Poss. explanation for multiple clicks in tracking script

Can this be it?

[ari11210]

It seems to me that all of these tracking scripts use onFocus to check for the click on the iFrame. Doesnt onfocus also get called when the user uses the tab key to the tab through the iframe? Can this be the reason for the multiple clicks people are seeing? This can register a 'hit' in the tracking script although the user never clicked on the iframe. Does this make sense to anyone else? Is there a better way to check for the click on the iframe (onClick, etc)?

Aaron

[annej]

I mentioned this elsewhere. Could it be a person on a slow dialup impatiently clicking away to get the page to come up?

[jonathanleger]

Tabbing into the iframe wont trigger a click, at least not as long as the script is checking window.status for content (which most of them posted here are). Yes, the log() function gets called, but that function verifies that window.status has text (which it wont if you tabbed into it), and so no click is recorded.

The OnClick event can't be captured in an iframe from an external domain (at least not the OnClick of any objects on the external page, which is what would be needed).

[jonathanleger]

Also, the latest version of the tracking script at:

[webmasterworld.com...]

filters repeated clicks of the same ad by the same visitor.

[richmondsteve]

jonathanleger wrote:

Also, the latest version of the tracking script at:
[webmasterworld.com...]

filters repeated clicks of the same ad by the same visitor.

I meant to mention something when that change was first posted, but apparently it slipped my mind. Any publisher who is using the click tracker in part to identify abuse and take action, may be best served by not implementing that change and handling the repeats in the reporting layer instead. The definition of abuse isn't important - it could be more than X clicks of the same ad from the same page, more than X clicks total from the same page or more than X clicks total across the site (on a per IP, per user or per user-session basis). On a client's site I'm currently triggering an IPTABLES (software based firewall in modern Linux kernels) rule to block port 80 access for 12 hours to any IP that clicks more than 4 ads while on the same page or more than 8 ads during a visit.

Just something to think about.

[jonathanleger]

Very smart idea, blocking visits from users abusing the ads. Would it be smarter, though, to put those "abusing" IPs in a database table and simply not write the AdSense code onto pages for those IPs over the next X hours? This would work only, of course, if you have a server-side script-driven site.

[richmondsteve]

jonathanleger, your solution is what I'm planning on implementing on my site and in general I think that's a better solution. My client has had problems with various system attacks and visitor abuse (some of which was definitely at the hands of competitors) and dictated that I handle potential AdSense abusers the way we handle the other incidents - at least for the time being. She decided to err on the side of treating all incidents that trigger the firewall rule as malicious, even though that might not be the case. So far (a week) there have only been 2 triggering incidents and neither resulted in an email or web inquiry after the block was purged 12 hours later.