I am struggling with a project and am reminded of a lecture at university that plotted two applications with differing rates of 'learnability' and 'expert usability' against each other:

a) Harder to learn, but faster to use once acquainted.
b) Easy to learn but cumbersome to use once acquainted.

I always thought good systems should not treat learnability and 'expert usability' as mutually exclusive and both went hand in hand. But it isn't always the case; the application I am working on necessitates the sacrifice of some learning functions so that it is faster for the user in the long term.

What has your experience been? To what degree have expert usability and learnability been a factor on the projects you have worked on?

http://www.chess.com/article/view/spectacular-queen-sacrifice strikes me as relevant.


Similar to your experience, I’ve found that learnability and expert efficiency are as often positively correlated as negatively correlated. As an example of positive correlation, having too many simple windows tends to hurt both: experts spend too much time navigating rather than “doing,” and novices get lost in the maze of interconnections.

Often learnability and expert efficiency are simply unrelated, and you can attend to the experts without significant detriment to your novices. The main barrier I see in design processes is simply convincing or reminding stakeholders that expert performance is important and requires special attention.

How you do that depends on what you mean by “experts”:

Experts at the task (who use your app daily) benefit when you:

  • Support shortcuts such as accelerators, toolbars, drag-and-drop, context menus, and double-clicking.

  • Minimize navigation and maximize awareness by designing a few large complex windows, rather than many small simple windows.

  • Use terse to-the-point text and labels in each window.

  • Use proportional design: easy to access commonly needed things, harder to access rarely needed things.

  • Have high internal consistency.

Experts on the system platform (e.g., Windows vs. Mac) benefit when you:

  • Have high external consistency, following platform standards and conventions.

  • Support the usual GUI capabilities, such as multi-tasking, and maximizing, minimizing, and resizing windows.

  • Avoid inventing new controls.

  • Provide Help.

Experts on a legacy system benefit when you:

  • Design for user performance superior to the legacy system.

  • Provide equivalents of the time-saving features in the legacy system.

  • Remove inefficiencies carried over from the legacy system.

  • Support a smooth transition from the legacy, such as by supporting the same keyboard shortcuts or having help specifically tailored for migrating users.

Domain experts, who have exceptional and subtle understanding of the operation you’re supporting, benefit when you:

  • Use an object-centered UI structure.

  • Provide on the periphery use-for-anything general-purpose commands.

  • Make these commands potentially powerful (e.g., by supporting multiple selection, scripting, and saving).

  • Provide overrides, such as by using sanity checks rather than hard validation.

Details at What Have You Done for Your Power Users Lately?

In cases where learnability and expert efficiency are negatively correlated, the solution is to apply a human-systems integration approach and figure the total cost (preferably mathematically) of task performance, training and support, manpower, and personnel selection for a given design and choose the least costly/most profitable one. As a rule of thumb, the more users use the app, the more it makes sense to pay the user training/learning/selection overhead to bias the design towards experts.

  • Thankyou so much Michael; the last paragraph is especially pertinent because it reinforces my position... – dave Nov 2 '10 at 13:38
  • Designing for different groups (novices and perpetual intermediates) is more work. Even if many members of the first group will grow into the second group, they need different things, and providing different things (that is, MORE things) costs more. Quality user experiences, including usable and efficient ones, almost always have a higher development cost that reduces the total cost of ownership or TCO. – JeromeR Nov 8 '10 at 2:47

Ideally, products would have no learning curve: users would walk up to them for the very first time and achieve instant mastery. In practice, all applications and services, no matter how simple, will display a learning curve.

source: http://www.asktog.com/basics/firstPrinciples.html#learnability

Free PDF: Usability or Learnability: What are we testing? http://sandrinebalbo.com/wp-content/uploads/sneu_v20071028_finalversion.pdf


Interfaces for the real experts tend to strip away all the niceties, so you get command line interfaces with say UNIX...

And the interface on a Forumla 1 car looks rather different from a normal car:


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy