User testing is more influential than you might believe

okcancel.jpgRecently the HATT list had its annual debate on the “Click the OK” or “Click OK” issue. (Yes, user assistance developers actually argue over this stuff. And it’s “Click OK“, if you’re wondering.)

Did you know that the OK button had its origins in usability testing? During interface development for their pioneering Lisa software, Apple designers noticed that users were having problems. When the software required positive or negative confirmation from the user, a small window appeared with two buttons: Do It and Cancel. The testers noticed that many users wouldn’t click Do It, and some were visibly annoyed. Via Folklore.org:

The team noticed one user that was particularly flummoxed by the dialog box, who even seemed to be getting a bit angry. The moderator interrupted the test and asked him what the problem was. He replied, “I’m not a dolt, why is the software calling me a dolt?

Get it? The nature of the low-pitch monitors required the use of sans-serif fonts, and most were reading the “i” in “It” as a lowercase “l” as in “Losery dude who writes about buttons in his blog.” The Lisa interface designers made the change to OK, it performed better in usability testing, and the rest is history.

What is even more intriguing is that this simple change reflected a new way of thinking about human-machine interaction. HOTB goes on to suggest:

“Do it! is the same as previous versions of Enter or Execute. It’s commanding the machine to do something. OK is acquiescing to the machine, forming a partnership. In the end, the simple OK button may have contributed to the success of the Macintosh. It changed the relationship between person and computer, away from the master and slave mentality toward a friendlier world where the computer is a partner.

The moral? Any interface between human and machine – be it mechanical, electrical, or graphical – should be tested as early as possible during development. Make the case to those who fund you.

2 Responses to “User testing is more influential than you might believe”


  • It certainly was a long debate on HATT, though with many side discussions along the way. What I find most annoying with OK/Cancel (not the toon) is the rampant misuse (or ubiquitous use) in inappropriate settings. I’ve seen it many times in cases where the buttons should have read Yes/No. But it’s now a familiar convention, list like the elipses at the end of some menu items. Most people don’t associate them with launching dialogs with additional action required. They pay them no mind. But they’re there, and some developer somewhere is smiling, knowing that their creation is continuing to be used.

  • I think this area is where Tech Comms really can push things forward.

    With an understanding of the task being worked on, the verbiage/tense/word choice can be crucial, and that level of ‘view’ of an application isn’t always available to the developer (who might be working on a subset of functionality which makes up part of an overall task).

Leave a Reply