Pere Tubert Juhé

<rant>

Like any other stat packages, AWStats is only valuable when it gives you information you actually do something with. With every piece of analytics data – whether or not it comes from AWStats, Google Analytics, Webtrends, Omniture, Mint, Woopra or whatever – you need to ask yourself “Knowing this information, what changes can I make to my web site to make it better?”

If the answer is “dunno”, then don’t waste your time and attention on it.

Not everyone has the same information needs, so what’s useful for you may be ‘dunno’ data for someone else. The art is in the detail (tips hat to Somerset Maugham), and if you are serious about using analytics to incrementally improve web site performance, you need that detail.

What’s that again? Use analytics to incrementally improve web site performance?

Explanation: Web analytics has the potential to close the feedback loop between cause and effect.

Typically a cause is something that’s on your web site, or a web-based process, that customers need to interact with in order to transact business with you. Web analytics can (for example) show that part of the process is faulty (such as high abandonment halfway through filling out a form). With this information, you redesign the form, abandonment rates drop, and sales go up.

One (previously unknown) cause of low sales is a poorly designed form. Analytics shows the form may be to blame. Web master responds by redesigning the form. The effect is that sales improve.

That is how analytics closes the feedback loop. But as a web site manager, you need to know (i) what analytics data is important, and (ii) what to do with it.

Problem is, every analytics package offers reams of ‘dunno’ data. So much so, that volume overrides common sense and buries the important stuff. So my goal over the next few posts is to go through AWStats in detail, so you can figure out whether any particular piece of analytics data is useful in your particular case. I’ll try to do the same with other popular analytics packages, as well.

One other point before I finish my rant – I haven’t yet seen an analytics package that provides all the information you’ll ever need. My preferred two-set is AWStats and Google Analytics, but there are good alternatives to these. And there are lots of situations where business success strongly correlates with one or two chunks of analytics data (improve the data, improve the business), so that time spent on understanding the full picture is time poorly spent. If you can just make sense of those few pieces of valuable data and adjust your web site or marketing strategy accordingly, then analytics will have closed the feedback loop and served you well.

</rant>

In my next post in this series, I will finally start discussing the AWStats dashboard (teaser – t’ing’s ain’t exactly what they seem).

If you would like to learn more about traffic generation,
search engine optimization and web analytics,
subscribe to my full feed RSS. My RSS feed is updated daily.