Google's looming public relations nightmare: click fraud

Henry Blodget has covered click fraud heavily, and it remains a concern to him. It should be a concern to anyone who uses Google Search -- searcher or advertiser. Although the Google folks continue to downplay the click fraud problem -- including a gloss-over response at yesterday's Press Day 2006 --  my belief is that it can not be underestimated, regardless of the sophisticated algorithms in place to detect such mischievous and damaging activity. This Wired article further demonstrates the potential earth-shaking problem at hand. If fraudsters weren't successful, then computer viruses, worms and phishing scams would not be such a large problem.

Ok, I know -- comparing the web search customer to an artificially intelligent computer program designed to spot abnormal click trends is a stretch. But can even the smartest computer programmer possibly anticipate the limitless number of variables an equally-sophisticated programmer may use to unleash click fraud schemes --  and especially for black-coat high-paying customers? It's not just click fraud when a human or manned click farm simply clicks on a Google ad repeatedly -- the bot farms that could (or are) developed most likely appear to click from countries all over the world, spend varying amounts of time on the clicked site, and come from IP addresses that appear unconnected. This is what intelligent click bots do, and they are probably getting better every day.

I'm sure Google -- of all companies -- is way in the forefront of detecting click fraud on its global search network. After all, advertising around search results makes up nearly all of Google's revenue. It's easily in the best interest of Google to keep the tightest of tabs on this kind of activity. However, you can throw that out the window if click fraud becomes a bad public relations situation for Google.

An adage I'm fond of quoting is that perception is reality, and Google's response, so far, is that click fraud is not a problem and they are on top of what little click fraud there is. Google's AI engineers may have this looming issue under control, but the way in which they communicate does not really instill confidence in Google's advertising customers. It will only become worse as a perceived problem unless Google can better communicate that they can quash this issue. Perception is reality to the rest of the world, including the public market.

Update: this blog post over at Cnet tells of a lower-than-30% click fraud rate that purports to be more realistic. We aren't so sure.