There has been a lot of discussion about the practice of progressive disclosure in reducing complexity for users both in terms of content and interactions. On the other hand progressive reduction has also become popular with the trend of simplifying UI designs. I was wondering whether both approaches have been combined in a sensible way help improve the user experience not just for content (since this is where most of the posts and examples have been focusing on), but also for user interaction. For example, if I usually take to clicks to perform an action the application might create a shortcut for me instead. Or if a particular menu option has not been used then it gets moved further down the list. The idea of trying to build a one-size-fits all application seems to just create headaches for UX designers, so why not take the opposite approach of adapting the application's functionality based on usage information?

This is not to say that the entire application needs to be fully adaptive, because then you lose the core design principles. But in areas where there are endless debates about how people actually think users would use something that even after testing doesn't provide a definitive answer, I think a combination of both approaches would solve the problem of having to make a compromise or only design for one group of user. Are there any good examples where both approaches are used to balance the complexity of user preferences/behaviour?

2 Answers 2


Well, any standard menu is a classic example of progressive disclosure, where first you see just the title "File" and then you see the options it contains. "Progressive Reduction" is apparently a new name for what used to be called "Adaptive UI" or just "Personalization". The classic example here is MS Office 2003, where they did just what you described - if a menu item weren't used, it would move down the list. AFAIK this approach failed miserably, users were never able to predict where a specific rarely-used option will be located in the menu, and also why their often-used options tend to move around. MS dropped the idea in Office 2007 so apparently they arrived at the same conclusion.

  • Would you expect that users have different expectations for desktop versus web applications? Although I suspect that it is a bit more blurred these days with more desktop application moving to the web.
    – Michael Lai
    Jun 4, 2014 at 7:11
  • The idea that someone failed miserably implementing something, does not make it wrong :)
    – Velkommen
    Jun 12, 2014 at 18:02
  • @Bluewater And where did I say that it's wrong? Jun 12, 2014 at 18:06

Progressive reduction does not sound like it plays well with muscle memory and for this reason I would avoid it.

Other methods to deal with complexity are to use a 'natural language' interface, typing or saying commands instead of having a specific word or icon linked to an action you can then attach multiple words to an action. As opposed to an old command line interface it isn't a matter of you knowing how to word an action in a way the computer understands, but to express an action and have the computer understand you or perhaps prompt if it is unsure. Expect to see this area expand as apple siri, MS cortana and initiatives such as ibm watson develop.

A method we use at work for a complicated programme are to setup a series of default user types. Much like the workspaces in adobe products it reduces a larger set of controls, to those used just for one particular task or role. These can then be further customised should the default not fit your specific needs. Additional elements can be found via a search and pinned, or using a master page of all commands.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.