Integral Ad Science Launches Real-Time Fraud Prevention Platform


ADOTAS — Integral Ad Science just announced the release of real-time fraud prevention for digital advertisers and publishers. This solution provides the industry’s first and only actionable technology that leverages in-stream signals to detect fraud and eliminate it in real time, before an ad is delivered or sold, which ensures that advertisers only pay for ads served to humans.

This technology works across multiple platforms and around the globe, and it is now available to Integral’s clients, including over 70 of Ad Age’s 100 leading national advertisers. It is designed to specifically improve upon and solve the challenges and limitations of the industry’s current solution set, such as log-level reports, white lists and black lists that are reactionary, and limit scale.

“We are thrilled to offer the industry’s only real-time fraud prevention technology for both the buy and sell side,” said Scott Knoll, CEO of Integral Ad Science, in a press release. “Online fraud is getting more and more sophisticated, and our mission is to stay one step ahead of the game and keep the advertising industry honest and safe. The real-time component is essential because the bad guys are aware of basic tactics, like white lists and black lists, and are too dynamic and savvy to get caught in this trap. The only true way to prevent fraud is to take action in real time a thousandth of a second prior to the delivery of an ad, evaluating each and every impression.”


  1. Interesting, but…quite underwhelming.

    I am very familiar with fraud detection from my days as GM of the US search business at

    The problem is, real-time detection identifies only a small portion of invalid clicks.

    (I don’t say “fraud” because there is also unintentional but invalid click activity. Double-clicking an element, for example, might send 2 click events, but they are clearly meant to be 1.)

    Regardless, being able to analyze clicks over time is actually far more important than doing only real-time detection.

    BTW, the search engines, at least, are really good at detecting fraud. Can’t say much about other ad tech vendors. (And of course a publisher has no way of doing it.) But they are certainly not fool-proof; they do let some fraud through.

    The problem is, how CAN they detect the stuff that currently gets through? They are already really good at detecting fraud, so what ELSE can they do?

    One approach is having the ad servers, publishers and search engines actually cooperate in order to do the BEST job. That’s because each one individually sees only what IT sees. It cannot analyze patterns across the other ad servers/sites/engines. But that cooperation ain’t gonna happen anytime soon.


Please enter your comment!
Please enter your name here