Googlebot Knows How to Use Search Engines

Recently I noticed that Google's "Googlebot" was using this website's search engine. This is not exactly surprising since a website's search engine may be able to find pages that are not well linked to by others. Nevertheless, it is interesting to see how the Googlebot performs its task. Here are some entries from the log (truncated so that its easier to fit into the page):


Google AdSense Account has Been Disabled

Some visitors may have noticed that the Google avertisements on this website have disappeared. This is due to Google deciding to disable my account because they detected "invalid clicks." Last month I noticed an unusual spike in AdSense earnings and expected them to drop again. Well, the earnings didn't drop at all. It looks as if these earnings were due to people deliberately clicking on advertisements that they have no interest in. Whilst I did suspect that this was happening (all the clicks were on the top banner, and there were too many of them), I am personally not responsible for them; nor did I know of any way to counter this. Naturally, the Google Adsense "invalid click" detection algorithms picked this up, and my account was promptly disabled without warning. I did appeal, citing that I am not responsible for these clicks, but I received a standard email reply stating that the account would not be reinstated.


Site-Wide Search Now Available

This website now features a search function which can be found in the left column. I have opted to use Google custom search since I have found that Google's search engine often is better at locating information than a website's own search engine. Google's search technology is simply far superior to many of the other search engines provided with various Content Management Systems (CMS). Adding search functionality was relatively easy, although it did require that I create a custom page type for the search results page. The reason for using a custom page type is because the HTML insert functionality in the Silverstripe CMS won't embed the Google search results page code properly; it performs some sort of operation after closing the HTML edit window that destroys it.


Google Analytics Misses Things

I have suspected that Google Analytics hides things it doesn't understand for a few weeks now. Whilst I am not certain about that, it definitely misses some things, and gets others wrong. Let us examine what it misses first. Some old browsers register only the first page visited, and the rest are not counted. This results in Google Analytics marking their visits as "bounces," when these visitors did, in fact, view multiple pages. For example, Mozilla 4 and lower appear to cause this. Other browsers do not even register at all, resulting in uncounted visits. These details were obtained by comparing the results from Google Analytics, to this site's server logs. 


Google Polska?

Google Webmasters Tools has a statistics page that tells you what the top searches were that your site was listed in, and what your position was in the search. To my surprise, this site's top ranking was in 3rd place for the search "Google Polska." How did that happen? I am not Polish; I am a Dutch Kiwi. Whilst someone did once ask if I had any Polish ancestry, there is no way that the Googlebot could see my blond hair and ask that question. Not to mention that the Poles are not the only blond-haired Europeans in this world.


Why use XHTML at all?

I decided to add a Google Adsense bar discretely at the bottom of the page, and once again, I had trouble with Firefox. Internet Explorer displayed the page just fine, but Firefox added the new advertisement below the existing ads at the top of the page.


Google has added this site to their index...

... but just the login page.


XHTML, Google Adsense, and Firefox 2

EDIT (2008/06/26): I have since encountered more problems with XHTML and Adsense, and have come up with a much simpler, and more effective, solution.