Four Reasons Why We Won’t See Third-Party Apple Watch Faces (And What Apple Is Doing Instead)
Apple Watch users have wanted third-party watch faces since Apple unveiled the watch in 2015, and developers have wanted to create custom watch faces for just as long. But despite the pleas of users, developers, and a well-known podcaster, they’re probably not coming.
As an Apple software engineer, I worked on the first two releases of watchOS, so I’m familiar with many of the Apple Watch’s internal trade-offs. While I don’t have any inside information about current versions of watchOS and Apple Watch hardware, there are at least four reasons to think Apple won’t support third-party watch faces any time soon, if ever.
Reason #1: Battery Life
The main reason Apple doesn’t allow third-party watch faces is battery life. On the Apple Watch Series 5 and Series 6, the watch face is displayed almost all the time. It’s vital that the code driving the watch face consumes as little power as possible. Apple engineers go to great lengths to ensure the watch face code is power-efficient.
These efforts go well beyond simple tricks like hiding the second hand when the face dims since animation takes more power than a static display. Apple engineers have intimate knowledge of how watchOS displays graphics and how the Apple Watch’s GPU works, and for better or worse, this information is proprietary. They know which graphics techniques use the least power. Their animation techniques are the most energy-efficient possible. They have access to private graphics APIs that aren’t available to third-party developers. And they have internal testing and measurement tools that the company doesn’t provide to third-party developers.
The Apple Watch doesn’t achieve all-day battery life by accident. Apple engineers spend thousands of hours fine-tuning the code to be power-efficient. Every night, Apple’s automated build system creates a new build of watchOS, called the daily build, using the latest code changes checked into source control. (This is true for all Apple operating systems; see “How to Decode Apple Version and Build Numbers,” 8 July 2020.) Apple engineers use the daily build, so they’re all running the latest version of watchOS.
The daily build is also loaded onto a rack of Apple Watches in the power testing lab. They run through a set of scripts that simulate normal use to see how long the batteries last. The results are published to an internal Web dashboard that tracks battery life for every watchOS build. If battery life goes down, engineers are assigned to find out why and to fix it. Battery life is an obsession with the Apple Watch team.
No matter how capable or diligent they might be, third-party developers simply won’t have the internal graphics knowledge, the private API access, or the test tools to be as rigorous about battery life as Apple demands. If Apple were to open up watch-face development to third-party developers right now, battery life would almost certainly decline, which would make for a worse Apple Watch experience for users.
Reason #2: Buggy Code
The watchOS code that drives the watch face runs 24 hours a day, for months on end. It must be as utterly bug-free as possible. It’s unacceptable to glance at your watch and find the face has frozen, crashed, or has a visible glitch. The Apple Watch team does a tremendous amount of testing on watch code. Those engineers run automated tests, manual tests, and recruit thousands of Apple employees to use beta copies of watchOS and report any problems.
In my years working at Apple, I developed a deep respect for Apple’s Software Quality Assurance (SQA) engineers. They do a tremendous job. When software ships with bugs, it’s usually not because SQA didn’t report a problem, but because the schedule didn’t allow time to track it down and fix it (see “How to Report Bugs to Apple So They Get Fixed,” 17 June 2020).
Only the largest third-party developers, like Microsoft and Google, have equivalent testing resources. Smaller developers probably can’t guarantee the quality required for watch face code. And even if they could, it wouldn’t be economically feasible for them to spend that much time testing watch faces, which would quickly become a commodity in the App Store. Nor does Apple want to take on that level of testing itself as part of allowing third-party watch faces into the App Store.
Reason #3: Apple’s Image
It’s an understatement to say Apple is highly image-conscious. Apple obsesses over every detail of its public persona, from the Mac Desktop background to iPhone colors, and even extending to the exact shade of wood in Apple store tables. Steve Jobs examined dozens of shades of white cables before selecting the particular shade of white in the iconic iPod and iPhone earbuds.
The watch face is the public facade of the Apple Watch. It’s what everyone sees when they spot someone wearing an Apple Watch. It must be beautiful, contemporary, and polished.
There are certainly some third-party developers who create beautiful apps. However, as you browse through the App Store, you also find plenty of apps whose design is uninspiring and many more that are outright ugly.
Apple does not want the Apple Watch sporting a gaudy, grating, or downright dissonant face, even if that’s what you might like. Nor does Apple want to be in the business of deciding which watch faces are stylish enough to appear on the watch. Apple thinks its own designers do a fine job of creating a wide range of highly customizable faces, and it adds more faces with each watchOS update.
Reason #4: Copyright Worries
If there’s one department within Apple that you don’t mess with, it’s Apple Legal. Apple doesn’t want to waste time and money fighting copyright infringement lawsuits over watch faces. Many classic watch faces, like the Hermès face that Apple licenses for the Apple Watch Hermès, are copyrighted. Vintage faces were also designed before smartwatches existed, so licensing such faces for digital use means negotiating with a copyright owner that may not understand the issues involved in digital licenses. Worse, the App Store supports dozens of countries, and the copyright owner could be different in each country. It’s a legal nightmare.
Apple is all too familiar with watch face copyrights. Years before the Apple Watch existed, the company had to pay the Swiss Federal Railway service $21 million for a license after “adapting” its iconic Mondaine watch face for iOS 6.
Apple could require the developer to certify that they had a legal right to sell each watch face. But the company has no easy way to verify that the developer is telling the truth, and Apple would inevitably be named in any copyright infringement lawsuit because Apple has deep pockets.
When Apple originally started the iTunes Music Store, it took a small army of lawyers to acquire the legal rights to sell all those songs, in all the countries in which Apple operated. One reason it took so long for Apple Music to offer song lyrics is that written lyrics are licensed separately from the music.
Complications to the Rescue?
Apple thinks it has a feature that will satisfy users’ desire to customize their watch faces: third-party complications. The term complication comes from the world of mechanical watches, where it refers to extra information that’s displayed in addition to the time. Common mechanical watch complications include the date, the day of the week, and the time in another time zone.
The Apple Watch’s complications allow a third-party app to display additional information—usually a tiny amount of text and graphics—on one of Apple’s watch faces. watchOS provides several different shapes and styles of complications, and most watch faces can show several different complications.
What complications don’t show is the time—that’s reserved for the watch face proper. A popular Apple Watch complication is the weather, showing the temperature and a small icon for the current conditions.
Complications are part of an app, but they have special restrictions. The complication’s code gathers data, like the weather prediction for the day, and passes it to watchOS in a static data structure with details about when to display each data record. watchOS displays one data record at a time, updating the display over time until it’s time for the complication to load more data. The system is ideal for predictable, slowly changing data, like weather, ocean tides, or the phase of the moon. It doesn’t work well for data that needs to be frequently refreshed, such as stock prices.
Apple designed third-party complications this way to save battery life. The complication’s code only runs for a few seconds, several times an hour, which prevents it from sucking too much power. Most of the time, watchOS is just showing the complication using static data. That usually works well, but it does limit what can be represented well in a complication.
watchOS 7 Expands Customization
With watchOS 7, you can see just how much Apple wants to answer the desire for third-party watch faces without actually going that far. Changes include multiple complications from the same app, watch face sharing, and quite a few new faces.
- Allowing multiple complications from the same app in a single watch face may sound relatively minor, but it gives apps a lot more flexibility, presenting more data across multiple complications, and it lets users arrange those complications however they like.
- Watch face sharing may also seem somewhat unimpressive, but users spend a lot of time customizing watch faces, including the complications. It can take quite a bit of effort to get things just so, and if you’re showing off your customized watch face to a friend, you can now share it with them easily.
- Apple added seven new watch faces in watchOS 7. The new faces this time around include Artist, Chronograph Pro, Count Up, GMT, Memoji, Stripes, and Typograph. As you can see, Artist, Memoji, and Typograph are relatively bare-bones, but the rest of them offer lots of complication slots and other opportunities for customization.
These new features let users make their watches more personal than ever before. I’m sure Apple has additional customization options planned for future versions of watchOS as well. But third-party watch faces are probably not on the horizon.
Good article that I mostly agree with. A few questions about this article, though:
The APIs do not have to be private. That is a decision made by Apple.
Part of the approval process for a hypothetical face store could be that Apple subjects the watch face to the same tools that it uses to test its own faces. (To pay for this, Apple could require that all faces in the Face Store charge a minimum fee, with Apple getting its normal 30% cut, and/or they could charge a fee to a developer to have a face in the Face Store. That fee could be pretty high in order to limit the number of submissions that they would get.)
Doesn’t Apple already do this with the App Store - reject apps don’t follow specific rules? Why can’t Apple specify specific guidelines for faces which appear in a hypothetical Face Store along the lines of Human Interface Guidelines, except mandatory rather than guidelines?
Frankly, this is the biggest reason why I think Apple will not have a Face Store.
I don’t think that Apple is very good at designing watch faces, particularly the analog faces. Just as a few examples, on the Meridian watch face, if the face is black, the hands themselves are black with white outlines - not the best design, in my opinion. The same goes for most of the California faces - the hands are closer to the background color rather than a contrasting color. I’d love to see what a good designer could do with a watch face that wasn’t constrained that way (and, honestly, Apple could fix that by allowing the user to specify black or white hands on the face of their choice.)
There is only one reason Apple won’t let third party watch faces and that’s because Apple uses watch faces to sell the latest Apple Watch. You want the new watch face well you better go buy a new watch.
Yes, but not on aesthetic grounds. Which is why there are so many ugly iOS apps.
The main issue here really is battery life, I suspect, with the copyright concerns second. The other worries could theoretically be dispensed with, though it’s hard to see Apple having any incentive to do so.
In addition to Hermes, Watch’s Mickey Mouse face and voice most certainly had to be licensed from Disney. I agree that Apple is just being very selective, and of course, design sensitive.
I have to agree with Adam. Apple has no real incentive to support 3rd party watch faces. Apple is already selling millions of watches. They obviously have plans to add new features centering around health and fitness, since that’s what customers seem to want.
If customers wanted 3rd party watch faces, Apple would respond to that, but I don’t see any indication many customers want that. Apple marketing has lots of customer data, they know what customers want.
Remember, Apple’s interests are always: Apple, customers, developers. In that order.
I know when I first got my Apple Watch I inquired from Emerald Sequoia Software if their beautiful watch faces that they made for the iPhone (and later iPad) would be available for the Watch.
They said the same thing that was mentioned in the article, and also that the resolution for the watch (at least at that time) could not handle their complexities. But they are an example of a company doing quality work that Apple would benefit from teaming up with for great faces.
Wow, those are gorgeous, thanks for sharing!
Thanks for an excellent article!
The peek behind the scenes at their QA practices are particularly fascinating. I figured something like this was happening, but it has all been conjecture in my mind.
I was also going to post a reply about the APIs. It makes perfect sense that access to the HARDWARE is proprietary. Such access can cause instability and poor performance. But that’s the whole reason why APIs exist: they’re basically wrappers that hide everything under the hood, and expose just what Apple wants the developer to be able to access, in a safe, controlled manner. So, I’m still looking for an explanation to why they are not sharing the APIs.
I think battery life, stability, and appearance are the main reason Apple has blocked 3rd party faces. These are the same reasons they have made similar decisions over the years.
Thanks for a great story!
That’s not entirely true. Sure, there are some watch faces that are exclusive to certain models, like the Hermes, but my Series 4 has all the new watch faces introduced with watchOS 7. There are several the Series 3 and earlier didn’t receive because their screens aren’t large enough.
Back in the day of the classic MacOS (1980’s - 1990’s), Apple published API’s for developers fairly often. There was care taken to make sure the APIs were well thought out, but no one really considered that they might be supporting the APIs for decades. An API would be superseded by a new improved API, but Apple would have to support the old API, because lots of apps still used it. Over time, this became unwieldy, and it was bad for developers too. When there are 3 printing APIs which one are you supposed to use?
With Mac OS X and the new (NeXT) regime, publishing APIs became much more tightly regulated. APIs for new services are refined for months by senior engineers, and ultimately approved by an API committee before being published.
Often an Apple app does something new (eg: pull to refresh), which they implement themselves. Other Apple and third party apps like it, and start to imitate it, writing their own code to implement it. Usually several apps inside Apple need to be doing the same thing, duplicating code, before engineering management decides to publish a common API to support that. And then it usually takes another OS cycle (another year) before the API is refined, approved, and shipped. From the developer’s point of view this slow progress is frustrating. But Apple has managed to create a set of (mostly) coherent, maintainable APIs without excess duplication, even after 2 decades.
Apple doesn’t publish low level graphics APIs on the watch because it doesn’t want developers doing that sort of thing on the watch, and Apple doesn’t want to commit to maintaining those APIs for decades.
It was actually Loren Brichter who invented pull-to-refresh after he left Apple in his Twitter client Tweetie, which was eventually purchased by Twitter.
Good to know!
I appreciate the behind the scenes of this article. But I don’t really buy any of the reasoning except for #3. The rest could be addressed if the company felt like they wanted to invest in the effort to allow third party watch faces.
Public API’s could be created that are optimized around power savings. (And Apple isn’t exactly married to any API forever, given the marriage of watchOS to iOS; an old Watch becomes obsolete pretty quickly.)
Buggy faces will receive negative App Store reviews Plus, faces are mostly decoration; I can imagine, if Apple set their mind to it, that primary logic components could be handled by API, and that the third party face could be largely, say, an XML document plus components. And lawyers could protect Apple from real copyright concerns, or so I would think.
I think it’s nothing more than that Apple likes to be in control, and likes to think they’re making something beautiful. They don’t want a third party in the way of their relationship with their customer. A developer can’t write a skin for an iPhone or iPad, either, and I don’t think battery, copyright, or buggy code are the primary concerns there. They might be factors, but I think Apple just wants users to have an Apple experience.
Personally, I find it ironic, because to me, an Apple Watch is hardly something timeless — to the contrary, a few years is the most you might hope to use it. However beautiful Apple thinks its faces might be, it’s just a digital watch. The idea that third parties would sully the experience is silly to me, and the Watch’s boringness is one reason I haven’t had one since the original. Third party faces might change that for me. But, as the article says, I’m not holding my breath.
I was surprised to learn this as I don’t have an Apple Watch and had simply assumed that you could put any “background “ on a watch face and possibly select the appearance of the hands. I based this on my Fossil Collider HR which allows me to put any design on the face plus up to 4 complications, albeit using e-ink and physical watch hands. Complications and user designed backgrounds have been around on the HR for quite a while, so this seems like Apple playing catch up.
As the owner of two Garmin watches which allow third party watch faces, all I can say is thank the Cosmic Timekeeper that Apple doesn’t. Most of the ones I tried on my Epix were ugly, and there are fads that make most of them useless to me. (Haven’t even looked for the Fenix 5+.) For example, I don’t care about “steps”, and don’t need a watch face showing them, but there was a time when there were very few watch faces in the ConnectIQ “store” that didn’t have them and no way to customize what a face displayed. (I haven’t looked in a long while, so maybe the situation has improved with Garmin.) On my Apple Watch I use Modular with my favorite complications, and that’s perfectly adequate for me. I care far more for function than looks, though.
I don’t understand the concern. It’s not as if anybody is making you install, download or even look at them. But they’re available for people to search for and install, should they want to. And everybody’s sense of style is different - the same face you hate is one that somebody else loves.
Although I don’t use a smart watch, I use other skinnable apps. Like Firefox. There are literally thousands of themes, but they’re all optional. You have to actively search for and download them, and the product works just fine if you don’t (and want to stick with the default theme).
Nobody refuses to run Firefox because of the existence of themes, and I’ve never heard of anyone claiming that the Mozilla project should remove the capability to only permit their own curated themes to be installed.
So why should a watch be any different?
Join the discussion in the TidBITS Discourse forum