Whenever I see someone using a fancy new software interface, such as at a bank, I love to ask if they like it. I think without exception, the answer has always been no. The answer to the follow-up question, why, is invariably the new software is too slow.

Watching them use the new software, I can see why. Often they switch between multiple input methods, where before, a keyboard was all that was required. But is there more to it than just bad design?

Thinking of my personal history, I remember working at McDonald’s. I actually worked at two different restaurants that had different tills (POS terminals). One was an old style - a big array of buttons - and the other was a new style - a hierarchical touch screen. The old one was not as pretty, but between the two, I much preferred the old style because it was so much faster.

In my opinion, interfaces are often designed for the occasional user at the expense of the frequent user. Rick Omanson from User Centric gives a great talk Heavily Used Interfaces: Designing for Efficiency about, quite obviously, creating very efficient interfaces for the frequent user.

If you won’t want to watch the entire webcast, the main points for improving efficiency are:

  • Make information needed for decisions visible
  • Use dense screens - if location is content
  • Keep controls control to content being acted upon
  • Rely on knowledge in the head
  • Design for scanning (highlighting, headers)
  • Avoid switching input device
  • Allow flexibility (forms instead of wizards)

In the end, all this discussion of efficient interfaces leaves me wondering. Were these considerations in the designers mind when they designed these older systems, or were the system limitations determining the design?