On the hand, such lack of clarity very much hinders genuine advertisers from identifing clicks they believe are bad and for which they are still being forced to pay.
We're setting up our own tool to help clients identify which clicks from Adwords we believe are clearly fraudulent & to allow us request a refund. This should be based only on the most blatant abuse as we know everything else will be ignored by Google.
So far we've come up with the rules below. Any thoughts or suggestions would be welcome.
Rule 1 - If there are at least 4 clicks from one IP address using the same ad tracking code within 3 minutes of each other, this is fraud.
Rule 2 - If there are at least 4 clicks from one IP address using the same ad tracking code within 30 minutes of each other, this is highly suspicious.
Rule 3 - If there are at least 2 clicks from one IP address using the same ad tracking code within 3 minutes of each other & repeated at least 5 times in one 24 hour period this is highly suspicious.
Thanks
J
Also, IMHO it not just the click from the same IP address that can indicate its fraud, but what visitors from said IPs are actually doing - if they don't do thinkgs that most of your visitors are supposed to do (ie download images, execute bits of JavaScript (could be good check actually), then these clicks are probably made by humans.
If so, then how do you determine:
AdSense affiliates gaming the Google system with bots that shift IPs? Definitely fraudulent, but often tough to separate from "real" users in a similar IP block (i.e. AOL, MSN, etc.)
The range of IPs to be used as an identifier for AOL-type users, who can (and do) have their IPs switched dynamically every few minutes?
When a "real" user hits "reload"?
And how can you be confident you are presenting an accurate list of fraudulent log entries that Google actually charged you for? Do you use a second form of authentication, or just log entries?
I ask because Overture maintains that their anti-fraud system stops charging you when it kicks in, however the visitors will still be sent to your site and log an entry with your tracking code. Google's is no doubt similar in that they won't STOP a visitor from clicking the ad, but they would stop charging you for the click...and you would have a log entry, anyway.
(This thread hosts an excellent and valuable topic.)
I helped a client set up an Adwords account and showed him how to put the code in his pages.
Thing is is that it worked great for a while and he made like 200.00 in his first 30 days with legitimate click thrus, as far as I can see anyway.
Anyhow, NOW when this guy goes to his web pages that have google ads in them, his IE Browser just crashes. BUT if he removes the Google Adwords from his pages, the browser will display the site correctly.
I have heard rumors that this is a tracking virus or IP tracking tool that keeps people from clicking too many times on their own adwords accounts. and thus crashes the browser.
Other rumors I have heard is that it is a virus from someone else, that does not allow you to view ANY Google Adwords, and if there is an Adword stream on a particular page, your IE browser will crash and close.
Anybody else seen this happen? Please let us know so we can let our readers know.
