Posted by: Rob Hof on February 28, 2007
Google soon will introduce some new tools to fight click fraud, such as the ability to blacklist suspected fraudsters and prevent them from even viewing ads to click on. But I can’t help but feel that the announcement tonight is really intended more to convince advertisers that they don’t really need them. Most of what Google’s announcing today are numbers that it hasn’t released before, backing up its longstanding claim that actual click fraud is very low.
Google has complained for some time about consultants’ and analysts’ estimates—outlined in a BusinessWeek cover story last year—that click fraud constitutes 10% to 15% or more of clicks. For one, Google claims, invalid clicks—that is, clicks that may be fraudulent or just spurious—are less than 10%, and that it catches those before advertisers even get charged. Moreover, it claims that invalid clicks identified after advertisers asked for them to be checked out are less than .02% of all clicks. In other words, it’s saying the rate of actual click fraud is incredibly small.
Gord Hotchkiss, CEO of Enquiro Search Solutions, a search marketing firm briefed by Google (and, full disclosure, offered by Google itself as an expert), thinks Google’s numbers are accurate as far as they go—more accurate, in his mind, than the higher claims of consultants. I can’t judge that myself, and I’m sure some advertisers will point out that Google isn’t saying how much click fraud it think it doesn’t catch in the first place. That’s “unknowable,” says Shuman Ghosemajumder, business product manager for trust and safety at Google.
For that reason, Google’s numbers won’t mollify everyone. But the more it opens up that black box and helps advertisers better understand the murky art of search advertising, the more it will build confidence among advertisers that they’re not getting conned.
Danny Sullivan at Search Engine Land has a more in-depth review.