- We’ll define usability in an academic way
- Consider the problems with it
- And provide a shorter, actionable meaning of usability.
After reading the article, you’ll:
- know how to use the term usability in a correct way
- know when people use it wrong (and feel a smarter person)
- distinguish academic usability experts from the practical ones.
What Is Usability?
Usability is metric, a number. What does it mean?
It means we should measure usability to know if it’s high or low. If we add or remove a feature, we should re-measure usability again.
What exactly should we measure?
Traditional Definition of Usability
According to Jacob Nielsen, usability industry’s prophet, there are 5 components to measure:
- Learnability: for the first time users, how easy it is to learn?
- Efficiency: for repeating users, how fast is it?
- Memorability: how easy is it to remember?
- Errors: how many errors users do?
- Satisfaction: how pleasant is it to use?
Problem with Usability Definition
It’s a hell of measuring.
When you add a feature, you have to run multiple tests, some of them are time-consuming. Take memorability: how long are you prepared to wait before re-testing the feature?
Worse, if you create the whole new app or redesign an existing one.
Example of Redesign
Consider our recent redesign of Icons8:
Old web app for icons
It’s well-tested for usability. In particular, I’ve tested these tasks:
Some of the tests that I’ve run on this design. There are more for the collections.
I find this aged. When I’ve shown it to the cool guys from Clay, they said it’s “pretty okay.” I credit them for being polite.
Here’s what we’ve come with:
One of the iterations of the redesign of our icons app
This design is a vast improvement.
- It’s looking aged, thanks to Kyrilo, our UX designer.
- Instead of the left side panel for icon details, we’ve got the inline (inspired by Google Images). The theory behind it was offered by our friends from Usethics: moving attention between the list of icons and the left panel is a hard task and the major source of the cognitive load. Make it closer, they said.
- We’ve redesigned the collections, providing a shortcut for the default “Favorites” collection.
Should we run all the usability tests for this redesign, it will take 30 measurements not counting the collections. Collections will easily double that.
Let me provide a more straightforward set of metrics. I find it cheap, dirty, and good enough.
Practical Definition of Usability
For me, usability is a set of three metrics. Let’s call it Sucers:
From these 3, you only have to test 2:
- Success — test it
- Errors — test it
- Speed — don’t test it.
Here we go with our expanded definition.
1. Usability Is a Success Rate
Usability is a value. Usually, it’s a success rate for a given task.
Example: if 80% of users can buy a product, usability is 80%. Usually, it’s a high value.
2. Usability Is the Number of Errors
It could be the number of errors that users made. How long did a user wander all over the website before successfully ordering her product?
3. Usability Is Speed
Or, it could be the speed. In real life, I never use it. In part, because it’s much harder to measure. It requires many tests, and the results aren’t accurate.
It makes sense to invest the resources in eliminating the errors and adding some heuristics. To speed up:
- Fewer errors (as in #2)
- Fewer clicks
- Fitt’s law
- More shortcuts
What to Measure
Back to our example with our icons web app, we have to measure only two things:
- How many people can complete the tasks?
- How many errors do they make on the way?
All it takes are 10 measurements, down from 30.
For speed, we use the heuristics:
- Showing the icon details in the middle of the icons grid, we are doing good
- Adding a shortcut icon for adding to an icon to the Favorites collection, we’re doing good.
Call it a day.
How do you guys test your apps? Showing it to your wife counts as testing, as well as a formal testing lab equipped with a one-view mirror. Share your examples in the comments.