The Log File Lowdown

Checking your website log files is an essential part of any online marketing campaign. However, do you know what your log files are? How to access them? Or more importantly, what do these files actually mean?

Your web log files, which are often provided by your web hosting company as a part of your hosting package, track visitors coming to your website. Your web hosting company should give you a login so that you can view these files. Each visitor to your site is recorded, either as a unique visitor or site, or as a returning visitor. ‘Unique visitors’, or ‘sites’, record visits by their IP address. If an IP address is shared by several people within a company, then those people will all be recorded once. A returning visitor means a visit from the same IP address has returned to the website within a given time frame. Knowing how many people visit your website is very important, as well as what your visitors are actually doing on your website.

The term ‘hits’, which is bandied around by a lot of people, really is quite a meaningless term. When your web browser requests a web page, the page is formed of several files or ‘hits’, which make up the page. The page could just have a couple of files on it, for example an image and a java script file, or it could have tens or hundreds of files, such as style sheets, lots of image files etc. Therefore, if you are looking at the number of hits as visitors to your website, and about 50 files make up your webpage, then your idea of how many people arrive at your website is vastly inflated. It is best to ignore this statistic. ‘Files’ are also unreliable, since a file is a ‘hit’ which has returned data. Not all hits return data, so again I would ignore this stat.

Now that we have seen which of the stats that are often used in web log file packages are meaningless, lets get to the ones which you really should be looking at.

Other than unique visitors, one of the most important stats is referring sites. This shows you where your visitors are actually coming from. They will show you if a visitor has arrived from a search engine, or from another website that carries a link to your site. You may often see some of your own web pages as referring sites too, which pages of your site have referred visitors to other pages. These stats are very helpful in knowing which search engines are sending the most traffic, and which external links are working the best for your website.

You can find out which keywords visitors used in a search engine query to find your website through the keywords list. This again is very useful as you can see which keywords work, and even find some keywords that people are using to find your site, which you hadn’t thought of.

Other important stats to look at are your page views, i.e. which pages people are viewing, as well as entry pages and exit pages. If a large proportion of visits are leaving the home page, or the page they arrive at the site on that this can give an indication that some changes may need to be made to try and retain more visitors on the website.

Other stats you can discover through checking your log files are how long visits last, visitor paths, e.g which pages are the most popular, which countries and towns visitors are from, which operating system visitors are using, eg Windows XP, which browser, e.g Internet Explorer or Firefox and finally how often the search engines visit, by User Agent.

It can take a while to get to grips with understanding the log files, but once you understand the basics, you can glean a whole host of information to help you improve your online marketing campaigns.

If your web hosting company does not provide you with any log file tracking or you find the tracking you have is very basic and doesn’t give you much useful information, there are other packages available which you can use by pasting code onto your web pages. These providers are listed in this month's resources.

Further Reading:
The Feedback Loop Gap - by Jim Sterne of Emetrics.org