TAG, an association of ad-industry trade groups, worked with ad buyers at GroupM, IPG Mediabrands and Horizon Media and the metrics firms Moat, Integral Ad Science and DoubleVerify to scrutinize 6.5 billion display and video ad impressions on TAG-certified distribution channels from July through October. They estimated the so-called invalid traffic—typically bots simulating humans to run up bills on automated ad platforms—at 97 million, or 1.48 percent.
To put that number in perspective, the industry average for invalid traffic across display and video is 12 percent, TAG says.
The results are valid, according to John Montgomery, global executive VP of brand safety at GroupM, the world's largest buyer of advertising on behalf of marketers. "They used data that used TAG sourced inventory and then compared it to fraud levels that would have occurred had we not done so," he says. They checked it with used verification companies accredited by the industry's Media Rating Council.
"Not every impression across the web was measured," Montgomery says. "But it's a fair and representative sample."
Of course, it's unclear how much money marketers lose to ad fraud each year because the digital ecosystem is so complex. Some experts peg the number at $6 billion and falling while others believe it is climbing to as much as $16 billion this year.
"That's always the concern with fraud," Montgomery says. "We can't afford to be complacent. We are continually questioning our partners methodology in an effort to make it as strong as possible. The key here is to use technology to mitigate fraud and work with organizations like TAG to try and stay on top of it. To not do so, is to throw away a chunk of your budget because a percentage of your impressions will be invalid."
If TAG didn't completely eliminate invalidate invalid traffic, it made the problem a lot more managable, according to CEO Mike Zaneis. He compares the remaining bad traffic to damaged electronics a retailer may receive, or produce that goes bad at supermarkets—simply an aspect of doing business in the space.
"No matter how good one company is at identifying fraud, there is no silver bullet," Zaneis says. "We need every entity in the supply chain to identify and flag what they are buying. And it starts with the agencies, demand-side platforms, exchanges and down to the publisher as well."
It's that last part the industry as whole is struggling to overcome. Zaneis says if the digital supply chain took proactive measures and worked with TAG-certified companies that the entire ad fraud problem would essentially become negligible. Obviously, that's not the case today, and it will take significant effort to reach such an end.
TAG's Certified Against Fraud initiative awards so-called seals to buyers, sellers and intermediaries after they complete "rigorous anti-fraud requirements." Those who are awarded such seals —there are 49 today—receive a unique ID that identifies its ads to trading partners in the supply chain. Those identifiers can also be matched with the payment ID system to verify that payments aren't going to criminals.
"Fraud thrives in the dark crevices of the supply chain, so we knew that we had to get the legitimate participants in the supply chain to adopt the same high standards for this effort to be successful," Zaneis says. "When the industry links its arms and stands together, there is no place left for the criminals to hide."
Companies will often point to their own safety mechanisms—anti-fraud security vendors, for example—to thwart ad fraud, but Zaneis says that's not enough.
"When you pour yourself a cup of water, you want triple filtered water," he says. "That first filter may remove most impurities, but some will still get through. It's that second and third filter that removes everything else."