Forensiq Releases Next Generation Fraud Detection Algorithm
Ad Fraud Prevention Innovator Exposes Sports Bot Using Latest Machine Learning Techniques
NEWS PROVIDED BY
Oct 19, 2017, 09:00 ET
SHARE THIS ARTICLE
NEW YORK, Oct. 19, 2017 /PRNewswire/ — Forensiq, a global leader in ad fraud detection and prevention, today announced the release of its next generation of botnet detection algorithms, focused on exposing and preventing the latest methodologies to generate sophisticated invalid traffic and steal advertiser dollars.
Forensiq’s second generation algorithms introduce dramatic improvements to its automated traffic detection mechanism, primarily through expansion of its machine learning methodologies. The algorithms introduce improvements in automated traffic detection by using deeper behavioral analyses of browsing behaviors to identify patterns associated with automated browsing that simulates human traffic. As fraudsters come up with even more insidious ways to generate human-like bots, the algorithms will allow Forensiq’s team of data scientists to detect this type of malicious behavior more rapidly. Supply-side players, in particular, will benefit from the improvement by allowing them to detect suspicious sources and keep their ad supply as clean as possible.
Notably, the new algorithms have already flagged up to 75% of pre-bid requests to NFL team domains running through Forensiq’s supply-side partners on suspicion of fraud. This prompted a comprehensive study spanning a much broader set of team sports and sports media websites covering the NFL, NBA, NHL, and MLB, leading to the discovery of Sports Bot. Through the course of the study more than 9.7 billion pre-bid requests analyzed translating to an estimated cost of $200-$250 million.
Sports Bot used highly sophisticated techniques to fraudulently load ads on the affected sites without the site owners’ consent, leveraging a new methodology that allows it to monetize inventory on premium domains without resorting to tactics such as domain spoofing or masking. Although the bot’s specific deployment methodology is not yet known, it may originate with malware downloaded unwittingly onto users’ computers that allows fraudsters to both hijack browser activity and manipulate browser content. Such devices would allow fraudsters to invisibly appropriate the reputations of premium websites and unwitting users without either discovering that they’ve been taken advantage of, siphoning money from advertisers in the process.
To understand the potential impact of the bot, Forensiq partnered with AD/FIN, a leading digital media intelligence company. Together, we analyzed media spend and pricing data across impacted domains, combining it with Forensiq’s estimate for the scale of the bot to assess the economic impact of Sports Bot.
According to AD/FIN CEO Andrew Altersohn, “Forensiq’s findings are both fascinating and scary. While progress is being made against fraud, it’s findings like these that tell us we need to double down on our efforts. It’s unfortunately too common for advertisers to pay for unwanted or worthless ads, and at the same time top-shelf publishers take a hit on reputation and revenue. Partnering with Forensiq is another critical step towards addressing the problem.”
David Sendroff, CEO of Forensiq added, “Advertisers view sports websites as prime real estate, but fraudsters have taken advantage of these properties’ high viewability and high value traffic by hijacking computers with bots. Advertisers are bidding for ads that no one is seeing. The key for advertisers and publishers is to be vigilant–they need to identify and monitor fraud and ideally, stamp it out before they bid.”
Over the course of the last year, Forensiq has seen a marked increase in the ability of botnets to mimic human behavior and fake key advertiser performance indicators, such as viewability and engagement. For example, many of the sites visited by Sports Bot scored as high as 90% viewable as measured by Forensiq; however, the ads were actually shown to bots that were able to simulate viewability.
With its acquisition by Impact Radius last year, Forensiq has tapped into the expertise of an expanded technology team to create Forensiq Fraud Labs, a unit comprised of data scientists and ethical hackers to drive faster development of new algorithms. Core development has been accelerated by the combined teams and their ability to leverage the company’s combined data streams.
“TAG is excited to see these types of investments in improving the detection of non-human, bot traffic. As one of the first companies to achieve TAG’s Certified Against Fraud Seal, Forensiq has continued to demonstrate leadership in helping the industry fight fraud. We also applaud their commitment to having their practices independently validated, a top priority for leading marketers,” said Mike Zaneis, CEO, TAG.
Forensiq, an Impact Radius company, is the market leader in full-funnel digital fraud detection offering a series of leading solutions to fight ad fraud, measure viewability, and brand safety. We combine the latest machine learning technology with a dedicated team of data scientists obsessed with helping clients stay ahead of bad actors to eliminate wasteful spend and maximize their ROI across their marketing channels. Forensiq was founded in 2010 and is headquartered in New York City with seven offices worldwide. The company has won three LeadsCouncil LEADER Awards, including Best In Class for Fraud Detection as well as DM News Most Innovative Marketing Tech to Watch. To learn more, visit www.forensiq.com.