Speaking for web and mobile application UI design, do you think that "Flat" UI aesthetic is inherently better than more skeuomorphic designs?

For applications where your users might be considered "professionals" or power users it seems like it would be safe, but for more mainstream applications and websites do you think that the general public needs a little help?

Do any of you know of any user research that compared minimal "flat" buttons or UI widgets vs. more "buttony" looking UI elements? (Or any research in this area at all, really)

In terms of usability, I can't answer concretely by citing a study. Anecdotally, our customers have said that our flat UI is one of the best and most beautiful that they've used.

I can say that a small, lean company that pursues a cleaner design is more consistent its thinking about every aspect of its product and that one that doesn't betrays its own ideals.

I also find it more beautiful. There's always a shitty design trend that's fashionable. Geoff will remember the 2000s when guys like Joshua Davis were creating interfaces in flash (ps2.praystation.com). It's still beautiful.

Callous as it sounds, I don't think it matters whether the general public needs help by way of heavy realism. They'll eventually disappear.

"Callous as it sounds, I don't think it matters whether the general public needs help by way of heavy realism. They'll eventually disappear."

I've watched more than a couple user tests where users will just not realize they can click on something that we as designers/developers think should be obvious (this goes for everything, not just buttons). So my point is really more along the lines of: Why promote a certain design aesthetic over something else without proof that what you are promoting is actually better? Is making something aesthetically more pleasing worth a potential drop in usability?

Usability can be botched with any style of design. Is usability even the yardstick for a well designed interface?

I'd nonetheless argue the skeumorphism has the opportunity to create usability issues more often. Is a calendar app that looks like a paper desk calendar more usable because it's obvious what its function is, or less usable because it introduces the constraints of paper?

I don't think comparing _exactly_ the same interface in a glossy vs. flat war is the way to do it. You have to approach visual problems differently when working with flat design elements, so a one-to-one translation from glossy to flat is doomed. I find that when working with a flat aesthetic, my solutions themselves have to get simpler.

But I also think that Allan is right in saying that the average user's average to flat interfaces is eroding. Facebook is the perfect example of interface trending towards flatness. Their major actions buttons still have a slight bevel, but lots and lots of the secondary actions are completely flat. And judging by the fact that they haven't changed their interface style, it must be working.

I find this post entertaining.

Obviously it's in their favor to inflate this matter so they can promote themselves/their product. They must be well aware that "flat" is as much a fashion trend as "skeuomorphic". It's pretentious to call it "an era". It's no different than runway fashion– everybody starts doing the same thing; when it's everywhere, to stand out as a designer, you have to do the exact opposite. This repeats over and over in cycles. Same in graphic design, industrial design, architecture...

Bonus: check out Laver's Law.

I think if you look at it purely as design then yes, it's a superficial choice, but I'm much more interested in the possible gain or loss of usability by choosing one style of design over another. LayerVault (and many web things these days) are more of an application than a static set of linked web pages. Clickable elements aren't just linking to documents, they are performing actions, so making that apparent to users is important.

My hunch would be that less experienced users would miss some non-button looking buttons.

Making something look like leather is one thing, or a calendar looking like a paper calendar, but what about the UI elements that do things?

Related: How long until apple changes iBooks to not have page flip animations?

Addressing Geoff's original question about research in the area: During my undergrad human centered computing studies, I read dozens and dozens of formal research papers on HCI and CS, and rarely came across anything of merit discussing visual interface styles. I even went hunting through the research db's for things considering specific design styles when applied to research problems. I think the problem is that visual styles and trends a bit too time sensitive for true formal research to be done on them.

There are two arguments I'm seeing are 1. There is no right way, there's only what's popular, and 2. There's a loss in usability by making things look simpler.

"Flat" as a style may at some point pass in and out of fashion. That don't really matter — the point isn't to flatten UIs. An aesthetic isn't some kind of panacea to interface problems.

I'd strongly disagree that there's no right way to design for the web. More importantly, I'd disagree that what's good is subject to the combined tastes of whomever has the most influence.

@Geoff, my gut tells me the opposite. My hunch re:buttons is that users are scanning for things that really stand out. Take the "compose" link at the bottom of this page, for example.

"There's a loss in usability by making things look simpler."

My point is more that there may be a loss of usability by making things look less actionable. I'm going under the assumption that people make buttons look like physical buttons so people know they can press it. If that button doesn't look enough like a button, perhaps less people will click it simply because they don't know that it's clickable.

This was already addressed above somewhat by pointing out that "flat" doesn't have to mean "it doesn't look like a button." But I'm still curious about any usability studies that have been done in that area if they exist.

My experience is that flat or "realistic" doesn't matter. What matters is that hierarchy is evident at a quick scan and that users can find their way to the next actionable item easily. We stare at these UI's all day long as we design and critique them like pieces of art. That's not how the user sees them - they view most of the page through peripheral vision. Contrasting items will draw thier eye and lead them around the page. After doing a high fidelity mockup, I take it into Photoshop and give it a 1pt gaussian blur. The hierarchy of the page needs to be evident in the blurry version. What I've found does help users is having a good amount of white space. It simply makes content areas and action items easier to pick out.

My point of view as front-end developer. We back in time with flat design. Sometimes I ask myself "why we didn't do that a nine years ago?" Everything is simple, minimalistic, and everything looks great. Usability is quite simpler too. Flat gradients are more elegant than glossy aqua style. We could do this a years a go. But we had a time to experiments, to try every option, but now most important is maintenance and maybe this is a key of "flating" everything too.

About usability, we must remember about key patterns - new look, modern look in new operating systems will update our feeling. I thing, that sometimes in usability we forget about experimenting and go out of the box.