The complexity and sheer level of data presented by analytics packages can often cause confusion and lead to mistakes. These mistakes are usually the result of misinterpretations and a lack clear goals. Before beginning any SEO campaigns it is important to set out some clear goals or KPI’s and have a defined plan of action. You need to know what you are measuring and what you are trying to achieve. Without this knowledge you will often get little from your analytics data.

Of course there are a couple of packages out there to track your websites performance, with Google analytics being the most popular. Yahoo’s offering has also started to get some attention while some higher end packages such as Omniture exist.

Some of the most common analytics mistakes can be found below:

Making Decisions Based On An Unrealistic Sample Size
Treat it like a scientific experiment. You wouldn’t base your conclusions on a sample size of two people would you? This would not give you nearly enough data to base a decision on. For example if you see that a keyword has 4 visits and converted 3 times you couldn’t make realistically make drastic changes to your campaign based on this sample set. You need to give it time.

I think Google Optimizer is particularly good at highlighting this point. In the reports section of Google Optimiser it will give you feedback about your testing. For example it will tell you that it is too early to make a decision if you have online had 30 visits. It will also provide you with recommendations on sample sizes. Worth a look.

Using Biased Data
There are many ways in which data bias can occur. Seasonal trending can be a factor, especially if you base your plans on data from high performing months. You can use Google insights to look for probable spikes and dips in traffic for certain keywords.

IP Exclusion Filters
David Whitehouse highlights in his post on analytics mistakes that many users forget to exclude their IP addresses from profiles. As a developer or SEO you are likely to visit the site many times. However the data recorded from your visits are likely to skew results slightly as you are not a genuine visitor. You can exclude your IP address in profile settings of Google Analytics.

Goal Conversion Setup
In order to track those goals and KPI’s effectively we often setup goals. These create a record of any visitor that performs a specific action. These often include making a purchase or enquiry. Although they are often quite straight forward to setup many fall into the trap of recording a goal step in the funnel twice. This is because people will add the goal page in to the funnel. This is not needed as you have set the destination page at the beginning of the setup process. That part is already recorded. The guys at ROI Revolution discuss it further.

Not Recognising That You Have Enough Data
We live in a fast changing environment and acting at the right moment can be crucial. Say for example you are planning for your christmas campaign and testing some keyword ranking techniques. You need to not only gather enough information to base your decision on but also implement it at the right time.

There are of course many more intricate mistakes we can make when using such packages. The data provided by such packages can be a little overwhelming. Your best option is find what you want to achieve and how you want to track that. Then add these essential items to the dashboard.

For some great advice on Google analytics you might want to check out the Advanced Web Metrics blog. Full of great info. Also if you think there are any things i should add to my list drop me a comment. : )