Is it appropriate to rely on accidental discovery for non-vital or less commonly used UI features?

Question

Every function of an interface demands consideration concerning how users will be instructed on how to find and use it.

Certain features, like Android's push-shift-twice-for-caps-lock on the physical keyboard, aren't explicitly mentioned anywhere in the UI itself. (Maybe the instruction manual, but who reads those?) The feature is discovered almost assuredly by accident when a user tries to toggle Shift off, and finds it in Caps Lock mode. Because stumbling upon this particular feature doesn't really do any major damage, relying on accidental discovery seems effective enough in this scenario, and users will remember how to use it from that point on.

There are other ways of alerting a user to a feature, but I don't see where such instructions would be useful here. When a user pushes Shift, a message might inform them of the Caps Lock function, or the message could appear every time until the user finally uses it or tells the phone to stop reminding them. Either way, such messages take real estate, and I imagine they'd be much more annoying than helpful.

Answer 1

I would say "Yes", but would perhaps rephrase the question a bit to "accidental discovery of otherwise accessible features".

Let's take the iPhone for example:

• Yesterday I've found out that a friend of mine wasn't aware of the "horizontal-swipe" to show the delete button (for example messages in your inbox), although using the iPhone intensively for several months. Same goes for pressing "Edit" to mass-delete emails.
  Why?
  Since he could delete the message from within each one.
• Same thing happened with another person that wasn't aware that pressing the top of the screen scrolls all the way up. They used to simply scroll regularly multiple times till reaching the top.

In short - if the action is a "bonus" (easier/faster) for an otherwise evident functionality, it's fine.

If you have a feature that cannot be accessed any other way, I would hesitate to completely hide it without indication (although it might mean users will explicitly search for it :)). In this case a small "i" or similar icon shown by the action one or very few times (with the ability to "not show again") can be acceptable, but it depends on the specific application. Some desktop applications might even show "tips on startup", screens (which most ignore 95% of the time).

Answer 2

I'd take the opposite stance as Glen. Counter-intuitive controls are a bad idea, and you shouldn't have to memorize dozens of arcane shortcuts in order to be proficient.

That said, I fully support the "accidental but probable" discovery method, for the exact reasons you stated.

In cases where you have so many features that showing them all would be a horrible idea, I support using a modal interface such as the Ribbon bar that Microsoft put in Office 2007 and above. When the user selects a task, take away everything else and show only things pertaining to that task. There's no reason the user should have to hunt through endless menus. Those used to the menus hated the new version of Office, but almost everyone has come to appreciate just how effective it is.

Answer 3

To take the iPod as an example. I don't know anyone who read the instructions...I certainly didn't. It took me a while to understand that UP (in the circle UI) meant LEFT in the screen. And CENTER meant RIGHT.

Totally counter-intuitive to my brain at the time and I kept messing it up. These are the most used buttons in the whole device. But I did figure it out (as did the rest of the world) and got used to it.

Another example: Photoshop has tons of keyboard shortcuts. If you don't know them, you can still work. But knowing them makes you a "guru" and you are more loyal the product as a result. Its good to let power users find hidden gems.

Short answer: Yes, its appropriate.

Answer 4

The answer depends on whether we're talking about touch UI or mouse/keyboard. With mice, we have ability to provide lots of hints and visual clues based on hover. IMO, this makes accidental discovery more acceptable than with touch UIs.

Having said that, listen to Dan: it should be "accidental discovery of otherwise accessible features" -- never the only way to do something.

Also, make discovery easier by using consistent patterns. E.g. many mouse users will right-click on various parts of the UI to see if they can get a contextual menu. On iPhone, a common shortcut is to tap and hold, and I often find myself trying to do just that to see what goodies I can unearth.

Answer 5

Both ways could be good.

As Glen said , it can be good to let people just discover them and then be happy that they found it. I think that makes people more happy, to find it themselves as opposed to let the app point them at something. You get a proud feeling and love the designers smartness at the same time.

But, for functions that you still would like to push a bit you can do something in between. As you already said, a message could appear to point at a certain feature. If you spend some time making the message not being irritating, I think it can work. To use the shift/caps-lock example:

• only show the message the third time they use the shift key
• only when they used the shift earlier to type more then one letter in capitals
• only when they discovered other features themselves and thus may be interested in becoming an expert

Yes, it takes a bit more designing on forehand and computing/database entries. But that it already takes for just the showing something until they dismiss it.

Answer 6

I think that those features have to be hidden in a place where users will likely search the first minute they'll feel the need to. OR, if it's really to be accidental, should not freak out users, but rather be pleasantly surprised.

Funny how I figured out the shortcut for taking a picture with my iPhone just before iOS 4.0 when they took it out or changed it. :-)

Funny how from time to time I manage to take a screenshot of the iPhone, by accident, but I can't tell how exactly how I do it. But didn't bother to read the manual.

Answer 7

These hidden features enforce users engagement to the product. Even if they felt they made a mistake at first, this will boost their competence and make them more explorative. So yes, I would say they are not only appropriate but highly considered. But considered wisely. As you said, this only should apply to non-vital functions with remarkable undo features or/and high fault-tolerance.