Bad UI: What I learned from the world’s worst stove about UX design
By: UX designer Robyn Bragg
As we all know, lessons can be learned in the oddest of places. Learn about the ineffectiveness of poor user interfaces with this real world story from a Nonlinear UX design expert.
Anyone who has been in the trenches of user interface design knows that it’s incredibly difficult for users to learn a poor interface. While I could give lots of web-based examples from my day to day at work, there is one bad interface that stands out above the rest, so much so that I am constantly going to back to it as a prime example of just how to create a terrible interface. Let me lay it out for you:
I rented an apartment a while back; it came with a stove. This stove had one of the worst interfaces I’ve ever used.
After four months of daily use, I was still turning on the wrong burner at least half the time and had accidentally left the stove on more times than I could count. It was far from an ideal (or safe!) situation.
Of course, as a good usability professional, I set about deconstructing WHY it was so bad. Why, despite my best efforts, had I failed to learn this interface? And, more importantly, by learning what didn’t work, what could it teach me about how to do things right?
Check it out for yourself:
So why was this interface so hard to learn? Take a look at this diagram showing which knob controlled each burner.
This interface breaks several laws of good interface design:
Interface designers use proximity to create relationships between items. If two items are placed close to each other, our brains automatically assume they go together. It’s a basic principle of visual perception that this interface badly violates. One would assume the top two knobs control the top two burners, since they are the closest, right? But no, they controlled the two leftmost burners.
This is very important. Once you pick an interface approach you need to stick to it throughout the whole user experience. Changing things within the system (e.g. using a certain type of button to do one thing and then have it do something else later on will cause massive confusion). In this case, I could probably have figured out the top two knobs meant “left side” and the bottom two knobs meant “right side” if not for the fact that in one case the top knob of the pair controlled the top burner, but in the other pair, the top knob controlled the bottom burner (which violates the proximity law as well). So, really the arrangement was pretty much random, which made it extremely hard to learn. All the cues that I was familiar with didn’t work, so I just kept turning on the wrong burner again, and again and again…
This stove did not match any of my previous expectations of stove behaviour (there’s a phrase you don’t hear everyday!). I expected it to be similar to other interfaces I was familiar with (since it looked very much like every other stove I’d used) but it wasn’t. This isn’t to say that a new interface can’t be wonderful and revolutionary, but new interfaces should be used with extreme caution and thoroughly tested before launch.
This is a minor point, but it demonstrates the overall lack of thought put into this interface. The little icons that tell you which knob activates each burner were diamond-shaped. This bore no relation to the shape of the stove. I didn’t have too much trouble with this, but it really seemed to bother and confuse my husband. The diamond shape was just another random, poorly thought through element that added to the overall cognitive load of the interface.
Would this bother you?
There was no individual feedback for each burner, just one general light which came on when any burner was on. And, as it was an electric stove where the burners didn’t change colour when hot, there was no secondary feedback to tell me which burner was activated. This led me to often turn the wrong burner off when I had several going at once.
Given enough time and burnt food I would have likely learned how to use this interface (I moved before I could prove this hypothesis), but only because I was a captive audience, using the interface every day, and hot food is a strong motivator. If I had any other choice would I have use a different stove?
Am I likely to EVER buy a new stove from this company?
With websites, users have several options when using poor interfaces (in order of preference):
- Go elsewhere: If users find another site providing the same content/service with a better interface, they will likely go there and never come back.
Cost: lost revenue, lost audience, lost engagement
- Struggle: In some cases users have no choice but to use the interface. I don’t know about you, but struggling with poor interfaces makes me feel victimized and angry. I leave the experience with a lot of resentment towards the site and the organization and tend to avoid using said interface if I can.
Cost: lost engagement, lost trust, lost reputation, decreased brand value
- Learn the interface (at a high cost of time and mental energy): This is true of sites we must use daily, like intranets, but it’s worth noting that if the interface is too difficult or frustrating users will prefer workarounds (like email, Dropbox or Google Drive) rather than uploading files to an intranet.
Cost: lost productivity, loss of security
Don’t reinvent the wheel
With interfaces, familiarity is good. The less a user has to think about it the better. There are a lot of great, easy to use stove interfaces out there, and if the designers had just stuck with familiar patterns, I would have had a reasonably good experience.
Test, test test
Always test the solution with real users in a realistic context before launch. The usability problems with the stove were glaring and could have been identified and fixed easily before launch if tested in real world conditions.
If you take away one thing, let it be this:
I’ve heard people dismiss usability concerns with the offhand remark “oh well, users will figure it out.” Maybe they will or maybe they’ll leave and never come back. In any case this attitude can be dangerous, especially when your interface is linked to revenue or productivity.
Taking time to put ourselves in users’ shoes is always worthwhile. Simple usability testing can be done fast without adding huge costs to a project, and changes can be made quickly, especially if the testing is done before the start of development.
How about you? Are there sites which cause you daily frustration? What’s the least intuitive interface you have encountered?