Analytics

An example of the Google Analytics dashboardOnce you’ve carried out the various methods described on this website to optimise your website and linking strategy for SEO, you’ll need to ascertain what effect your efforts are having by analysing data about your website visitors. This will allow you to see how effective your measures have been and fine tune your content and linking methods so you can see how changes improve (or worsen) your rankings.

The sort of information you’ll need to gather includes the numbers, duration, location and return frequency of visitors, as well as how they got there and where they were when they left your website.

Choosing an analytics tool

Choosing an analytics tool is a key decision – there are many on the market, with varying levels of functionality at a range of different prices. Many smaller website owners can attain a good level of analysis using of the many free tools on the market (perhaps the most notable is Google Analytics), whereas larger companies wishing to track many pages and tied in marketing campaigns may find they need a more complex commercial system.

Different types of analytics tools

Log analysis tools (for example: Sawmill, AW Stats, Web Analyzer Pro) read web server log files and create reports based on them. To use these sorts of tools, you must have access to your web server log files and often need to install software on the web server. These tools can be rather slow, tricky to set up and limited with the data produced, but have the advantage that all requests will be logged, so they’re generally more accurate and include media files.

Script based tools (for example: Google Analytics, Yahoo web analytics) involve a javascript snippet being placed on all web pages (normally placed on an omnipresent master page or include) and when each page loads, the script analyses data based on the client side environment variables. These tools have the advantage of being lightweight, easy to install and can provide quite detailed analysis, but since they rely on javascript and can only be applied to pages containing html, they aren’t entirely accurate and don’t necessarily record media file related stats (although they can be setup to record links to any file type).

Combined tools (for example: Webtrends, Omniture) can get the best of both worlds by tracking both logs and script tags. The down side is expense and setup time.

Interpreting the data

The real point of collecting data is so that you can interpret it to find what works well and what doesn’t. The real value comes in finding being able to identify where there’s room for improvement and fine tuning to acheive better results.