March 30, 2011.
Recently I've been spending more time discussing website usability, and kept getting caught in the endless pit of differing opinions. You know, the one where two people have different opinions about something and the conversation falls through the floor while accelerating towards escape velocity. There are some great tools for resolving these kinds of discussions: user testing and A/B testing are two popular ones. Unfortunately, right now we don't have too much experience with user testing and we're still gaining confidence in our internal analytics, so I turned towards the literature.
Having read not a damn thing on the topic, I somehow ran into "Don't Make Me Think" by Steve Krug, which seemed like a reasonable place to start. I asked a coworker who studied HCI about it and he suggested that it would probably be too basic to be interesting. Unfortunately, I had already bought it at that point. More unfortunately, he was right.
On the positive side of things, the book confirmed much of what it's taken years of questionable (and in a discussion, inadmissible) intuition to arrive at. It does a fantastic job of creating memorable phrases, in particular:
We don't figure out how things work. We muddle through.
jives with my experiences. (A few days ago at work four developers were discussing approaches to distinguishing between on-site and off-site links, and it turned out that the site actually did have an existing convention, but three of the developers–all who had been working on the site for the past six months–hadn't picked up on it.)
Krug's discussion on navigation had a few interesting tidbits as well, in particular emphasizing the need to make it obvious to users to be able to know, at all times, where the hell they are on your site. In redoing my blog I dropped all indications of what storylist a user was looking at, which combined with the stateless navigation to become a rather felonious usability crime. So, I'll never claim this book had no effect on me: it did make me rethink this blog's navigation.
Helpful navigation suggestions aside, a twitch of annoyance grew into an enduring complaint as I read further: there is no evidence to support any of the claims. One could easily argue that Krug never intended to make a book which could win in a firefight, but finishing the book I'm no better to authoritatively argue usability points than I was when my iPad first put its grubby digital paws on it. Perhaps, my fault for unreasonable expectations of such a short book (216 pages), but I can't help but feel a bit disappointed.
One usability book later, I'm still just another jackass with an opinion.