Log in

No account? Create an account

Previous Entry | Next Entry

and the results are in ...

Fair Isaac's results indicate click fraud is a potentially large problem in certain online ad segments. However, they have concluded that competitive ad clicking (employees of one company click on those of another) is not too harmful to advertisers.

They are asking for more advertisers to submit data so they can study the problem in more depth.


( 2 comments — Leave a comment )
May. 19th, 2007 05:06 am (UTC)
I don't fully understand the issue -- but if they can detect fraudulent clicks well enough to claim with reasonable certainty that 10-15% of billed click traffic is fraudulent, couldn't the same technology/techniques be used to avoid charging for the fraudulent clicks in the first place?
May. 19th, 2007 11:51 pm (UTC)
Re: *confused*
For more background information about the issue, you can read the report prepared by Dr. Alexander Tuzhilin, who was retained by the courts to evaluate Google's fraud detection practices. (A link to the report, and my comments on it, are here.)

The same technology could be used to avoid charging for fraudulent clicks. Google, Yahoo! and some of the others do not charge for clicks they consider "invalid". The question remains, do they capture all, or even most of the fraudulent clicks? There is also the issue that not charging for "invalid" clicks creates an expectation in the advertisers' minds that these clicks will never be charged for. The engines and ad networks incur a cost of processing all of these clicks, and essentially, spend money to determine that they cannot charge for a service that they render. As click fraud (and other types of ad fraud) increase, due to increased sophistication of fraudsters, the costs of processing increase as well. These costs, if not absorbed by the advertisers, will come out of profits.
( 2 comments — Leave a comment )

Latest Month

July 2018

Page Summary

Powered by LiveJournal.com
Designed by Tiffany Chow