's job is to get them to notice you"...the viewable ad issue is a real one since as you said, if there is no opportunity to see, it doesn't matter how great the creative or the call to action is. We decided to solve this by creating a simple technology that only renders the ad on our storypages if it is within an individual user's viewable browser window. Meaning every ad is viewable and billable as such. No brainer from a value standpoint for the advertiser...much tougher for us as a publisher to do given that we anticipated anywhere from 15-30% inventory loss, thus revenue. However, with design/UI tweaks, better and more compelling rich media ads, and commitment to really trying to solve this simple problem (and support from our board) we were able to launch ServeView: http://www.msnbc.msn.com/id/40078155
We were hoping that more premium publishers would join us but that hasn't happened yet...although with folks like Comscore and the IAB 3M initiative, there seems to be more momentum.
Actually, if I'm not completely wrong, the 31% is related to not only fraud, but non-human interaction (which may not necessarily be fraud at all).
But I could be wrong. Right now, there seems to be a lot of information flying around about viewable impressions.
It's my point of view that the viewable impression is valuable because it eliminates the concept of "above the fold" and "below the fold". Below the fold was 'bad' before the viewable impression because it was perceived to be unviewable. What we're learning is that below the fold ads ARE viewable, but usually at not as high a rate as above the fold.
On the other hand, it's possible that below the fold ads have a longer time spent in the viewing frame of the user - thus making engagement with these ads more 'valuable' than headline banner ads which disappear from the screen with the slightest scroll down.
While more creative ads, and better advertising in general, would be welcome, you can't avoid the quantifiable nature of this (or any other) medium, regardless of whether you believe the methods used are viable or not. The only question is whether or not a majority of people agree that the method of quantification is acceptable as a currency. It's easy to see the lure of, and value applied to, a viewable impression.
What's not to like? Rather than worrying if an ad is rendered and not seen, you can verify if it was in a viewing frame - which makes it as valuable as an ad in a TV show. In either case, you know the ad ran in a placement which eyeballs have the opportunity to see it.
In addition, it limits inventory. Given the questionable nature of so many publishers where ads are placed when running in ad networks, we can determine if the money spent on these placements are worth it at all. If I spend $1 on a network that has 5% viewability, then I really place a $20 CPM for an ad that was really seen, and that's a great basis for comparison to other premium publishers.
Doug Ferguson mentions Ad Block, and the reality is the viewable impression takes this into account. If an ad is rendered, but blocked at the viewer's computer, we will know this and not have to pay for it.
I'm not sure why people are upset or worried about the viewable impression. We all know there are problems with any method of quantification - this just reduces the impact of one of those factors.
What kind of ads? Banners?
It looks like the point being made here is to establish different pricing based on where an ad is, or its "viewability quotient." What is the problem we are trying to solve here? We are trying to assure that advertisers receive the appropriate value for their spend and that publishers deliver value. If we had appropriate measurement systems for the value of online advertising beyond clicks and actions, true attirbution, then it wouldn't really matter how "viewable" the ad is. What would matter is what your ROI is on the spend on that placement. This then makes all placements on the same playing field in terms of providing value. DR advertisers know this and the eCPA of a campaign drives determination of inventory/campaign value. If we had a similar metric for branding campaigns, we wouldn't need to create havoc in our industry trying to adapt to a radical concept such as 'viewable' with all the tremendous cost of change that goes with it.
You're absolutely correct, Cory, especially with regards to mobile. The problem, in my opinion, is the publishers, who obviously benefit from getting paid for unseen ads. There would need to be a technical solution for above-the-fold detection in this world of ad networks and DSPs where we are often far removed from the actual placement of the ad.
The assumption that 100% of above the fold placements have the potential to be seen is wrong. Realvu, the only MRC accredited metric for the viewable impression, has documented that there are a multitude (20) of reasons why an ad served never gets to the page at all. Advertisers have and continue to pay for an ad that never ever has the opportunity to be seen. The total value of dollars wasted is incomprehensible. The only viable solution is to move to the viewable impression metric -- It;s the only metric that works.
Great article but I think everyone is avoiding the “elephant in the room”, which in this case, is frauds… While 30% of ads are below the fold, probably the same number or even higher ads are presented through non human traffic or a non existing site. With the popularity of non transparent URL’s for cheap advertising you can just imagine what's happening in that space…
We ran a test for one of our client for driving traffic to his site and found that ~40% of the traffic to the site was non existing traffic. Thanks to our brand safety measures, which are an integral component of our online video advertising platform, we blocked the vast majority of this fraudulent traffic.
Whether it's 31%, whether it's fraud and/or non-fraud, bothers me less than the sheer fact that apparently some large percentage of impressions cannot be perceived. If I'm running some campaigns for brand value rather than direct response as such, I now have to discount the value of my spend by a considerable amount.