Commentary
To Combat Fraud, Pixalate Brings In The Hardware
- by Tyler Loechner , Staff Writer @mp_tyler, June 25, 2015
The programmatic ad industry is obsessed with
the invisible. It runs on software and data that floats somewhere in “the cloud.” The servers that act as the brains behind it all are tucked away in “data centers.”
Putting everything behind the scenes and on autopilot allows marketers to focus on what they say they do best. You know, buzzwords like “strategy” and “creativity.”
But veiled threats exist, too.
In a bid to combat ad fraud, Pixalate is turning toward actual hardware. The ad tech firm this week introduced the wordy “Pixalate Security Threat Intelligence Dome,” a hardware appliance that cuts off suspicious activity before it reaches the exchanges or networks. The new offering comes out several months after Pixalate unveiled its fraud blocking technology.
Per a release, the hardware keeps tabs on “IP reputation,” identifies URLs associated with malware of phishing activity (and blacklists them, along with bots), and identifies proxies, such as TOR nodes or network gateways that are used to hide real IP addresses. All of this is meant to clean up the supply -- not by finding bots that exist in the supply, but by not letting them enter the pool in the first place.
Pixalate’s tech is the latest in a string of recent innovations in the fraud fighting field. Last week, Are You A Human introduced its human whitelist, which verifies humans instead of blacklisting bots. Earlier this week, PubMatic struck a partnership with Integral to give buyers more control over the risks they’re willing to take with suspicious activity. This month also saw AdTheorent introduce its own fraud detection tech for mobile and AppNexus broaden the steps it's taking to fight fraud.
Pixelate is also offering its new hardware appliance in “virtual form” -- also known as the most hyperbolic way to say “software.”
Tyler Loechner is a MediaPost Reporter covering real-time, programmatic, RTB and more. You can reach Tyler at tyler@mediapost.com.
How accurate is this method, assuming it has been run enough for Pixalate to know? One hundred percent fraud elimination is not possible, so shouldn't the vendors and buyers settle on an acceptable threshold level, then try for incremental improvements from that point? The fraud problem will persist as long as there is enough money to be made doing it in comparison to the effort of pulling it off. Perhaps the goal should be to make the monetary loss inconsequential, akin to how retailers deal with shoplifting.
@henry -- they claim to block all "known" fraud (which is of course what they would do, why would they let known fraud through?). No word on the overall % blocked.
The idea of isolating frauds and possible fraudulent traffic before it reached any ad-tech vendor boxes is great, should have been done years ago. If Pixalate filters are good, this will help to reduce hardware overhead and increase effective QPS. There are hardware appliance for all sort of things such as antivirus, anti spam, etc, this is long overdue.
The one company that approaches digital ad fraud from a different level is Grey Cloak Tech. Even with Pixalate running it's systems on Site Scout traffic, Grey Cloak is still seeing highly suspicious traffic coming through on auto dealers known to go through agencies using Site Scout. It's a real arms race out there. Suggest advertisers take matters into their own hands - Grey Cloak works on the Advertiser side in most cases putting the onus on Ad agencies to come with legit traffic. If they don't, and just trust the supply side, they'll be writing quite a few credits and refunds.