This doesn't happen with all searches, of course. Here's one where it does. Search for "oracle robust" WITH the quotes http://www.google.com/search?hl=en&lr=&ie=UTF-8&c2coff=1&q="oracle robust" The URL of the first result is http://www.google.com/url?sa=U&start=1&q=http://www.oracle.com/timesten&e=42 when you click on it, you go to the correct oracle page. Anyone know why they do that, or under what circumstances?
Let me clarify if others find that unclear. Usually, the results link straight to the destination. In this case, www.oracle.com/... is the first result. But instead of linking to that, Google links to themselves, and has a 302 redirect to the destination. HTTP/1.1 302 Found Cache-Control: private Location: http://www.oracle.com/timesten Content-Type: text/html Server: GWS/2.1 Content-Length: 227 Date: Sat, 18 Mar 2006 15:28:11 GMT This is really strange. What's even more strange is that all the results on the page lead to Google and 302 redirect to the proper destination.
Google is click tracking at times. Why is unclear. Here are some previous posts on the subject. The last one is probably most informative. http://forums.digitalpoint.com/showthread.php?p=193064 http://forums.digitalpoint.com/showthread.php?p=201314 http://forums.digitalpoint.com/showthread.php?p=219698 http://forums.digitalpoint.com/showthread.php?p=267576 http://forums.digitalpoint.com/showthread.php?p=648099
Actually, they click track ALL the time... it's just done with JavaScript usually. My guess is they do it without JavaScript sometimes just to get samples of data to compare against other things.
Isn't using Javascript to track clicks rather unrepresentative of the entire population that clicks on the links? I still know of people who use early generation browsers, and others who use extensions/mods to disable Javascript.
I think this is perhaps the reason for tracking clicks via the 302 redirection method. It gives them data to compare and adjust for the popularity of the term for non-JS users. Google using a 302 redirect is ironic, since if we webmasters use a 302 for click tracking, it can trigger a bug in Google) and is not a recommended method anymore. Google is finally attempting to fix it in this BigDaddy update... If turulillo is correct about logging out of Google to avoid JS tracking, it also gives them data for the users of Google as a whole. If the info given by gford about pulling data from toolbars and logged in accounts is correct for Google as well, it would support that.
More on google's URL tracking in this thread from a few days ago: http://forums.digitalpoint.com/showthread.php?t=66438