This Leaked Document Shows How Big Brands' Ad Budgets Get Spent On Asian Porn Sites
Upon further inspection, the performance data the client was receiving had been encrypted to hide the fact that many of the clicks registered in the campaign had originated from Asian porn sites with names like "realmalesexscandals.com" and "boyfriendoffers.com," the document says.
In other words, the clients' ads had been placed on porn sites, and the clicks they generated were disguised to look like traffic from mainstream web brands.
In the online advertising business, there is a widespread terror of saying publicly what everyone actually knows: That a huge portion of clients' ad budgets are wasted on fraudulent clicks. It's common for people to complain about adtech fraud, but extremely uncommon for anyone to describe specific clients who have lost money, or the companies who handled their spending.
It is estimated that advertisers waste $7 million a month on click fraud, much of it from botnets. Augustine Fou of the Marketing Science Consulting Group estimates that up to 22% of all clicks on web ads are fraudulent.
So we were interested when someone sent us this report from Telemetry, an adtech auditing firm, which purports to show that an international video ad budget for groceries ended up being spent on "false referrers, [who played video ads with] sound muted, thumbnail sized, targeted at Asian & the Middle Eastern audiences and hosted within sites of an explicit almost 'sex tourism' faceted nature."
Unfortunately - and this is a measure of just how scared adtech execs are of telling the truth - the document has been redacted for legal reasons to disguise the company whose money was wasted. (We contacted executives we believe had knowledge of the campaign but they either did not respond or declined comment.)
The document does, however, describe step-by-step how adtech fraud works. It contains a chart showing what normal traffic looks like coming from clicks on ads hosted by reputable online media brands. The chart indicates that incoming traffic is random and irregular - just what you'd expect:
A second chart shows traffic coming from suspicious web sites bought through an ad network the client had thought was reputable. The traffic is exactly even across all the different brands - a highly unlikely scenario:
The media brands were also a weird mix: CyclingNews.com is a reputable site. But Momsational hasn't been updated since 2010. Yet the two appear to be generating clicks on the client's ads at exactly the same rate.
So Telemetry dug a bit deeper. The ad network's results were being served using software that encrypted the original data. Once that data was decoded, Telemetry found that traffic wasn't coming from Momsational or CyclingNews at all. Some of it was coming from Asian porn sites. Here's a look at the encrypted data, followed by the real data underneath it:
Other traffic was coming from no-name search engines used on the home screens of internet cafes in Singapore and Vietnam, such as "search-results.com." This graphic from the document shows how one of the client's ads ran as a tiny, barely visible box on search-results.com, but the referring URL was disguised to make it look as if the traffic was coming from Momsational:
Telemetry's conclusion? That this was a "sharp practice at the very least":
A source tells us that the client stopped spending money with the offending ad network, but there were no further consequences.
That begs a question: Why would a large, sophisticated advertiser tolerate such lousy partners? Part of the answer is that, in the adtech world, almost nobody has any incentive to stop click fraud:
- Client executives don't want the embarrassment of admitting they were fooled.
- Chief marketing officers dare not go to their CFOs - the people who control their annual ad budgets - and confess they have been wasting their money.
- Advertisers seek the cheapest clicks possible, but reputable mainstream brands are expensive. So they search for clicks en masse at less-reputable networks that sell their ad space cheaply.
- When those low-quality networks generate a lot of clicks, it makes the campaign's results look better.
- As web ad buyers - agency trading desks and demand-side platforms (DSPs) - compete for clients' business by touting the effectiveness of their campaign results, they are loathe to lower those metrics by weeding out worthless incoming clicks.
And as long as nobody complains, everyone keeps getting paid.