Inspired by "Is coffee a good excuse for a slow application start-up time?" I tried to look for examples where having higher speeds could be bad for the user experience but I couldn't find any.

Are there any scenarios where faster {startup times, interactions, animations, etc} can harm the user experience? Do we always have to optimize looking for the highest speed?

  • 11
    Just a quick example (hence comment), computer startup times are getting very fast, sometimes I wish I had long enough to read the fact that I need to press esc or del or F8 or ... to get to the BIOS. Nov 19, 2012 at 21:14
  • 1
    Nice point George. I need to get to the BIOS so rarely that I never remember the correct key. It usually takes me 2 or 3 re-starts, each time hitting as many function keys as I can, to get into the BIOS settings. I still never know which key actually did it.
    – Bill Dagg
    Nov 21, 2012 at 20:20
  • 2
    My suggestion would be: get a Mac ;-)
    – edgarator
    Nov 27, 2012 at 11:41

6 Answers 6


I know some people advocate introducing a deliberate delay after certain actions in order to make it feel like some sort of processing has happened. Consider a 'save' function for example. Even if the system is able to save changes almost instantly without delay, introducing an artificial delay may well instil confidence in the user that their changes really have been saved. This is based on the idea that people are accustomed to computers "taking a while to think" when performing certain actions.

Here's a real-world example of how Coinstar used this idea: Adding delays to increase perceived value: does it work?

Other than this, I'm sure we all introduce subtle delays in one way or another. One example is a dropdown menu that opens and closes based on mouse hover. It is often necessary to introduce a delay so that the menu doesn't open immediately when the mouse pointer passes over it on the way to something else, and equally so that it doesn't close immediately if the mouse pointer accidentally moves away momentarily.

  • 13
    The Coinstar blurb in the link above is a great example: “Coinstar is a great example of this. The machine is able to calculate the total change deposited almost instantly. Yet, during testing the company learned that consumers did not trust the machines. Customers though it was impossible for a machine to count change accurately at such a high rate. Faced with the issues of trust and preconceived expectations of necessary effort, the company began to rework the user experience. (continued...) Nov 19, 2012 at 13:15
  • 7
    (...continued) The solution was fairly simple. The machine still counted at the same pace but displayed the results at a significantly slower rate. In fact, the sound of change working the way through the machine is just a recording that is played through a speaker. Altering the user experience to match expectations created trust and met the customers expectation of the necessary effort to complete the task.” Nov 19, 2012 at 13:16
  • 4
    Hm the mouse-over delay is different from most delays; it's to improve usability not to pander. As long as there's visual feedback that "yes, I'm done saving" I don't think a delay is necessary.
    – Ben Brocka
    Nov 19, 2012 at 13:18
  • 1
    If the Save action lacks feedback, deliberately making it slower is not a solution to the lack of feedback. We can give such feedback by greying the Save button after saving. The Mac has the dot in the red light disappearing, and the icon getting brighter. Apr 27, 2013 at 18:48

How about Pacman? Surely with computing power these days, you could speed the ghosts and yellow chomper to Superman-like speeds, but then the game would only be fun for those who can click faster than a speeding bullet.

The general rule is that any time you are simulating a graphical interaction or sequence, you must know your audience and render that interaction in a speed optimal for them.

  • Upvoted. I think the Pacman example is a bit of a stretch, bit the general rule is well stated.
    – Freiheit
    Nov 19, 2012 at 17:09
  • 6
    In fact most video games, from the most basic Atari game clones to Mass Effect, have to make sure that they can be played properly despite the pace of Moore's Law. Many older games were designed to run on a particular generation of architecture, and accounted for rendering in their game sepeed delays, so if you loaded them up on a modern PC they'd be unplayably fast. Nowadays most such games split the rendering off to another thread, and do time-based game logic updates rather than delay-based, allowing the computer to make the rendering engine cry without affecting play speed.
    – KeithS
    Nov 19, 2012 at 17:18
  • 6
    I think the reason for the downvote is that a really fast Pacman is more of a bad game mechanic rather than a bad user experience.
    – Freiheit
    Nov 19, 2012 at 20:24
  • 3
    Old games that assumed a clock speed were such a common problem that they hardwired clock-swapping (turbo) buttons into PCs to allow users to work around the issue. See pcguide.com/ref/case/switchTurbo-c.html Nov 19, 2012 at 20:35
  • 3
    @Briguy37 I have to admit, Pacman was the first that came to mind for me too. I installed the original Pacman on a Pentium II a long while ago and it was impossible to play because they didn't have a frame limiter. Frame limiters in a game is a bit different to a user interface, but I think you do have a valid point in that we shouldn't be implementing interfaces that are too fast for the user to handle. I think a good modern example of this is George Duckett's comment on the original question regarding BIOS load screens being so fast that you don't have a chance to see what button to press.
    – Nathan
    Nov 20, 2012 at 0:46

Actually, the "perceived time" is the important subjective UX factor.

But since UX and usability also embraces objective factors (such as effectiveness), slow performance can per definition decrease the UX/usability.

One example of such illogical "perceived time" tasks is the installation process of 5 floppy discs vs the installation with a CD-ROM.

The 10 minutes "active" installation process with 5 floppy discs and all that disk switching were actually perceived as faster (and thus preferred!) over the 5 minutes passive CD-ROM installation with nothing to do other that wait for the progress bar to finish...

The conclusion is, of course, not to slow down our application!
But we cannot judge the UX from the system performance alone. We need to evaluate how the system interacts with the person using it.

  • +1, great example - however: "The 10 minutes "active" installation process with 5 floppy discs .. were actually perceived as faster (and thus preferred!).." - I wonder if this still holds true in times of ubiquitous internet, when there's always something you could do in fron of your computer.
    – peterchen
    Nov 20, 2012 at 9:13
  • 1
    Liked the angle of the perceived time and the observation that it is actually a subjective topic. Nov 28, 2012 at 20:29
  • I wonder who would prefer putting 5 floppy disks than putting one CD‑ROM. Feb 16, 2013 at 12:11
  • Nicolas: The people who thought this was the fastest way to get the job done... Feb 17, 2013 at 15:07

I'm intrigued by the question. However, there is one thing I haven't seen mentioned. Very often, the most stunning looking graphical decisions are an annoyance to the audience that does not expect it to be that stunning.

For instance, when I first started with jQuery, I felt tempted to animate EVERYTHING I could find, menubars, dropdowns, loading gifs etc.

However, more recently I found that elderly, or impaired people, actually find menu's that "suddenly dissapear" quite annoying. They just don't expect them to fade out when they take the cursor of it, while trying to read whatever the cursor was blocking.

Now, this might seem off-topic. But I believe indeed that animation and loading speeds work the same way. If something moves quicker than I feel comfortable with, I'll refrain from using it too much ( And yes there are exceptions to those rules ).

I therefore think that Briguy37 is right. It's all about the intended audience.

Also, the windows 7 start menu pops up and hides instantly after a mouse click. That feels natural because you want something from it. However, a program minimized to the taskbar takes a while to popup. Why? Because it might be touched accidentally while reaching for something at the bottom of the screen. Immediately showing the program works confusing in this case.

Bottomline: Interesting question, fun topic, no answer. These decisions must be based on the target audience and the intended purpose the element or software.

Not the answer, just mine.

  • I'm sixteen years old (i.e. not elderly) and I hate menus that disappear when the cursor is moved off. Has this ever been considered good ux?
    – minseong
    Nov 9, 2017 at 17:29

I think that things should happen in the speed that the user expects them to execute.

In the 386-era, if something happened exactly after you typed or clicked, something would seem to have gone wrong, because almost everything that was done at those computers took a certain time, and users were used to it.

With a new computer, one user would expect things to happen faster, but only to those things that were taking more time than the user would require to perceive it. Example: the program could save the document faster, but if the user couldn't see a change (like a icon changing color, etc) he wouldn't trust that something really happened.


I think the key is the UI feedback. Generally, doing the action faster cannot harm. But giving UI feedback too fast can harm.

I have a few examples.

When I type some letters in Firefox' address bar, Firefox does show me suggestions, but only after a delay of 1 second. This is a good compromise. If I want the suggestions, they come quickly. But showing the suggestions at once would be too disturbing.

In the same vein, things related to mouse position, such as tooltips appearing and disappearing, usually have delays.

On the Mac, activating a menu item — even with its keyboard equivalent — makes the menu blink in blue. If the blink were too fast, it would be unnoticeable, or blinding. Animations are something special. There must be not too many animations, and they must not be too slow. They can be irritating. But some animations help the user understand what happened (where is my window ??).

When I leave my finger on N, key repetitionnnnnnnnn is not just a while loop adding a character at each iteration. Otherwise, after two seconds I would already have a gazillion letters N. The repetition rate is even settable on the Mac.

Throwing a surprise dialog in front of the user is generally a bad practice. However, computers do that over and over. And, in some few cases, it is justified. I am using an application, and I am going to click on something, and the Windows Update dialog jumps in, with two buttons : Restart now and Restart later. If luck is bad, the first button comes where I intented to click, I click, and… bye bye ! The same happens with the keyboard. If I am writing the letter R of “keyboard” just when the dialog jumps in, bye bye ! So a good practice when showing such dialog is to have its buttons ignore action during the first two seconds. Firefox has a similar delay on some controls, to prevent click hijacking.

When you try to authenticate in a system, the system often delays its response on purpose. This intentionally harms your user experience. This is for a — good — security reason.

Slowing pure processing speed on purpose is rarer than slowing UI on purpose. But it can have good reasons too.

For example, Mac laptop computers have an energy saver option, on by default, which makes them run their processor more slowly when on battery.

I think of the telephone. When you hang up, the link is not cut at once. If you quickly take back the phone, you can continue your conversation. I don't know whether there is an intentional delay at play or just a slow process.

When you are in a hurry, the lift doors open and close too slowly. Their engine can certainly make them move much more quickly. Having the lift doors move more quickly would improve the experience of most users. But they would knock out a few fragile things called children, so their little user experience would be harmed by the improved speed. :-) The same goes for metro and tramway doors.

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.