Brands Lose Millions To 'Ad Collision'


An analysis of billions of ad impressions reveals that 5% to 7% of ads from the same campaign serve up on one Web page simultaneously, according to a study released Tuesday. The viewability report from AdSafe Media estimates that it costs advertisers between $60,000 and $70,000 on a $1 million ad buy.

The findings from the Q1 and Q2, 2012 Semi-Annual Review show that pages with a high percentage of what AdSafe CEO Scott Knoll calls "ad collision" are more likely to produce click and impression fraud. The report analyzes ad collision and ad viewability. The latter is defined as 50% or more of the ad being seen in the browser window for at least one second.

"The industry is so focused on cookie targeting and buying a user, they've lost track of targeting quality media," Knoll said.



When it comes to viewability by browser, Knoll said there is little difference. The study found about a 40% average across Chrome, Internet Explorer, Firefox and Safari, he said.

AdSafe's study found that directly placed ads are most likely to be viewable, according to the proposed 3MS standard, with 49.9% of ads in view for one second. On average, 20% or less of all ads remain in view for 15 seconds or more. 

When looking at viewability in relation to content and context, religion and spirituality gain the attention of Web site visitors the longest. On average, Web site visitors spend nearly 200 seconds on pages related to religion and spirituality, alongside ads that remain in view for nearly eight seconds. Home and garden follows as the second most captive category, with about 75 seconds on the page and nearly seven seconds with ads in view.

It makes sense that sites used to accomplish some task -- such as booking a hotel room or buying a house --see a slightly higher level of engagement overall, compared with sites that just deliver content.

Of the more than 40% of directly placed ads, close to half stay in view for at least one second, and less than a quarter remain in view for 15 seconds.

Adult content and illegal downloads represent about half of today’s high-risk ad inventory. These categories accounted for less than a tenth of high-risk inventory in late 2011. The study attributes real-time bidding to the growth of this year's higher levels of illegally downloaded content. Compare that to the declining percentage of content related to drugs and offensive language.

The appearance of hate speech rose from 2011 to 2012 on publishers and networks and fell on exchanges. Alcohol-related content -- which had declined in both the fourth quarter of 2011 and the first quarter of 2012 --rose across the board in the second quarter of 2012, according to the report.

9 comments about "Brands Lose Millions To 'Ad Collision'".
Check to receive email when comments are posted.
  1. Kevin Lee from Didit, August 28, 2012 at 10:56 a.m.

    So, let me get this straight. The NY Times charges extra for a "Surround Session" same or interrelated ad units on the same page and when it happens on a publisher site by accident due to Ad Exchanges and/or networks then it's an "ad collision" and automatically defined as bad? I thought studies showed that an ad replicated twice on the page dramatically increased recall? And if that's the case then at the low prices of exchange inventory doubling up may actually be a good thing in many cases.

  2. Steven Ustaris from OwnerIQ, Inc., August 28, 2012 at 1:08 p.m.

    Great insight in Kevin Lee's comments. We agree that 'ad collision' -or whatever you want to call it - occurs. But to assume that this is a 'bad' thing in all instances, especially when, as Kevin points out, there is data to suggest otherwise is self-serving. If this were true, then Media buyers should use this study as leverage to reduce costs of "roadblock" packages from publishers - b/c apparently having multiple ads on the same page is a bad thing

  3. Erik Ford from Kaizen, August 28, 2012 at 2:28 p.m.

    Agreed with Kevin and to build on that, I've worked with premium publishers where multiple ads on page, labeled 'road blocks' command much higher CPMs. So dual exposure, at a fraction of the CPM would more times than not, be considered a savings, not a loss.

  4. Arlo Laitin from Cardlytics, Inc, August 28, 2012 at 3:15 p.m.

    I think you buried the lede Laurie. The larger issue here is viewability. Ad collision is minor compared to the issue of non-viewable ads, and placement alongside illicit content. 50% non-viewable is outrageous - this # is actually higher among networks and RTB.

  5. Lorenzo Moreno from The Trade Desk, August 29, 2012 at 9:32 a.m.

    Yes, while its true that in cases where the goal of the campaign is brand exposure, multiple ads on the page are a good thing--basically free roadblocks.

    But if the goal of the campaign is reach, this is a problem, and in RTB, you could end up driving the cost up on yourself.

  6. Don Scott from BH Media Group, Inc., August 29, 2012 at 11:15 a.m.

    AdSafe needs to get their methodology right on this one. As prior comments demonstrate and I concur, "roadblocks" and other intentional multi-element campaigns are a significant part of online ad strategy. These must be factored out before making any statement about actual duplication of ad creative on any given website page.

  7. Kiril Tsemekhman from AdSafe, August 29, 2012 at 4:15 p.m.

    'Collisions' are good when they are intentional: in this case same data can actually be used to monitor compliance with roadblocking requirements. When collisions are uncontrolled and accidental they can lead to waste and drive up the cost. Especially on sites with 10-15 ads on a page - which are not a rarity on exchanges.

  8. Pierre Bouvard from Cumulus Westwood One, August 29, 2012 at 9:08 p.m.

    "The study found about a 40% (viewability) average across Chrome, Internet Explorer, Firefox and Safari, he said."

    Do that mean 60% of the ads were NOT viewable??

  9. Burke Scott from Microsoft, July 1, 2013 at 2:45 p.m.

    The problem with this methodology is that AdSafe starts off with the false assumption that "un-intended" ad collisions are wasteful with regard to Direct Response (DR) advertising. It's an easy false assumption to make though: "If there are two ads on a page and only one gets clicked on then the other ad must have been a wasted impression, right?" The problem with this assumption is that its not predictably accurate. Sometimes higher CTRs occur as a result of having more than one ad on a page, whether intentional or not. Therefore, it is false for AdSafe to assume when a click occurs on an ad that only one ad was necessary on the page to garner that click (and any duplicate ads are wasted impressions). We have run many DR campaigns with advertisers across network/RTB inventory that achieve very high CTR with low eCPC yet still have what would be considered high collision. When we reduced the collision rate the performance suffered causing the CTR to go down significantly and the eCPC went up. Therefore, it appears that ad collisions can actually be beneficial to the performance of DR campaigns rather than "wasteful."

Next story loading loading..