Expert Voices: Peter Morville on Why Information Architecture Matters - ' Next Page' (
Page 4 of 4 )
: Making Web sites Count">
Making Web sites Count
Are methods emerging for tracking the value of improvements to the information on a Web site thanks to intelligent architecture?
There are a lot of different ways to try to track value. None of them are perfect, and attempting to apply quantitative measures to information architecture improvements is really difficult.
I'll tell you a positive story, and then I'll talk about all the difficulties. I recently worked with the National Cancer Institute on a redesign of their
cancer.gov Web site. We were doing a fairly shallow redesign, not changing any of the content. We weren't even changing the deep structure for the document level. We really did a visual redesign and a restructuring of the first few levels of the site. We were able to make some really dramatic improvements in the ability of visitors to the site to find what they were looking for, and that resulted in the National Cancer Institute winning quite a number of awards for the site. More important, it moved right up to the top of the
American Customer Satisfaction Index for e-government. So we were able to really get some positive improvement in the site and some metrics from a third party to show that.
The problem is that most of my clients focus on that customer satisfaction index as an important metric, when, in fact, it's a holistic measurement. It can't tell you whether it was the visual design that made the difference, or whether it was the content. The metric tracks what's important: the overall user experience. So it's very difficult to isolate the information architecture from the other elements of the user experience. You could certainly do that in a research lab, but in the real world all of these factors work together. It's quite possible to do a beautiful information architecture redesign but completely destroy the experience by messing up the graphic side of things.
Another thing that's interesting about metrics is how easy it is to oversimplify. Some of the metrics that a lot of analyst firms put out depend on managers thinking, "If we could just reduce the number of clicks—the amount of time it takes somebody searching to find what they're looking for—then in an intranet setting, for instance, we could save millions of dollars a year of lost employee time."
There's some truth to that. A poorly designed site that's wasting employees' time means you are losing money. But I would also argue that the searching and browsing experience is one of the most valuable ways knowledge workers learn, and it's not a straight line from point A to point B. It's an iterative and interactive process where we might enter a couple of keywords, do a search, find a few documents, and learn that we should actually be searching for something a little bit differently.
So we don't necessarily want to optimize for reduced numbers of clicks and speed if that means we're reducing our employees' learning opportunities.
Is there an overarching goal to information architecture?
That's one of those big-picture philosophical questions. I guess part of what excites me about the Web, and makes me feel good about doing information architecture, is that it helps people get better at sharing information. People have used the metaphor that in creating the Internet we're creating a central nervous system for the planet, wiring ourselves together and making the ways that we're able to communicate and share information much more fluid. Of course, in order to feel good about that, you have to have a certain amount of optimism that if we enable people to communicate and share information better, they'll ultimately make better decisions.
Expert Voices 
