An Apple Watch Heuristic Evaluation

At UX Launchpad, we teach design courses. So I went to my local Apple Store with a notebook and a camera to analyze the design of Apple Watch as a way to learn some things they did well and point out some things they can improve. The design issues I found fell into a few distinct categories:

1. Straight Up Buggin, Y’all
2. Troubling But Easily Fixed Issues
3. Fundamental Concerns

And then of course there’s a grab bag of other insights and closing thoughts. Note that this isn’t an issue of our popular Design Explosions series, because those are all about staying positive. Whereas the point of a heuristic evaluation is to tease out and discuss problems with a design, so that’s where I’ll be focusing.

First up, bugs!

1. Straight Up Buggin, Y’all

My first experience with Apple Watch was pretty bad. I was attempting to broadcast on Periscope but I kept tapping things on Apple Watch that didn’t work. Afterwards, I talked to an Apple Store employee who said he found the demo units to be buggy. So I tried again, this time without trying to use the device through a viewfinder, and everything worked as expected.

Since then, I’ve tried the device a few more times, and I never struggled as much as I did with that first demo unit. But I think my wife has the best explanation why: I was trying to tap a tiny screen by aiming my finger at what I saw through a phone screen, while a hundred people chatted with me (including some trolls), while trying to narrate. I was probably distracted.

But there’s a third factor that explains my original impression, and it’s a big one. Certain things in the demo unit just don’t work, but they fail silently. For example, you can try to turn on Airplane Mode or compose a new message, but you’ll tap several times before you realize it’s not a mis-tap, the demo won’t allow you to perform those actions.

(This video is ugly when I embed it so please enjoy a link instead)

This is a software design no-no. An interface most always displays a response, instantly, even when it can’t succeed. Other places on Apple Watch’s demo get this right. For example, if you try to use Siri to set up a calendar appointment, it apologizes and says the software can’t yet do that.

Of course the final units will be able to perform all advertised features. So those issues, while extremely concerning to me at first (and I assume many other early adoptors who hit similar problems), don’t actually point to a bad interface design. They point to some early bugs in the demo units. Not ideal for the initial rollout, but not a huge concern in the long run.

2. Troubling But Easily Fixed Issues

My time working on the Windows Phone design team helps me know where some tricky interaction design challenges live in built-in apps. For example, when setting alarms, you can sometimes get a few levels deep in the hierarchy. (for example, when deciding when an alarm should repeat) So what do you do when a user cancels out of the app? Do you save their changes made until that point? Or do you destroy the changes they made? (There are other options, like not letting you leave an app until you select “Save” or “Discard” from a prompt, but that’s a non-starter.)

Windows Phone went with destroying changes if you don’t tap save, a choice made worse by certain flows (again, setting repeating alarms) that actually make you tap save twice. It’s a flaw in Windows Phone’s interaction design, but I’m happy to report that Apple Watch always saves your progress on the fly. I believe it’s a good choice because it’s one less thing to think about, and I doubt anyone will miss the PC-style “Save” and “Discard Changes” dialog.

(Video of making a choice, exiting out, and having the changes saved)

But that doesn’t mean all dialogs are bad. For example, destructive actions (like deleting an alarm) should always ask you to confirm, and ideally let you undo. Apple Watch does neither. If you accidentally delete an alarm, it’s gone. (Check out this video — I perform a bunch of actions, then at the end I tap delete and whoooosh a destructive action without a confirmation.) I chalk this up to a silly oversight or a bug. It’s easy to fix and there’s no reason to believe they won’t.

I found a few other minor things like this, but none really worth calling out. Just odds and ends and things left undone. Anyone who’s worked in software knows how it goes. You address priority 1/severity 1 bugs first, and a bunch of pri3/sev5 stuff goes unaddressed. I bet Apple has a bunch of bug fixes waiting and ready to go in a 1.01 update.

Like category #1, these are things that jumped out at me as problems, because I’m a designer it’s my job to know this stuff. They damped my enthusiasm a bit, but they’re not deep structural problems. They’re just bugs that they didn’t get to yet.

But that’s not all I found. I also found bigger stuff. Grab a comfy chair.

3. Fundamental Concerns

I think the Apple Watch is pretty solid overall. The word “concerned” doesn’t mean I think they’re doomed or that I’d expect to see an embarrassing number of returns. But they are definitely going to cause friction, and they’re inherent in the design of the system and the constraints Apple is working under. There’s no quick fix for these.

Tiny touch targets

In 2007, Apple was very explicit about touch target sizing. They knew the average finger size, they knew what factors make it easier or harder to tap a target on a screen, and they informed their developer community to exercise judgement when designing tap targets. But the Watch screen is crazy small compared to an iPhone, meaning some tap targets are tiny. When you go to tap a back button to move you back a screen, you’re going to miss sometimes. And whether you miss 50% of the time or 5% of the time, it just feels more fiddly than you’re used to on an iPhone.

Which is why the Digital Crown is such an experiential breakthrough. When I can use it to manipulate the UI, the phone feels smooth and beautiful. When I can’t, and I have to tap a little target to complete a task, my heart sinks a little bit. It’s not that the touch are a disaster, but they’re certainly harder to access than I’m used to. This is a video of me having to tap multiple times frequently. And I have small fingers!

Apple could have made the Watch bigger to account for this, but that only gets you so far. There’s an immovable law in play here — smaller things are harder to tap. And you don’t want a watch to be phone sized. So tapping a reasonably sized watch screen will always be harder than a phone. Thank goodness for the digital crown.

Some Overloaded Actions & Gestures

On the Maps app on iOS, you can do a few things, using a few distinct gestures. In order of most common to least common, you can:

Drag to pan the map around
Tap to toggle the chrome
Stretch/pinch to zoom in or out
Double-tap to zoom in
Long-press to drop a pin

This means even if you never discover anything other than scrolling and tapping, you’re probably ok. Tap the address bar and type an address, and you will probably complete your task successful. Power users might use the long-press, but you certainly can’t rely on that gesture, so they don’t.

Whereas Apple Watch has to rely on it if you don’t come through Siri, and that adds a learning curve. And it’s not just that you have to learn a new thing, it actually seems to work inconsistently. Here’s a video of me using Maps and getting confused.

Digital Crown zooming works. Panning works. I don’t think tap does anything. But long-press (aka Force Touch) seems to drop a purple pin. Except when it brings up Search and Contacts. I couldn’t manage to figure out how to trigger one versus the other, and that’s concerning.

Putting myself in the designer’s shoes for a moment, there a few things the design is attempting to allow, beyond zooming and panning:

Orient the map on your location
Let you drop a pin
Type an address
Go to a contact’s location

So orientation comes via a button on the bottom left of the screen. But then the other three are all behind force touch, with no clear (to me) way to distinguish between them. I suspect a lot of people are going to dead-end in the Maps app if they load it directly. And of the people to try Force Touch, some are going to get a confusing purple pin when they were expecting a search field.

Similar problems are found throughout the interface. You can’t clutter the UI with buttons (I remember when designing for iPhone felt cramped!) but as an app creator you only have touch and force touch to work with. That leads to some situations where people will struggle to complete their tasks because you have to guess at how to complete your task.

Launching An App Is Hard

I love the visual design of the app screen. But the icons feel microscopic, certainly too small to tap. I find myself panning to get the icon somewhere near the middle of the display, then zooming with the Digital Crown. When it works, whew. But it often doesn’t. If it’s that hard to launch with the larger size of the watch, I’m going to struggle with the smaller one I bought for me and my dainty Tiny Person Wrists.

But people are resilient. “Satisficing” is a real thing when people use software, and I’m sure my experience will be no different. You can launch apps via Glances, the Widget or Google Now-esque mini-apps available on Watch, and I suspect that will be my primary entry point. In a perfect world, I’d be able to launch apps reliably every time, but it’s a small screen. Tradeoffs about in software, doubly so for software this tiny, so thank goodness for Glances.

Two Buttons

I should say right up front that I think two buttons was the right call. If there was only one button, it would take waaaaaaay too long to load up a contact. Digital Crown press->Apps->Load the Contacts App (see above, this is not trivial)->Find the Person->Finally start completing your task. Whew.

You know how lots of people hate the idea of Apple Watch and can’t understand why you wouldn’t just use your phone? Well, most of the time they’re willfully ignoring the benefits of wearables to make a point. But on this, they’d be 100% right. If I have to fiddle for 15 seconds to contact my wife, yeah, I might as well just be using my phone.

Which is why having a dedicated button is the right call. But it does lead to cognitive overhead. Personally, I think people will acclimate to it just fine. One button is like the home button on an iPhone. The other loads contacts. It’s not as easy as iPhone, but it’s not as complex as Android. It’s harder, relatively speaking, but I wouldn’t call it hard in an absolute sense.

Still, it’s a little more cognitive load to consider. For example, my son never did figure out that he could press the Digital Crown. He found the Contacts button, but that was it. Whereas I have video footage of him as a baby navigating through the iPhone lock screen, to apps, and back again by tapping the Home screen, all within a few seconds. So.

(Note: I also have footage of him using Windows Phone. Notable insight — he and his friends always hold it with his palms and his fingers splayed because he kept accidentally getting whisked out of his game and taken to Bing, thanks to the overly sensitive capacitive buttons.)

So two buttons is something people will get used to, and I understand the design tradeoff, but it’s one more little bit of cognitive load. Should I click here? Or here? But the buttons are nothing compared to Apple Watch’s modes.

Modes Complicate Things

Oh, modes. Early in my design education, I was always taught that modes cause a ton of problems in a design, and that’s been borne out by all the data and research I’ve been involved in or read about in my entire career. Sometimes modes work well (like “edit” mode when you’re going to delete a bunch of messages or “wiggle mode” to rearrange app icons) but they should be used sparingly and carefully.

Y’all, Apple Watch has a lot of modes.

If you’re looking at the clock, you can swipe from the top or bottom to get notification center stuff or Glances. This is the only place you can access these locations. (after all, if every swipe down loaded notification center, you’d accidentally trigger it constantly on such a small screen)

If you’re looking at Glances you can swipe down to dismiss, or right and left to see other glances. But you can’t get to apps or the clock without exiting out with the Digital Crown.

If you’re in an app you can use it fully. But you can’t get to Glances or Notification Center. This means if you go from a Glance to the app and then decide you want to get to another Glance, you press the Digital Crown to return to home, then drag up to get back to where you were in Glances. On one hand, it’s only two steps. You’ll figure it out. On the other hand, there’s a clear cognitive burden this places on you while using the app. Here’s a video of me going from the music’s Glance to the full app, then having to stop to consider how to get back to where I was.

(One of my favorite articles about these small but important moments is Your App Makes Me Fat. If you haven’t read it, I highly recommend it.)

It’s Trying To Do A Lot

People have written about this, but they usually mean at a high level, like it’s trying to perform a lot of different tasks, or install a lot of apps. But I actually noticed it most when I was trying to personalize the watch face. Check out this demo at Apple’s website and watch the steps involved in customization.

At a high level, it’s just “Get into edit mode, then spin the dial to pick what you want, then confirm when you’re done.” Sounds reasonable. But the sheer number of personalization options means the UI has to flex and scale awkwardly to address a visual smorgasbord of options.

This explains why Apple has their reputation for products that quote-unquote just work. Historically they’ve reduced the number of options to reduce fiddling, while their competition bets on choice at the expense of simplicity. This time around the tables have turned and Apple has gone big on personalization. There’s no doubt it complicates the overall experience, and it feels un-Apple to me.

But hold on to your hats. I actually want all these choices that are bloating the experience. I did not see that coming. I’ve leaerned that I don’t want a dizzying array of options for a laptop, or a smartphone, or even the watch hardware, but software is different. If the UI needs to be more complicated in order to help me feel like a super special unique snowflake with great fashion sense, fine. That’s a price I’m willing to pay.

All is vanity.

Other Than That, How Was The Play?

If you’re looking for me to throw the Watch under the bus, sorry. Write me off as an Apple fanboy (it wouldn’t be the first or last time!) but I’m still incredibly excited about Apple Watch. Despite all these design concerns, I’m definitely going to buy one. Why?

First, the small, bug-sized stuff doesn’t bother me. And even the larger, more fundamental design concerns I have can be boiled down it “it’s hard to navigate to different areas on the Watch.” But the device isn’t meant for bouncing between apps like a laptop or smartphone. I predict I’ll use it largely as designed, as a glance-and-go experience. To use a phrase my friend Jake Zukowski coined, the software isn’t trying to be “sticky,” it’s striving to be “slippy.” And in fact, I think it’ll do better than any other smartwatch on the market (both UX and sales) for the foreseeable future.

Why? Because when I think about ways to address the concerns I have, the tradeoffs come down to things like “make the device a lot bigger,” or “don’t let me customize as many things,” or “make it so I’m constantly accident-swiping into Notification Center,” or “make the beautiful display of apps into a grid of 2x2 icons that I have to swipe through, iPod mini style.” And none of those sound good to me. So I’m ok with where they landed. These are tradeoffs to be sure, but that’s what design is. I think they chose well.

But I have a personal consideration that’s much bigger than all that, and it makes all the discussion about fashion and vertical integration and marketshare and finding new billion dollar businesses and all the rest of it pale in comparison. It’s the peace of mind I get from a tap on the wrist.

In my family, we’re living with illness, and I have three young children. I have frequently missed important texts from my wife because it’s hard to find the balance on my phone between keeping it quiet but letting important things through. Sometimes I put it in my jacket pocket, so I don’t feel the vibration, but I’m in a movie theatre so it has to be on silent. And these considerations aren’t theoretical, they’ve happened and they add stress to how my family communicates, lives, coordinates, and plans.

What I see in Apple Watch is a way for my wife to be able to contact me from her Mac, iPhone, or Apple Watch (if she gets one) with almost perfect certainty that I’ll notice it. Other than running out of battery or forgetting to wear it in the morning, it seems like a solid way to stay in touch. The most reliable ever. And to me, that’s valuable on a life-and-death level.

There’s no one thing that Apple Watch can do that the iPhone (or Android devices) can’t do. But that’s the wrong metric. From the Uber app to my family members tapping me on the wrist to monitoring my heart to helping me be more playful through little drawings on my phone, the watch is helping me do things better. In some cases, significantly so.

So it’s not perfect, because nothing is. It’s complex, but I understand the tradeoffs. They have a lot to work to do, like any company shipping the first version of something. But I have no doubt it will change my family’s life and health for the better. That’s worth more to me than any family heirloom.