July 15, 2010
I recently went grocery shopping at a store I’d never been to before, Stew Leonard’s, and, no offense to Stew, I am unlikely to be a repeat customer. While there were some things about the store that were better than most grocery stores, the experience drove me nuts. Here’s why.
The store is laid out completely differently from standard grocery stores. Most stores, even unusual ones like Whole Foods or Trader Joe’s, have a nearly identical layout. One side is produce, frozen foods in the middle, meats in the back, dairy around the other side, standard aisles have bread, baking stuff, cans, cereals, drinks and snacks, etc. Go shopping enough, and you can generally find your way around any store in your sleep.
Stew Leonard’s decided to break the model. The store has no aisles and is linear – you just keep walking in one direction/flow and hit every single section of the store before you reach the end of the maze at the cashiers. One bonus is that they merchandise some things well and put logical items next to each other (burgers next to buns). But you can’t really go back if you missed something, you have no idea what’s coming up next, you can’t tell if you’ve seen all of a given class of item yet since different elements of every category keep popping up.
Sometimes that kind of a risk can pay off in a breakthrough new product design. Maybe people buy more items at Stew’s because things are set up differently. But the experience was very disorienting, the shop took twice as long as usual, and I couldn’t find a bunch of things so I still had to go to A&P afterwards – basically, the costs outweighed the benefits.
The obvious comparison here to our professional world is UI design. Breakthrough redesigns are always risky. They can produce better user experiences, but they can also confuse new visitors or less sophisticated users, and they risk an immediate reaction of “I can’t figure this site out, goodbye.”
UPDATE: Comments aren’t working today on the new blog, but my friend Pete Warden just emailed me a great comment about this post:
Your post reminded me of an incident at Apple that I wanted to share…One of the engineers was advocating for a UI change to an existing product. It clearly made the interface more elegant and logical, but our designer was pushing back hard. Finally the designer said “If you put that change in, I’m going to sneak into your house tonight and move all your furniture to different positions”. That analogy stuck with me; familiarity is what enables us to use a tool without having to stop and think, and so you need a really strong reason to change the structure of an interface.