Monthly Archives: March 2007

Google: Click Fraud at 0.02%

Interesting post on inside Adwords blog about the extent of click fraud. The upshot is that Google is claiming that they are only seeing 0.02% of all clicks as being invalid clicks but initially recorded by Google as valid clicks. Here is the interesting bit:

Our Click Quality team investigates every inquiry we receive from advertisers who believe they may have been affected by undetected click fraud. Many of these cases are misunderstandings, but in most cases where malicious activity is found, the clicks have already been filtered out (and not charged for) by our real-time filters. Because of the broad operation of our proactive detection, the relatively rare cases we find of advertisers being affected by undetected click fraud constitute less than 0.02% of all clicks.

Put another way, for every ten thousand clicks on Google AdWords ads, fewer than two are reactively detected cases of possible click fraud. This proportion has stayed within this range every quarter since we launched AdWords, even as the issue of click fraud has received more widespread media attention. In the cases of reactively detected invalid clicks, a refund or credit is provided to the advertiser, and we utilize the discovery as a feedback mechanism to improve our proactive detection systems.

They explain it with a diagram as follows:

The interesting question though is how many of these clicks are invalid that even the Google Click Quality Team is not able to detect? I just don’t know its possible with filtering or with humans to detect all kinds of click fraud scenarios. The is especially troublesome because Google does not allow its advertisers to control where there ads will be shown. This makes it really hard for customers and Google to detect the fraud based on more controllable set of conditions thereby making it a more manageable problem. Apparently some advertisers are getting frustrated with Google and switch to one of their upcoming competitors, profiled recently by NYT:

Google and Yahoo have been fighting it out over which company will dominate the online advertising business, with Google maintaining the upper hand so far.

But in the competition for contextual text ads — those small sponsored links that run adjacent to related articles online — both companies are facing a challenge from a tiny but growing adversary named Quigo Technologies, a New York-based ad service that bills itself as an alternative to the giants.

In the last year and a half, a trickle of large media sites like ESPN.com, FoxNews.com and Cox Newspapers’ 17 sites have stopped using Google and Yahoo and instead signed up with Quigo.

What Quigo offers is transparency and control in what can often be an opaque business: advertisers pay Yahoo and Google for contextual ad placement on a wide variety of Web pages, but get little say over where those ads run or even a list of sites where they do appear.

Quigo, by contrast, gives advertisers not only the list of specific sites where their ads have appeared but also the opportunity to buy only on specific Web sites or particular pages on those sites. It also allows media company sites like ESPN.com and FoxNews.com a chance to manage their own relationships with advertisers.

Although Quigo remains a small competitor, with less than 10 percent of the contextual ad business, its growing success has apparently persuaded Google, which is accustomed to calling the shots in all aspects of its business, that it has to change the way it sells the sponsored link ads in the future.

Quigo still has a long way to go, but its nice to see some of the advertisers and web-sites getting a little bit more say in their ad placements. This can only lead to good things for the overall online ad market.

I paid to get to Digg front page

Fascinating piece on the Wired web site, related to how the editors put together a nonsensical web site, and got it on the front page of Digg by paying for diggs at User/Submitter (check out our review here) site:



Ten hours after hiring U/S, I had 40 diggs. The vast majority of them had also dugg the Photoshop tutorial or the $35 offer. This was the moment when I reached a tipping point, and I began to get a lot of organic diggs and comments. The crowd on Digg is drawn to what’s popular, and many of them second-guessed themselves when they checked out my blog and saw how crappy it was. Quomen commented, “None of those photographs really appeal to me. Am I defective? or just a loner.”Despite their doubts, Diggers kept digging my blog. There’s a perverse incentive here: Diggers who vote early on stories that become wildly popular become more “reputable” in the Digg system. If you’re trying to move up the Digg ranks, it’s in your best interest to vote on anything that looks like it’s gaining popularity. And my blog, with its flurry of paid votes, fit the pattern.

Interesting crowd dynamics at the Digg and the incentive structure does not help either…I have actually seen people comment they are not sure what a particular topic means but still Digg it. In this case more than 1/2 the Diggs were unpaid…That is just unbelievable. I guess the momentum trading theory is alive and kicking.

“We find it interesting that Digg still allows anybody to view any user’s diggs,” U/S told me in an e-mail. “By way of this ‘feature,’ User/Submitter is able to verify that our users actually digg the stories they’re given. Without this feature, Digg users are given complete digging privacy, and User/Submitter cannot exist.”

This is another perverse incentive…but this is driven by Digg’s drive to build a community. To fix this one, they will need to change the business model…All in all great job by the editors for creating and publishing this interesting experiment.