One thing’s always bugged me the most in industry conversations about digital ad fraud. I mean, yes, finally, we’re having those conversations, and that’s great. But everyone seems set on totally separating the conversations about fraud problems between display ads versus video ads. The symptoms have been different. The tactics have been different.
Frankly, the fraud problems with display and video ads are just about the same.
Except that when it comes to video, the situation may actually be worse, something that was brought to light in a Wall Street Journal article earlier this week about just how mind-blowing video ad fraud is. Piling onto that is a Financial Times article revealing that more bots than humans saw a recent Mercedes-Benz digital video campaign.
The big takeaway is that, if we are going to take the issue of ad fraud seriously, the problems we have to tackle are even more pronounced and serious for video ads. A few months ago, the IAB launched its new guidelines on viewability -- an extremely big step in the right direction for the whole industry. But the video ad problems mentioned in the WSJ article also shed light on the ways in which those guidelines still don’t go far enough.
The most obvious? Fraud’s a different issue than viewability, though the two issues feed on each other. In video, this is particularly bad because fraudulent impressions still register as “viewable,” and thus the new standards in their current state don’t help. And video needs audio to be effective. There’s no standard for that.
More importantly, these fraudulent video impressions can actually perform. These are real peoples’ computers being used by botnets visiting pages with unviewable ads. Those real people, when they return to their computers, are likely to eventually purchase or sign up for something, like a trip or a credit card, sometimes thanks to retargeting from those fraudulent video views. This means the fraudulent ad can actually show the marketers behind it a perceived ROI, even though the owner of the computer never actually saw the ad.
So, since they can’t separate fraud from legitimate views, marketers who are purely ROI-focused need to depress the overall price they’re willing to pay for video impressions to factor in this waste. And this drives down the price that a quality publisher can get for its inventory.
To be perfectly blunt, this sucks for everyone -- even if we think we’re reaping the benefits.
Keep in mind, too, that when we talk about these fraudulent video ad views -- which are somewhere between a quarter and a third of all video ad views -- we aren’t incorporating the people who are using ad blockers. Or who use an ad as a reason to open a new browser tab and ignore the ad.
I’ve said here before that the way the ad industry is (often complicitly) packaging up millions of meaningless views, until they look like something good, bears a lot in common with the subprime mortgage crisis. And, similarly, we could be facing a crash. This is a problem in all kinds of online ads, but as the WSJ articulated, it’s particularly complex and pronounced in video because of how sophisticated that technology (and the ways to trick advertisers into paying for it) can get.
The new IAB guidelines focused overwhelmingly on display, but the association is going to need to start toughening up its take on video, too. Because when “drone pools” are something our industry considers acceptable, there’s a problem.