Dateline WordCamp 2007! Wow, so much to learn, and so many great geeks, but one of the most interesting presentations for me was from Information Architect/Usability Expert Liz Danzico of Happy Cog and Boxes and Arrows. She spoke to how great design for interaction is invisible to the user: you only notice design for function when it’s bad. When design works there’s no friction, and you can just go about doing what you intended. Good design anticipates what you need without making a big deal about it.
However, anticipating what people need isn’t the easiest thing in the world to do, and can have unanticipated pitfalls.
Liz told a very funny story about fancy new-fangled elevators in New York (apparently they have them at the New York Times) which have been designed to help you out. You tell them where you’re going, and you get directed to the right elevator bank and the elevator goes to the right floor. But what if you meet a co-worker on the way who tells you about a meeting elsewhere you should be going to? You have no way to redirect. So there’s been frustration about that. Fortunately, here in Silicon Valley, we usually have buildings with fewer elevators and fewer floors… so far anyway.
In some cases, if the landscape/interface is too hard to use as designed, people will modify it to suit their needs: look at all the dirt paths that show up, cutting across the corners, for instance on town squares and university quads. (Remember those well at Chapel Hill.)
Happy Cog has used some guiding principles in their usability analysis (and future redesign) of the admin interface for Word Press. One, as above, is to anticipate user needs, but in a more highly functional way. Anticipate peoples’ need to modify the landscape: allow users to customize the system to make it work for them. Especially for a system or task people use over and over again this makes so much sense.
For Word Press, Happy Cog constructed five user types, or personas, to help structure their usability research. These were based on expertise and familiarity with the application, as well as frequency of use. Then, they visited people in their own environments using Word Press. Apparently this involved lots of smoke and coffee in some cases 😉 That way, they got information through observation: how people actually use Word Press as opposed to how they say they use it. Liz did say that Word Press users tended to be a smart bunch, and this often coincided.
One very interesting bit she shared was the idea they tested of basing the admin functions on object-oriented nouns rather than verbs. Many new interfaces, for Vox, Tumblr, and iPhone, base functions in the interface on images that represent nouns, rather than verbs, as the Word Press admin interface does (Write, Manage). However, when she tried this model with Word Press users, by using card stacks and other methods, they couldn’t wrap their heads around it, it just didn’t make sense. The activity of writing seems to be the compelling concept, not the concept of a collection of things—posts or pages.
They took a serious look at the log in page, and found that most of the main panel contained content that people almost never use. So it’s likely to go. (This kind of analysis is related to the ideas Robert Hoekman discussed the previous day, more on that in a future post.) She previewed a few wireframes of design changes that will be coming in the 2.4 version of Word Press. Looks like it will be more widget-based (Netvibes, etc.) which would be wonderful.
Some more guiding principles:
- Try to BE the user of the software you’re developing.
- Don’t surprise people.
- Show, don’t tell.
- Don’t be brief at the cost of clarity. (Reminds me of that oft-quoted principle from Einstein, that things should be as simple as possible and no simpler.)
- Findability, consistency
- But most of all: transparent design, unself-consciously present.
Really a fabulous presentation, gave me lots to think about.