Mis-Understanding Functionalism

    September 6, 2006

I recently came across Matt Jabobs’ blog on Functionalism and I figured it was worth taking time out from my ongoing series to comment on it since I think it shows (unintentionally) just how much better Functional Analysis is than the schemes that preceeded it.

The tone of the post is a bit odd – and I can best describe it as “I don’t like your idea and besides, I had it first!”

Here is a quick summary of Matt’s “historical “scheme:

“A scoring method was created to tabulate the performance of each page based on how well each page served – and was intended to serve – a variety of functions (e.g. explanatory, navigational, information collection and sales functions, among others).

A portion of the scoring method was based on heuristics to analyze the function categories noted above. We also extended the heuristics to score how well each page addressed the following three value/benefit categories: Functional, Economic and Psychological

Another portion of the scoring method was based on historical web analytics data.”

There are several things that seem problematic with this approach and readily distinguish it from Functionalism. Let’s start with the fact that Functionalism doesn’t have any heuristics. That seems like a pretty big difference to me. There are no rules of thumb in Functionalism for scoring pages. There are no subjective design qualifications. Matt is right to be skeptical of the degree to which Functionalism reduces the need for brilliant practitioners and is worried about the amount of effort it requires. The system he believes it mirrors would be rife with just those issues.

Functionalism is all about web measurement and it is rooted in actual web measurement practice. True, the analyst has to classify pages. But he doesn’t score them, he measures them using specific KPIs. It’s a huge difference. Functionalism actually has no method whatsoever for scoring pages. It has techniques for measuring them correctly.

And the page classification system doesn’t require the analyst to decide whether a page is attempting address “psychological” or “economic” concerns. In fact, the classification system is designed to be extraordinarily easy to use – I think any reasonable observer would expect virtually everyone involved with a web site to be able to easily and accurately classify the pages. And the substantive judgments about page performance are entirely contained in the web analytics KPIs – which are rich, detailed, specific to function, and carefully thought out and documented.

Matt does mention that a second portion of the scoring method is based on historical web analytics data. That gets pretty short shrift if you ask me. One sentence? And I’d be interested to see the KPIs that measure how well a page is performing its “psychological” function. But things are even worse than this, because the general groupings Matt suggests are far too broad to be useful. Categories like “sales” are inherently much too broad to be interesting. Part of the goal of any classification scheme has to be to implicitly segment pages so that their distance from the goals you’re measuring is relatively constant. Where this isn’t true, your measurement will give fantastically misleading results. All you’ll really be measuring is how close to conversion your pages actually lie in a visitor’s path. This is exactly the kind of measurement we’ve seen far too many times – a complicated scheme to produce bad results that everyone immediately ignores.

The bottom line is this – the system Matt is describing isn’t better than Functionalism, it isn’t Functionalism and it isn’t a precursor to Functionalism. It’s just a complicated system for scoring pages based on some unarticulated set of rules about design and navigation coupled with some basic web measurement statistics.

Now obviously, something like Functionalism doesn’t spring full-born into the world. There are plenty of intellectual predecessors. Nor do I believe that Functionalism is the be all and end all of web analytics. There are plenty of other good techniques – many of which we ourselves also use. I hope that readers of the Blog series on Functionalism will appreciated how much sweat equity and actual practice goes into how we do and talk about web analytics. I can un-blushingly claim that I think we are more honest, less salesy and more useful than the overwhelming majority of blogs and white papers out there. The proof of that, of course, is in the pudding – but I’m more than willing to let the whitepaper speak for itself on that account!


Add to Del.icio.us | Digg | Yahoo! My Web | Furl

Bookmark WebProNews:

Gary Angel is the author of the “SEMAngel blog – Web Analytics and Search Engine Marketing practices and perspectives from a 10-year experienced guru.