I'm not speaking rhetorically here. I mean, if you're a web analyst, and you set out to make sense of the web site behavioral data that we all collect in web analytics applications like Omniture SiteCatalyst and Webtrends and Google Analytics, how do you frame a typical analysis project?
When I do web analysis I often turn to functionalism.
The concept of functionalism was developed several years ago by Gary Angel, Semphonic's founder and CTO. In essence, it is a framework for doing web site behavioral analysis. First, web pages are categorized into types depending on their function, and then specific KPIs are applied to each page type. I'll explain more later.
Although functionalism was Gary's "baby," over the past few years it has become the common language we all speak at Semphonic. It is now one of our web analytics consulting product offerings, and at any given time we've got someone in the office doing functional analysis. We're spreading the word through training, too; I recently taught a class on functionalism at the X Change 2009 Think Tank.
Since I believe functionalism is so useful for web analysts, I have decided to write my own blog post series on it.
I will begin by telling you how I got started with functional analysis and explaining why I think it's worth the effort. In subsequent posts I'll lay out the basics of functionalism, give a number of examples, and describe how you can run a functional analysis project of your own. I'll file everything in the Functionalism category on this blog.
As background material I recommend reading Semphonic's white paper on functionalism, as well as Gary's own blog series on the topic:
- Functionalism White Paper (pdf)
- Gary Angel's Blog: Functionalism and Web Analytics Part I
- Gary Angel's Blog: Functionalism and Web Analytics Part II
- Gary Angel's Blog: Functionalism and Web Analytics Part III
- Gary Angel's Blog: Functionalism and Web Analytics Part IV