I’ve tried split A/B testing banner ads, emails and even postcards, but I’ve not tried A/B splits when making a design decision as this article from usability news suggests. The concept of split testing a decision decision particular fits in well with the concept of an evolving site redesign rather than a radical redesign.

I’m an old hand at reading site stats going back to when I crunched log files with analog, then parsed them with Clicktraks to adding javascript tags from urchin (now Google Analytics). But, the real question comes down to what do you with the site stats once you have them. Or, how do the site stats change what you do. To work on that, I started reading Web Analytics: An Hour a Day by Avinash Kaushik this weekend to help me learn more about who is visiting my website.

And then, today, Mindy McAdams makes a great post how she is using Feedburner stats about her blog to know who to design to. As in what monitor resolution and what browser she should optimize her blog design to. Her blog stats only reflect the users that visit her blog (not the web as a whole) and she does have a more technical audience, but it is interesting to see how a third of her visitors are now using Firefox 2.0.