How good is Google’s fraud detection??

[jiff]

Hi,

A few months ago I noticed some suspect activity on our AdWords account, which I suspected to be competitors clicking our ads, so I added some free fraud detection software to our site.

After about 2 months I collected all the data I thought to be fraudulent and sent it off to our AdWords rep. I few days later I received an email stating they were crediting my account due to, I quote “potentially invalid clicks that weren't automatically filtered from your reports”.

Excellent I thought, but about a month later I noticed some more suspect activity so I collated some more data and sent it off again. Again I received another credit to our account due to, I quote again “clicks which did not fit a profile of normal user traffic”

On average we receive under £0.50 a month back in automated adjustments to our account, yet in the last 3 months I have received over £70 because I sent Google my suspect data. If Google spends so much time and effort on fraud detection, why can I, with FREE detection software, an hour or two to filter the results (which could be automated quite easily) find far more fraudulent clicks than they can?

A reply from a Google rep would be nice :eek:

[hardball]

According to Schmidt, Google's auction-based pay-per-click advertising model is inherently self-correcting. Schmidt's scenario for what would happen if Google did not police click fraud and it was "rampant":

"Eventually, the price that the advertiser is willing to pay for the conversion will decline, because the advertiser will realize that these are bad clicks, in other words, the value of the ad declines, so over some amount of time, the system is in-fact, self-correcting. In fact, there is a perfect economic solution which is to let it happen."

Apparently the CEO is content to let fraud "happen", so the issue is not technical but philisophical/moral. Its up to the advertiser to lower bids to account for the fraud factor, not googles problem.