Google image search keyword tracking

[FirstTimePokerPlayer]

I was wondering if I could find out with which keywords people find my site in Google images and found this article at yoast.com

I'm kinda clueless here. Do I just copy/paste that piece of code below my analytics tracking code?

[Michael Martinez]

You would have to download the scripts (you cannot use the example in the article) and modify the URLs (I think) to work with your site. I have never tried fiddling with Google Analytics scripts. I usually just look at my raw server logs, which provide better information.

[Mark]

I usually just look at my raw server logs, which provide better information.

Where do you find raw server logs for a site?

Mark

[FirstTimePokerPlayer]

Okay, well maybe I'd better not fiddle with them either then. I found the raw server logs and unzipped a 'gz' file (or something). When I tried to open it, some sort of DOS screen popped up and I got an error. Then I renamed it (removed the .com extension) and opened it in notepad and there it was. Is that normal, or should I have used some sort of special program to open these files?

I'm now trying to deceiver the log with the help of this page: studying web traffic.

On a side note: I found this piece "GET /favicon.ico HTTP/1.0" 404 etc." Why would a request for my favicon (which is there) result in a 404 error?

You would have to download the scripts (you cannot use the example in the article) and modify the URLs (I think) to work with your site. I have never tried fiddling with Google Analytics scripts. I usually just look at my raw server logs, which provide better information.

[Michael Martinez]

Where do you find raw server logs for a site?

Mark

It depends on how you're hosting your site. The two most popular platforms are Apache and Microsoft's IIS. I only have experience with Apache. Apache has to be told to collect log data.

I believe both platforms require that some sort of third-party reporting tool be installed. On the Apache side one of the most popular (and least flexible) packages is Webalizer. I wrote an article on SEO Theory about how to configure Webalizer last year. You can find it here. Some of those tips may also work for other packages like HTTP-Analyze.

Some Web hosting providers offer control panels (like WebMin, cPanel, and Plesk) that come with default configured statistics reporting tools (often it's Webalizer).

You have to see what statistics reporting your hosting provider makes available or installed for you.

Depending on what kind of hosting plan you use, you may be able to install your own reporting tool (you can buy licenses for better tools than the free ones in use).

[Michael Martinez]

Okay, well maybe I'd better not fiddle with them either then. I found the raw server logs and unzipped a 'gz' file (or something). When I tried to open it, some sort of DOS screen popped up and I got an error. Then I renamed it (removed the .com extension) and opened it in notepad and there it was. Is that normal, or should I have used some sort of special program to open these files?

I would recommend using Wordpad to look at the raw server logs. You can also purchase or maybe purchase an open source file to look at your server data.

I'm now trying to deceiver the log with the help of this page: studying web traffic.

On a side note: I found this piece "GET /favicon.ico HTTP/1.0" 404 etc." Why would a request for my favicon (which is there) result in a 404 error?

It's a case-sensitive file name on a UNIX/LINUX system (Apache Webserver in most cases). If the file is actually named "favicon.ICO" or "Favicon.ico" the GET would fail.

Otherwise, that might be indicative of a hard drive problem. Webservers tend to eat up hard drives like kids eat candy.