Sometimes online display advertisers end up overpaying for campaigns because of ne’er-do-wells who falsely click on ads or refresh Web pages, resulting in fake click or impressions.
Several ad tech firms have complained about the problem to AdSafe Media, said Will Luttrell co-founder and CTO of the brand protection and ad viewability company. Companies have been able to detect click fraud and report it after the fact, but now AdSafe said it has developed the ability to block it from happening in the first place through its Suspicious Activity Detection product.
The product features a predictive scoring system that rates site domains on a scale from zero to 1,000, much like an individual's credit score, that lets media buyers and ad exchanges prevent ads from running on sites that will likely expose advertisers to click or impression fraud.
“The approach in the past has been to find out what sites are causing problems and then create a blacklist," AdSafe CEO and president Scott Knoll said. "The problem is, the bad guys are very good, and they continually come up with new ways of masking their activity. So there’s no way to solve this with a blacklist."
Luttrell wouldn’t go into specifics about how the technology works, lest those who engage in the suspicious activity find ways to game the system. However, he said AdSafe could combine its own viewability data with its partners’ clickthrough rate data to see when the numbers don’t align properly. In one example, he said, “we saw one site with a 5 percent clickthrough rate, which is in and of itself really suspicious, but 98 percent of the time the ads were never in view."
AdSafe is making Suspicious Activity Detection available for ad networks and exchanges to use so that when buyers set up a campaign, they could specify that they don’t want their ads running on any site with a score that's below 700, for example. The system would help ad networks that might access media channels exposed to the suspicious activity and be kicked off exchanges so as not to subject exchange buyers to fraudulent charges. As for publishers that may be fraud victims and risk becoming blacklisted, AdSafe can put a monitoring pixel on their sites to detect the exposed sections of a publisher’s inventory so that the publisher can clean itself up.