There was a time, when I was in a team that was designing an important IT system for a multinational bank, the testers arranged for perfectly normal office workers from the bank to try the system out. This was long before the days of instant video. The software team watched from behind a two-way mirror. I would still heartily and unreservedly recommend this type of usability test.

It comes as a painful shock to developers when the software that they’ve created turns out to be unusable by ordinary humanity, for reasons that become immediately and painfully obvious. There are several natural emotional phases of denial before developers come to terms with the fact that their software isn’t right.

• Anger
  They are fools who are too old/stupid/impatient to use our divine software.
• Denial
  Surely anyone can see that we’ve provided a very usable software product.
• Helplessness
  What can one do? We are just following accepted standards within the industry. Surely everyone understands that?
• Bargaining/Guilt/blame-avoidance
  The whole product will have to be re-written, How are you going to explain that to the board/management, etc.? We should have had a better brief from the Business Analysts: The UI should have been spelled out in the design documents (Did Van Gogh have to be told how to paint sunflowers?)

For me, it was a time of quiet reminiscence, not tears. I’d been working for Xerox when they were trying to launch the Xerox Star, the pioneer of the visual interface that years later came to underly Windows and Apple’s OS. It was a superb and ambitious system that worked well, but had a fatal flaw. Although a Japanese version was eventually released, it only really worked when an American was using it. No effort had gone into the messy business of collations, character sets, currencies and other cultural issues. In fact, though it was a superb word processor, the Xerox Alto Star’s designers hadn’t appreciated the complexity of cultural necessities of worldwide business processes.

To create an international computer for the European market, we instead ended up buying in a plain but serviceable CP/M computer from a third party, putting it in a nice Xerox box, and selling it with reasonable success. It was cheap and supported the business processes of Xerox’s clients.

It may seem unfair to rummage so far back into history to jeer at the difficulties that the computer geek has in appreciating the requirements of the wider range of humanity. Actually, if anything, it has gotten worse. The first user interfaces had far better: ’transfer of training’, meaning that the more you learned, the easier it was to learn the next application because the user interface was more consistent.

Applications are nowadays poorer at accommodating disabilities, especially, visual, motor, interpretive, or coordination. Even the ‘case-sensitivity’ dinosaur of remote technological history, that we thought had long died out, keeps rearing its head and bellowing incoherently.

Each generation that provides technology is doomed to repeat the mistakes of the past if it forgets that technology was designed for humanity at large, in all its rich diversity, not the other way around.