The New Yorker’s Peter Steiner famously created the cartoon caption in July 1993: “On the Internet, nobody knows you’re a dog.” Fast forward 22 years, and those “dogs” on the Internet have become menacing ad fraud “bots” on phantom publisher sites, maliciously-designed to generate fraudulent advertising impressions and clicks that, in turn, steal from digital ad budgets.
With the exponential growth in digital advertising volume over the past decade, ad fraud has become a systemic challenge for both digital publishers and advertisers. In a December 2014 study, digital security software provider, White Ops found that ad bots can account for anywhere from a 5 to 50 percent inflation in monetized audiences, adding up to an estimated $6.3 billion cost to advertisers in 2015.
The solution to ad fraud must start with proactive measures taken by publishers and aggregated supply sources. At PubMatic, we have vigilantly managed ad fraud exposure to single digits as a percentage of total impressions across our platform. Our Brand Shield solution (made possible through a partnership with ad fraud protection leader Integral Ad Science) constantly and automatically limits and filters suspicious impressions and inventory. Additionally, we continue to believe that buyers should not incur the cost for any proven fraudulent impressions, even the marginal amount that may get through our Brand Shield solution.
PubMatic recognizes the critical importance of maintaining high quality-advertising inventory on our platform. As a provider of marketing automation software solutions to the premium segment of the publisher market (PubMatic works with approximately half of the comScore 100 publishers in the U.S.), we take our commitment to supporting only premium and legitimate impressions very seriously—to not only protect the value of publishers’ digital assets, but also to preserve the integrity of high-quality inventory for advertisers.
We also know that ad fraudsters are constantly in search of new ways to exploit online advertising campaigns for their own gains—often relentless in their endeavors. Therefore, we’ve been vigilant in identifying bad actors on our platform. Over the past several years, we have removed a significant number of companies and bad URLs from the platform. Following the announcement of our Brand Shield initiative, in the past three months alone, our dedicated ad fraud protection team has blocked over 50 billion impressions across more than 450 supply sources and among tens of thousands of URLs.
PubMatic understands that the best way to combat ad fraud is to proactively identify fraudulent ads in their tracks, and leverage automated technology to detect potentially suspect sites before they ever reach our platform. It’s a constant battle, but based on the data that we’ve collected, it’s one that we are winning more and more.
PubMatic is deeply committed to helping the online publishing industry thrive. To achieve this objective, we are adhering to the following marketplace principles.