Want to experience the joy of ad fraud in near real time? Here is your chance. Posters to Reddit reported an ad fraud program in the last 48 hours involving fake 300-by-250 video ads for McDonald’s.
Our understanding is that the value of the fraud is tiny. Instead, the interesting aspect of this story is that it is a nice little study on the extent to which the “benefits” of ad fraud are widely and quickly distributed across the ecosystem including to legitimate adtech players.
Who takes a cut? Clearly, the originator of the fake ads is making money from deliberate fraud.
“The originator is the seller of ‘traffic’. They sell this by volume on a click or server-request basis and they get their payments from the websites or ad networks who are their customers,” according to Shailin Dhar cofounder and director of research at adtech and ad fraud consultancy Method Media Intelligence.
He told Which-50, “Sometimes ad networks or specialists at publishers learn how to do it themselves and then localise the operation.”
An analysis by Mark Pilipczuk, Senior Digital Product Director, Digital Propositions, Audience Solutions at Acxiom reveals that India-based RTB (real-time bidder) Streamlyn is the display buyer and instigator, while the display DSP (demand side platform) is MediaMath and the display ad server is Zedo.
On the video side, Pilipczuk identifies Cedato as the video player, PubMatic as the video (SSP) supply-side platform and he says video verification is provided by Pixalate.
A Pixalate spokesperson however denied Pilipczuk’s suggestion saying their role was misunderstood. “It is a false claim that Pixalate has validated or has verified these ads or their content. Pixalate provides services to businesses transacting in digital advertising to ensure that ads were served to and viewed by real users via valid traffic detection and filtration services.”
“Pixalate has no relationship with the alleged perpetrator of these ads flagged by the Redditors. Pixalate’s business model is not scanning and validating ad creative – a model employed by numerous other businesses – perhaps leading to some confusion by the commenters regarding Pixalate’s role in this particular case.”
In the absence of any mitigation, all would ultimately benefit from the fraud.
According to Redditor MrMattCannon who first identified the fraud, a fake display ad is served which then makes requests to multiple SSPs. “Behind the scenes, this then makes multiple requests to video SSPs.”
“The creative has USD pricing but is serving in huge volumes to the UK.”
We asked Dhar how easy it should be for legitimate ad tech vendors to identify the fraud and mitigate against it.
According to Dhar, “This all starts with SSP’s vetting their ad-networks and publishers properly to find out what kind of sites they actually represent.”
The problem, he suggested is that this process is not always as robust as it should be.
“That does not happen across the board because salespeople at the SSP’s have their own incentives to onboard as many people as possible.”
Adtech vendors need to adjust their detection methods, he suggested. “Empty ghost ad-calls on rapid fire from the ad-server are easy to prevent and identify if you are exonerating human machines, and not trying to identify suspicious mouse strokes or IP addresses.”
“This type of IBV (in banner video) fraud is far too common. It’s the primary cause of video fraud which is in the billions of dollars per year,” he told Which-50.
It is important to note that ultimately companies like PubMatic and MediaMath do not really benefit from the fraud, as they have refund policies in place to compensate their clients where fraud is detected.
Jeff Hirsch, CMO and Head of US Publisher Development, PubMatic told Which-50, “At PubMatic, we have been an outspoken proponent of transparency and quality across the digital ecosystem and as such, we take fraud very seriously. When any fraud is detected across our platform, we shut it down immediately and block the culprits. In this case, we have taken action and blocked the creative as well as the buyer.”
“As always, with our fraud-free program, the buyer is protected when any fraud is detected. We will continue to monitor the situation and take any necessary action to prevent and mitigate fraud from existing across our platform in the future.”
But not every adtech vendor has such policies, although they are becoming more common.
And the adtech industry is a long way from real-time mitigation. The volume of transactions in the ad tech world, and their real time nature makes this almost impossible with current technology.
Instead, vendors adopt a protect and mitigate policy where the goal is to identify and stop the fraud, and identify any refunds during the period of the campaign or immediately afterward so that the clients are not out of pocket.