Aussie Web Analyst

My experiences in the past make me suspect that many people in the business world view web analytics as 100% science. That you input the data at one end and via some mathematic equation or formula, out the other end rolls reports, insights and recommendations. And that if the answer is not immediately obvious, all that is required is more data and/or a more complicated equation.

Clearly I disagree with this line of thought. I dislike the assumption that the web analyst (i.e. me) brings nothing to the party and that their their role has been reduced to simply extracting data from various systems and then mass producing reports and presentations. But more importantly, it leads to the belief that web analytics is easy and that all you have to do is look at a few reports, a few numbers to get to a solution.

I find web analytics very different from this. I do have core reports and core metrics that I will look at for every website to get a feel for what is going on. But when I see numbers that are off, I can’t tell you the answer straight away. Instead it is a trigger to look at more reports, and in multiple date and page configurations in order to understand what is going on be able to provide some recommended actions/solutions. Being able to recognise when a number is off, that is an art. Knowing which reports to look at next in order to understand what is going on and what data/page configurations to try with those reports, another art. I do not believe there is or will ever be a rule book which can tell you how to do this and as such, it is not a science.

So how do you learn this art? Is it possible to learn it? Yes, of course you can. It comes with experience, playing constantly with reports and data and thinking a lot about web analytics and reading a lot of articles about web analytics. Of course reading Occum’s Razor is the most important thing you can do… It helps to have numerical reasoning skills. I don’t think statistic skills are essential beyond the basics such as knowing there is a bell curve and alternative ways of defining an average but am sure that these skills would be useful. Creative thinking is very important too, you need to be able to conceive of potential causes of the problem before you can look for them. After that, a lot of it just comes down to developing a feel for the numbers and a good sense of where the answers may hide.

Tags: Web Analytics