Visible Measures, an ad tech company specializing in video distribution and measurement, on Tuesday said its platform is completely preventing ad fraud, a claim that few if any others would dare to make.
"There is a theoretically possibility that fraud can occur, but the reality is our clients are experiencing no fraud," said Brian Shin, CEO of Visible Measures.
That's despite estimates that billions of dollars are being wasted on ad fraud, in which bad actors charge marketers for ads deliberately served out of sight or to software routines simulating real people. It has seemed unlikely to completely prevent ad fraud in real time because ad tech connects ads with available impressions in just milliseconds, and because fraudsters evolve tactics as countermeasures arise.
"Without knowing what they are doing or what the small print is, it would be extremely unlikely that they can guarantee a 100% fraud-free rate," said David Hahn, senior VP of product management at Integral Ad Science, which helps marketers avoid fraud and maximize the viewability of their ads. "It is almost like saying, 'We are going to rid the world of crime and we are going to guarantee it.'"
"There are a lot of measures and protection capabilities you can put in place, which will reduce ad fraud to a very manageable number in the low single digit percentages," Mr. Hahn added. "But the rate in which things change and the rate in which these guys come up with new ways of defrauding is pretty significant, and there is a lot of rapid development happening on the bad guy side as well as the good guy side."
Two other ad tech players, DataXu and TubeMogul, offer refunds to clients for ad fraud they detect on campaigns after the fact. Unlike those companies, which generally claim their fraud rate is under 2%, Visible Measures has become confident enough to declare its platform entirely impervious to fraud.
"Our clients have really been the ones who have been pushing us to offer this, they really are concerned about fraud," Mr. Shin said. "We think this is the next logical step rather than just saying, 'If something happens we won't charge you money.'"
When pressed on Visible Measures' fraud rate, Mr. Shin said the company is seeing campaigns with zero percent while acknowledging that total perfection was elusive. "Less than 1% would be more comfortable for room for error," he said, "but it is closer to zero percent."
Mr. Shin said he defines ad fraud as "anything that is considered non-human" and "any time an ad is not visible to the user."
Visible Measures has accreditation from the Media Rating Council and works with powerhouse brands including WalMart, Coca-Cola Co., Unilever, Ford and Volkswagen. It recently helped Sam's Club score 24 million video views for a private-label diaper ad, and says it monitors over 500 million videos daily.
The company works with analytics vendors such as Moat and DoubleVerify to bolster its fraud detection. Mr. Shin said its fraud-fighting approach includes tracking real consumer viewing patterns, such as mouse movements and what users do during and after watching a video. "Eventually, it becomes pretty obvious which are non-human," he said.
Fraudsters likely to flock to video
Calling a video ad platform 100% fraud-free will strike many as particularly bold. Ad fraud is difficult enough to counter, but fraudsters are particularly motivated to work in video, where ad prices are higher than in display advertising.
Within video, inventory with a cost per thousand impressions of $15 or more has 2.7 times more bot activity than cheaprer inventory, according to a recent study by WhiteOps.
"The end goal for a bot is to sell you that it isn't a bot," said Amit Joshi, director of product and data science at Forensiq, another fraud and viewability specialist. "Bots are very smart. They know how to simulate scrolls, or sit on a page for 30 seconds to simulate user activity. It isn't like before where they just clicked something and bounced."
Nobody knows exactly how much is lost to ad fraud each year. The Interactive Advertising Bureau said in December that advertisers spent $4.4 billion on invalid traffic in the U.S. last year. That compares to roughly $60 billion in overal domestic digital ad spending.
The Association of National Advertisers, meanwhile, says more than $7 billion will be lost as a result of ad fraud in 2016.
Others claim the number is much higher, estimating that $18.5 billion, or a third of all digital advertising revenue, is being lost to fraud.