Counting Clickthroughs Correctly

[Squiggy2]

I've got a list of articles that users can click on. Clicking on an article takes the user to a page (countlink.cfm) that records their click and then takes them to the article they wanted.

When I look at which article links are being clicked, it seems that many articles are being clicked on that we didn't think would get any action at all. Is it possible that Googlebot and the rest of the spiders are causing the clickthrough numbers to be skewed?

[Fernis]

Not only is it possible that a search robot follows a link, it's almost certain.

A good place to start would be http://www.google.com/support/webmasters/bin/answer.py?answer=182072

I recommend using robots.txt

Another safety mechanism would be using Javascript.

While I don't know for sure, I'd imagine that most search robots only "follow" links on a page, and do not emulate fully javascript enabled browsers.

If you assume a bot doesn't "mouseclick" a link as a user would, you could use javascript's onClick event to throw users to the page you want, instead of having the target page in the HREF attribute directly. That could confuse the bot hopefully 🙂

-Fernis

[Squiggy2]

Good idea. I just implemented your suggestion, and we'll see how it goes from here. Thanks!