AppsFlyer helps mobile advertisers beat back ad fraud via machine learning

AppsFlyer is helping mobile game and app developers deal with the severe problem of fraudulent advertising responses via third-party ad networks. In doing so, the company said it has already saved brands tens of millions of dollars since 2016.

Today, the San Francisco-based maker of mobile attribution and marketing analytics is launching Active Fraud Insights 2.0, hoping to set a new marketing industry standard for detecting fraud. The platform leverages metadata from 98 percent of the world’s mobile devices, and it uses proprietary advances in big data and machine learning. Ad fraud is causing an estimated $7.2 billion to $16.4 billion losses a year.

AppsFlyer’s tool can review ad network partners in an effort to combat mobile fraud, where an ad network or other party creates a fake user that tricks the advertiser into believing that a real human is downloading or using an app or game.

Above: AppsFlyer is attacking the multibillion-dollar ad fraud problem.

Image Credit: AppsFlyer

Active Fraud Insights 2.0 detects different kinds of fraud, including DeviceID reset fraud, install hijacking, click flooding, and more. AppsFlyer works with more than 2,500 ad networks to regulate their activity to minimize fraudulent traffic.

“Over the last two years, the scale and sophistication of mobile fraud has grown at an alarming rate,” said Oren Kaniel, CEO of AppsFlyer, in a statement. “Thanks to AppsFlyer’s scale, we are in a unique position to detect and prevent mobile fraud faster and more consistently over time than anyone else in the industry. Further, since actions speak louder than words, we are going to regulate the AppsFlyer ecosystem and will remove bad actors from our platform in order to protect our clients — the marketers. As such, we have begun a continuous review and audit of all our integrated ad networks, ensuring they are aggressively doing everything in their power to minimize fraudulent traffic being sent to AppsFlyer clients.”

Rivals such as Tune and Adjust are also working on the problem, and the ability to do this well is becoming a differentiator in the mobile marketing and analytics industries.

Active Fraud Insights 2.0 is part of AppsFlyer’s Active Fraud Suite and is now available. The company said that since the introduction of the first version in 2016, dozens of large marketing teams have been able to identify and stop offending campaigns, networks, and even siteIDs before significant damage was done, detecting and addressing tens of millions of dollars in fraud. Active Fraud Insights 2.0 is grounded on a proprietary mobile engagement database — both the largest in the world and, with over 500 billion mobile events measured every month, the fastest growing.

AppsFlyer is used by Facebook, Google, Twitter, Snap, Tencent, HBO, Playtika, Waze, Alibaba, Kayak, Activision, and 12,000 other brands and partners.