I make no secret about the fact that I despise autoplay video ad units. You know what I’m talking about—you can’t stumble around the internet these days without eventually tripping over a video ad unit busy playing away, often with full volume, without ever having asked for it. As a consumer, the pain point is obvious. It violates an implied trust that sites have built with their visitors, that loosely rests on a trade-off of tolerating ads for free content. Bend that too far and consumers will balk, or so the theory goes.
But the problem isn’t on the consumer side at all, it’s a buy-side problem where media buyers are (sometimes unwillingly) fueling this practice with heaps of their precious media dollars.
Ask anyone who’s worked in the online video industry long enough and they will tell you horror stories of black box ad networks promising millions of video views at dirt cheap prices, littered with below-the-fold autoplay units firing away on sites you wouldn’t tell your mother about. And while paid syndication is a leg of online video that appears here to stay, it suffers from one of the long standing handicaps of our industry—the inability to effectively measure what the heck media buyers are actually buying.
How bad is the autoplay? Well brand-focused video marketing company TubeMogul wanted to figure out that question with some real data. So it ran a test of video ads running on some 7,000 publishers sites and found a whopping 400 sites were running what the company calls “fake pre-roll” ads totaling over 3.3 million impressions per day.
“Not only do these likely violate IAB Guidelines, but every media buyer we showed was (rightfully) mortified at how commonplace this appears to be across the industry,” said TubeMogul’s David Burch on their findings. Burch told me later that since their reported their findings that some site have stopped the below-the-fold autoplay practice while others unaware that they are doing it. He was unable to disclose which sites were the culprits due to contractual obligations.
Their research had a business purpose in this case, as the company launched a new ad verification service it calls PlaySafe, which will automatically block those fraudulent pre-roll ads from being served. “Pre-roll fraud is much more commonplace than people think, and preventing it is often like playing whack-a-mole. It’s disconcerting,” said Paul Kontonis, VP and Group Director for Brand Content at Digitas. “Verification technologies like this are reassuring.”
For TubeMogul, which now serves up ad units and tracks ads on some 55,000 publisher sites doing a combine 5 billion impressions per day, the new feature should help assuage fears amongst media buyers that those million dollar campaigns aren’t exactly performing as well as the brands think. The PlaySafe tech can distinguish whether an ad unit is true pre-roll or auto-play and match eligible buys accordingly, according to the company. And it’s now blocking an average of 3,315,465 fake pre-roll impressions per day.
A legitimate pre-roll video ad is 15 or 30 seconds in length and loads before the video content that a viewer has chosen to watch. In this way the pre-roll ad is similar to traditional TV commercial spots and is popular for advertisers and widely viewed as being effective. (TubeMogul)
Burch added that he does believe the effects of the fraudulent pre-rolls have put downward pricing pressure on the market for video pre-rolls which had otherwise remained a bright spot in online video with CPMs north of $10 a common sight for publishers.
What started as an analytics and distribution company back in 2006 now has a chance to build back some trust amongst brands in the multi-billion dollar video ad business. While the monitored set of sites is still a drop in the bucket of all site running rich video ad units, it’s a bold step towards rooting out one of this industry’s dirtiest little secrets.