We’ve had dozens of industry discussions, forums and initiatives, yet fraud levels in digital advertising have gone unchanged since a year ago, according to a new report carried out by White Ops, a fraud detection company, which was carried out in association with the US-based Association of National Advertisers (ANA). The study analysed more than 10 billion online advertising impressions across 1,300 campaigns for 49 ANA-member companies, who are for the most part leading US/international brands, and found that the estimated average annual loss to bots among the 2015 study participants was $10 million.
While it’s a commonly held belief in the industry that it’s mainly the bottom of the market that’s targeted by bots (i.e. scenarios where the inventory is suspiciously cheap), White Ops found that higher value CPMs are actually even more at risk of seeing bots, with video being particularly vulnerable:
However, the study also found that direct buys helped mitigate the risk. The report rather bluntly separates ‘programmatic’ and ‘direct’ into two distinct categories without accounting for ‘programmatic direct’, which would in all likelihood see lower bot rates for programmatic buys on quality sites.
That said, advertisers buying ‘programmatic video’ – and we’re guessing that the majority of this was bought on open exchanges – had 73 percent higher bots than the study average, with a bot range in video programmatic buys of 1 to 70 percent. However, the study found that only a small percentage of direct video buys were measurable due to publishers’ transparency and measurement policies that did not permit full fraud assessment on the buys.
Sourced Traffic Helps Fuel the Bot Economy
More than three-quarters of the study’s participants had higher bot percentages in traffic bought from third-party sources compared to unsourced traffic, with sourced traffic being three times more likely to contain bots than unsourced traffic. However, sourced traffic in 2015 showed a slight improvement over 2014, when sourced impressions were over four times more likely to come from bots.
Malware Threat Looming
Dan Kaminsky Co-Founder and Chief Scientist, White Ops, also commented in the report that the ad industry was likely to become the target of malware via: (a) outright remote compromise via “drive-by downloads” exploiting a vulnerability; (b) “bloatware” shipped with computers; (c) black-box libraries unwittingly embedded into otherwise legitimate applications; or (d) install wrappers that add remote-controlled services along with some functionality the user desires.
Kaminsky also notes that advertising fraud is perceived slightly different to other types of fraud for a couple of reasons. He said, “Advertising fraud has the curious status of almost seeming legitimate — you couldn’t expect to get away with raiding a bank account or accessing someone else’s Gmail account, but defrauding advertisers, even by using the host user’s identifying cookies, doesn’t seem nearly as criminal. While the ecosystem suffers, the end user sees very little impact from the fraud. For the bot operator, however, the scheme is quite profitable. Many do not even operate their own infrastructure. So this sort of fraud has a surprising number of “legitimate” participants. We’ve found companies where not everyone at the company knew they were fraudulent operations.”