Solving the Ad Fraud Mystery

Contributed by Scotty Pate, Data Scientist at Goodway Group As children, many of us eagerly participated in the whodunit stories of The Boxcar Children, Nancy Drew, Sherlock Holmes, and The Hardy Boys. We cultivated our inner sleuth solving the case alongside our fictional friends, and dreamed of being just like them. While most of us never donned a smoking pipe or magnifying glass as adults, some of us still became detectives of sorts. Throughout my career as a data scientist, I’ve uncovered a variety of programmatic ad fraud plots and foiled more than a couple nefarious individuals online. At Goodway Group, my job is to team up systematic defense mechanisms for preventing invalid traffic with vigilant human monitoring against anomalies. Basically, I use those code-cracking skills from my youth to get hot on the trail of fraudsters; then I call in my gang of analysts to scour the online scene for suspicious characters. What does this really mean in the world of programmatic media? I’ll give you the CliffsNotes: The Plot: Pretend that you are running a video campaign to soccer moms in the Midwest. Delivery is hitting its daily goals, performance is knocking it out of the park, and viewability is over 75%. Sounds too good to be true, right? Then, it probably is. Returning to the Scene of the Crime: First, Goodway Group partnered with Pixalate, an industry leader in programmatic fraud detection, to automatically blacklist troublesome domains and IP addresses on a daily basis from our campaigns. However, using third-party data only gets us so far in solving this ad fraud mystery. Following the Clues: Beyond applying Pixalate data, our team of data detectives and machine-learning algorithms investigated every ad served in the campaign using impression-level data. From questionable characteristics on domains to the users we serve impressions to, we analyzed every piece of evidence, including hard-to-detect domain registry fraud and bogus social activities. Narrowing in on the Suspects: In this case, our villain registered a maze of domains and set loose an army of click bots to hide his shady efforts at ad theft, but he was no match for our information logic. By tracing our swindler’s IP address, we were able to wipe out the entire fraud ring at once, blacklisting all sites hosted on that same IP address. All in a Day’s Work: Catching the bad guys isn’t the only happy ending for Goodway Group. Where the programmatic ecosystem comes in with 8.3% fraudulent inventory and site directs show 2.4% ad fraud, Goodway Group was recently measured at only .96% invalid traffic, according to leading ad verification company Integral Ad Science. The moral of the story: We offer our clients programmatic campaigns at a lower fraud percentage than if they ran their ads with the sites directly! While I may not carry a badge, a gun, or a magnifying glass, for that matter, I know I’m doing my part to police the online superhighway and keep our clients’ programmatic campaigns safe. If you want to learn more about how you can protect your campaigns from ad fraud, contact us now or watch our recent Goodway Anti-Fraud Initiative webinar and get actionable tips to reduce your exposure and maximize your campaigns.  Scotty is a seven-year data scientist veteran, seasoned in developing and executing creative terabyte scale design applications across federal, finance, healthcare, and retail industries. In past roles, he structured data analytic solutions and formulated strategic recommendations for $60M+ DoD initiatives. At Goodway Group, he is tasked with balancing machine learning with human processes, moving easily from a vision and strategy to measurement and accountability. Ultimately, his job is to bring science to the art of programmatic digital advertising.