As fraudsters get smarter, anti-fraud players must evolve to put up a worthy fight.
To scale its fraud detection capabilities, viewability and verification measurement vendor Integral Ad Science (IAS) has acquired bot detection company Swarm. Terms of the deal, which was announced Friday, were not disclosed.
“Fraud is not going away, it’s getting more sophisticated,” said IAS CEO and President Scott Knoll. “We’re seeing it take on a new chapter of its life where it’s not just focused on whether ads are viewed, but it’s scrolling on pages, mousing over things and replicating any metric advertisers are looking at. It’s acting more and more like humans every day.”
IAS detects fraudulent activity in two ways.
First, it analyzes massive amounts of data – the company views more than 10 billion ad insertions each day – from both the buy and sell sides to uncover abnormal distribution patterns and other anomalies.
Second, it analyzes individual impressions to determine if they are fraudulent, a process called side-channel analysis. Swarm’s technology will allow IAS to do side-channel analysis at full scale.
“You’re asking specific questions of the computer to find out whether or not it’s been affected by malware,” said David Hahn, head of product management at IAS.
Knoll added, “Side-channel analyzes every impression separately without any history, while big data looks for anomalies in the universe.”
IAS has been vetting side-channel analysis technologies for two years. The Swarm acquisition is part of a 12- to 18-month road map to enhance its abilities in the area.
Not only can Swarm’s technology easily integrate into IAS’s stack, but it complements IAS’s own side-channel analysis capabilities, Hahn added. And investing in both big data and side-channel analysis technologies differentiates IAS from other anti-fraud vendors.
The deal is also an acqui-hire: Swarm’s 18-month-old team will move to IAS’s new San Francisco offices and be folded into the IAS brand.
“We’re immediately going to leverage the technology for our own fraud detection to continue to expand it,” Knoll said. “We’re also very excited about the team itself – they’re very knowledgeable, and in combination with the team we’ve got, it will continue to accelerate our progress.”
