Click Fraud

Today's newsletter from eMarketer included a blurb about "Click Fraud":

"Click fraud" is the practice of using bots or people (usually hired in India or China) to click on advertising links displayed in search results with the aim of increasing the count and cost of the ads by making the traffic appear to rise. While it is impossible to work out the extent of click fraud, Google and the other sellers of search advertising are mute on the subject, some marketing professionals believe that they account for up to 20% of paid search fees in certain advertising categories. Others say as much as 50% of cost-per-click advertising dollars are spent on fraudulent clicks. Big problem.

I've always suspected that click fraud is a serious problem, but this is the first time I've seen an article about it. (I wonder if that's even an appropriate name for it.) We've been in the process of starting to do some online advertising, and this has been a concern. It would be easy for a competitor to click away on our ads just to hear the cash register ring. Heck, I've even done a little clicking myself that certainly went beyond the boundaries of genuine marketing research.

An obvious, albeit incomplete solution, would be to allow advertisers to exclude specific IP addresses from their clicks. At the very least you ought to be able to exclude your own IP! Advertisers also ought to be able to exclude IP's from outside their marketing region, and they should also be able to see a log of which IP's are creating the clicks. Perhaps the latter is already possible; I don't know.

Last week I specifically asked my Google rep if excluding IP addresses was possible. The answer was "no." He said it was the first time anyone had asked that question. I bet it won't be the last.

Update on Thursday, October 28, 2004 at 05:06PM by Rob

The term "click fraud" turned up in the October issue of Wired magazine.