• Wired, Thursday, July 13, 2006 10:45 AM

You'd think that of all the technology companies in the world, Google would be able to verify whether real users or automated programs are clicking on its ads. Network click fraud, the biggest risk to Google's $6 billion a year business, occurs when a site hosts Google AdSense ads, and the publisher runs a program to repeatedly click on the ads. That's easy enough for Google to spot, so clever fraudsters have turned to simulating IP addresses, and installing Trojan horses on other people's computers to generate fake clicks. The other kind of click fraud is competitive; it occurs when a company deploys a program to repeatedly click on the text ads of their competitors with the intended goal of depleting their AdWords budgets on wasted clicks. Firms usually hire shady third parties to do this for them; Wired links to one such firm that will commit click fraud for you. To be sure, the problem has become a security arms race. As Google and others improve fraud detection, fraudsters respond. While the search giant is putting money behind solving the problem, advertisers say it isn't doing enough. Wired says both sides are probably right: "It's in Google's interest both to solve and to downplay the importance of the problem." The click fraud question is a similar problem facing the online gaming industry: how can you tell if a player is computer-assisted? And what about online poker sites, where computer-assisted players have the potential to clean house. For its part, Google is now testing a new service model to deal with click fraud: cost-per-action. That's one way to change the rules of the game so click fraud doesn't matter.

Read the whole story at Wired »