The Problem With Flat Design, According To A UX Expert

Just because people like flat design doesn’t mean they get how to use it.


As a user experience expert at Nielsen Norman Group, Kate Meyer has watched a lot of people use a lot of websites. But her favorite quote comes from a young adult–under 25–talking recently about navigating a flat web design: “I don’t [know what’s a link]. I just start clicking and praying that it works.”


It’s a funny and telling comment, and it reveals that a lot of research remains to be done about whether young people intuitively understand flat UI better than their parents–and, even if they do, whether designers should design specifically for young people based on the idea that they’re better at navigating flat design. “There’s an underlying assumption that flat UIs are targeted at millennials,” Meyer writes. “So it’s okay if older users don’t quite ‘get’ how to use flat UIs, these designers argue.”

Over the past year, Meyer—who teaches conferences about designing for young users—has examined how different users approach flat interfaces, offering a compelling argument against using flat design as a way to appeal to young users. Or, as Nielsen Norman Group’s Don Norman and Bruce Tognazzini put it in their screed against Apple’s emphasis on flat design and minimalism, “So what if many people can’t read the text? It’s beautiful.”

Testing Flat Design

Flat design generally describes minimalist interfaces that eschew skeuomorphic flourishes: drop shadows, gradients, and the like. Recently, Meyer ran a study that asked young and old users to rate the attractiveness of five fictional websites. Four were “flat,” while one was skeuomorphic, rife with drop shadows and signifiers. NNG asked almost 500 users—half between 18 and 25, half over 35—to rate each site aesthetically.

A flat website design versus the skeuomorphic control design.

It turned out that younger users liked the flat websites a lot more than their older counterparts—by more than half a point on a 1 to 10 scale. Older users said the flattest website designs were “boring,” while younger subjects described them as “professional.” Fascinatingly, the youngs rated the skeuomorphic design as just as attractive as their “parents” did, with both age groups describing a shadow-stuffed website for a fictional steak house as “professional” and “trustworthy.”

So younger people found flat designs better-looking. Old people didn’t like them. That may not surprise you–especially if you’ve ever watched an older person try to navigate a super-flat website for a few minutes. But whether or not a user likes a design is only half the story. And it turns out, it doesn’t have terribly much to do with how well they navigate it.


Liking Something Doesn’t Make It Usable

In her past research on flat design, Meyer closely studied how well young adults could navigate flat sites. She observed something odd: While young people seemed faster at navigating the designs, they also indicated they didn’t really understand the UI intuitively. In fact, for the most part they seemed to have, uh, pretty much no idea what they were doing.

“It’s hard when you think something’s a link and it’s not. And you have to figure out how to get it another way,” one user said. Meyer compares this behavior to lab rats in conditioning studies:

This behavior is analogous to the behavior of laboratory rats in operant-conditioning experiments: If a rat gets a food pellet at random intervals after performing a specific action, the rat will keep doing that action in the hope of getting fed again. Similarly, users have discovered that clicking elements that don’t have strong signifiers sometimes works. Like the lab rats, users will stick to random clicking as long as they get rewarded from time to time.

Meyer and her colleagues use a term called “click uncertainty” to describe this confusion over where to click. “We see it manifested in a lot of different ways, and most of the time, users won’t even verbally comment on it because they’re too busy trying to figure things out,” she says. “Often their eyes flick across the page, [or]they might mumble some link labels or hover over things to see if they’re actually links.”

So young people didn’t necessarily “get” flat designs. They were just better at quickly testing where and how to get what they wanted in the face of click uncertainty. That runs counter to the way many designers think of young users–that because they’ve grown up with contemporary technology, they intuitively understand it when there are fewer affordances. “Please don’t think that because your younger users can adapt to poorly designed interfaces you’ve got a blank check to design careless, signifier-free interfaces,” as Meyer wrote a few months ago.

Inclusive UI

Meyer isn’t arguing that designers should give up flat design completely, but that good design fundamentals still need to be respected. Users who like a product don’t necessarily “get” how to properly use it, and that’s a problem. Companies that want to appeal to younger people might end up sacrificing too much by adopting super-flat designs. “How desperately do you want to be perceived as cool?” she writes. “What are you going to sacrifice to achieve it?”


Paying attention to the older users–the ones who don’t “like” flat design–might help to solve flat design’s usability issues sooner. This insight echoes the ideas of other technology companies moving toward “inclusive design,” or the notion that by designing for ignored or underserved users–including the elderly or disabled–products will become better for all. Inclusive design has quietly spurred some of the biggest technological leaps of our time: Cliff Kuang, writing about the evolution of this approach, recently pointed out that the typewriter, email, and even the telephone evolved out of designs for the blind and deaf.

Now, the same idea is being embraced at larger companies including Microsoft and Ford, leading a new wave of inclusive design in technology. Interface design, which is so deeply connected to competition between major operating systems and fashion in general, has been slower to adapt. Yet Meyer’s research, along with products Learn To Quit–an app designed alongside psychologists specifically for mentally ill users–show that the same ideas are making their way into UI.

What’s more, the scientific process is finally being leveraged by designers to differentiate between what users “like” and what they actually use. The difference, it turns out, is larger than you’d think.

About the author

Kelsey Campbell-Dollaghan is Co.Design's deputy editor.