I found a web analytics book worth reading. Although I knew almost everything Avinash Kaushik wrote, I learned more from this book than any other one source.
Finding out what your web site users think by talking to them comes up many times in the book. This is very important, however smaller companies may not have the resources to do this offline research in a scientific manner. In those cases you could make contact with people who converted. Although this is a very special segment, it is better than not contacting anyone. As Avinash emphasizes, with web analytics most of the data won’t be perfect and we have to base our actions on what we have.
Comparing traffic trends to competitor’s on Alexa is another good idea, but the figures will not be accurate for small companies (and not completely accurate for big ones either as he notes).
MSN drives very little traffic to most of my clients, even if we have a no. 1 position in natural search and I tend to ignore it. Avinash points out one tip I really liked — the MSN search funnel. It shows you what people search for before and after your keyword. This can give you many insights and it is a lot of fun. I tried it with keywords for one of my B2B clients and nothing came up, as I expected. However it is worth checking out for more mainstream keywords.
Although I didn’t count all the graphs shown in the book from different web analytics packages, I have the impression that graphs from ClickTracks outnumber all the others combined. This is probably due to its ease of use and ability to show important information if you know what kind of labels you need to segment visitors. Ease of configuration should not be overlooked when choosing a web analytics package. As you dig deeper into your statistics you will find more and more things to check and configure and if you have to rely on your IT people it will slow you down.