Some operating system releases mainly offer refinements and clean up details behind the scenes, like the mythical Mac OS X 10.6 Snow Leopard (in fact, Michael Steeber and Jeff Johnson suggest it was a myth). Other releases introduce major user interface changes, like iOS 7. watchOS 10 falls into the latter category. Many users have welcomed its interface changes, but others thought the old interface was just fine, thank you. This article is not a comprehensive review of watchOS 10; it is just an overview of the most notable changes, especially those with reassigned user actions. Then we’ll look at how and why Apple decides to change an established interface.
In watchOS 10, Apple updated the design language in most of the apps and watchOS screens to match iOS 17 more closely. Most notably, the company added the Smart Stack, with its customizable widgets. But Apple also changed how you access basic features using physical buttons, causing some users to exclaim, “Who moved my cheese?”
While watchOS 10 has enjoyed a generally positive reception—the Snoopy watch face is particularly popular—there has been some grumbling about how Apple redefined established actions to trigger new behaviors. It’s safe to assume that inside Apple, people also grumbled as familiar button presses generated entirely different results. But after using internal builds of watchOS 10 for months, Apple obviously decided the changes were worthwhile.
Apple is sufficiently proud of watchOS 10’s new Smart Stack collection of customizable widgets to make it a top-level feature. To ensure that users could access the widgets as easily as possible, Apple’s designers wanted to provide quick access from the watch face. Turning the digital crown is one of the most accessible actions on the Apple Watch, so Apple repurposed it to bring up the Smart Stack and scroll through the widgets.
The Smart Stack provides several interface improvements. Apple believes that most of the time, people are just looking for a snippet of information, like the current weather, and don’t need to open a full app. The Smart Stack comes with widgets connected with bundled apps, but you can add more, rearrange them, and pin your favorites to the top (touch and hold any widget in the Smart Stack to start editing). These widgets provide concise data, and you can customize the details by editing the widget. If you really want the full app, tap the widget; it doubles as an app launcher. Apple thinks the Smart Stack will provide the data you need most of the time, and it’s easier than opening an app.
In previous versions of watchOS, turning the Digital Crown from a watch face adjusted the watch face (for some faces). Now the crown brings up the Smart Stack; to adjust an interactive watch face like Kaleidoscope, you must tap first (and when you’re using the Siri face, turning the Digital Crown works only with that face and won’t bring up the Smart Stack). The engineers probably looked at user data and realized users seldom play with interactive watch faces once they’re past the novelty.
It’s unclear why Apple felt the need to add widgets to the Apple Watch, given that watchOS apps themselves are often widget-like companions to full iPhone apps and watch face complications provide at-a-glance access to even more discrete bits of information. When I worked on early versions of watchOS, we had no model for the “right” user interface for a tiny wrist-mounted screen, so much of what we were doing was throwing spaghetti against the wall to see what would stick. The introduction of Smart Stack widgets suggests that this may still be happening. I see three possibilities for how it came to be:
- Launching apps on the Apple Watch is awkward, requiring either a lot of scrolling in list view or tapping a tiny icon in grid view, and Apple’s analytics showed that most people seldom open apps. The Smart Stack attempts to surface the most relevant data from apps in a way people will use.
- Although complications provide another way of surfacing relevant data from apps, they’re tiny and fiddly to add, and Apple’s analytics may show that they’re infrequently used. Plus, quite a few watch faces, including the popular Snoopy, provide few or no complication spots. The Smart Stack lets users access data without relying on complications.
- Apple’s interface wonks may wield enough power to push something like the Smart Stack through, even if it doesn’t offer enough new functionality to make a difference. Not all interface decisions are backed by analytics.
Home Screen and App Switcher
Apple changed the Home Screen grid view of apps so it scrolls only up and down and no longer moves from side to side. I think this adjustment makes it easier to find apps on the small screen, although even easier yet is the list view that sorts all your apps alphabetically and doesn’t require you to see and interpret tiny icons. List view hasn’t changed much, though it now always defaults to the top of the list rather than remembering your position. You still enter the Home Screen by pressing the Digital Crown, and you can switch between the views using a button at the bottom.
To get into the App Switcher (which replaces the Dock), you now double-press the Digital Crown instead of pressing the side button. I think this button swap makes more interface sense because then you push the Digital Crown once to access all apps (Home Screen) and twice to access recently used apps (App Switcher). Recently used apps appear at the bottom of the App Switcher, represented by a static image of their last screen.
Previously, double-pressing the Digital Crown switched you back and forth between the last two apps, much like pressing Command-Tab on the Mac. Since most people probably don’t multitask much on the Apple Watch, losing that feature shouldn’t bother too much of the user base. However, transitioning from the Dock to the App Switcher means we lose the ability to pin and access favorite apps. Apple likely sees the Featured App widget in the Smart Stack providing that option, albeit only for three apps.
As a slight tangent, you can swipe left on an app in the App Switcher to force-quit it, just as you can force-quit an app in the iPhone App Switcher by swiping up. Some people do this routinely, mistakenly thinking they’re freeing up memory. In reality, iOS, iPadOS, and watchOS reclaim memory from background apps when more memory is needed. Force-quitting them in advance does nothing useful, and it’s actually counter-productive. The next time you launch that app, it must be loaded from storage, which takes extra time and consumes more battery power. You should only force-quit an app when it’s hung and not responding.
Control Center now appears when you press the side button, replacing the Dock. Previously, accessing Control Center required that you swipe up from a watch face; from other apps, you’d touch and hold, then swipe up. Pressing the side button is both more uniform and more discoverable. Apple’s user data probably showed that users frequently open Control Center, such as to ping a misplaced iPhone, so the engineers looked for an easier way to access it. (Remember that you can choose which buttons appear in Control Center—and in what order—by tapping Edit at the bottom of Control Center.)
Apple may have seen that people used Control Center a lot from the watch face but seldom accessed it directly from apps, instead returning to the watch face before invoking Control Center. Many people likely never realized they could touch and hold and then swipe up from an app. Or, Apple may have received more direct feedback about user behavior via Apple store employees, phone support, and marketing surveys. It’s even possible that Apple’s design group just made a unilateral decision.
In my experience, normal users rarely find interface elements that involve touch and hold. Apple tried incorporating the haptic 3D Touch technology on iPhone and Apple Watch models but eventually dropped it, presumably because it wasn’t used sufficiently to justify the added hardware cost.
Switching Watch Faces
Previously, swiping left or right on a watch face changed watch faces. In watchOS 10, you must touch and hold to enter a selection mode, then swipe to switch. (To add a new face, swipe left until you reach the New screen.) Apple’s user data may have shown people changing watch faces, then immediately changing back, suggesting an accidental change. It’s also likely that Apple support fielded calls from people who were confused when an accidental watch face switch completely changed their Apple Watch experience without them realizing, leaving them with no known way to get back. Overall, this change is likely for the best, although Apple bowed to user complaints and added a setting in watchOS 10.2 to revert to the old way for those who regularly switch between faces. Find it in Settings > Clock on the watch itself.
Gathering Data for Analytics
Apple has excellent telemetry on what people do on their Apple Watches and all other Apple devices. When you agree to share watch analytics with Apple during setup (you can adjust it later on the iPhone or Apple Watch in Settings > Privacy & Security > Analytics & Improvements), you’re helping Apple learn which apps and features are used the most, and which features users never seem to discover. The data includes how often you use your watch, which apps and features you use, which controls (Digital Crown rotate, Digital Crown button, side button) and gestures (tap, swipe, touch and hold) you use, and in what context. These stats tell Apple which parts of the product are used most and help Apple decide where to direct its limited development resources.
You might think a company as rich as Apple has essentially unlimited resources to do whatever it wants, but you’d be wrong. Throwing more money at a feature often doesn’t help since people are the bottleneck in creative work—and software development work is creative. Apple can’t necessarily find the developers it needs, or the developers with the necessary knowledge and experience to solve a problem correctly are needed elsewhere. Coordinating increasingly large teams and interlocking features brings additional challenges, especially in a system as tightly integrated as the Apple Watch,
The list of interesting new feature ideas is always longer than the development team can achieve. Implementing a new feature takes more than just writing the code to support it. The feature must be tested, documented, and translated into multiple languages. Marketing copy must be written, as well as training content for tech support and Apple store employees. Finally, Apple is committing to supporting this new feature for years to come. When people ask, “Why doesn’t Apple just do my favorite feature,” they don’t understand how significant a commitment that really is.
When thinking about analytics, it’s worth remembering that the data Apple collects is completely anonymous, and the company endeavors to collect the minimal amount of data possible. When I worked as an engineer on the Apple Watch, employees were trained on Apple’s security and privacy policies. These included collecting the least amount of data that satisfies the business need, never collecting additional data, and absolutely never collecting personally identifiable data.
When I added code to watchOS to collect user data, I wasn’t even allowed to check it into the source code repository until it had been reviewed by the privacy engineering panel to ensure I wasn’t accidentally collecting personal or extraneous information. Plus, when I reviewed the collected data, I couldn’t see data from individual Apple watches, only aggregate data. Apple may not be perfect, but the company tries very hard to protect user privacy.
How does Apple decide when to overhaul a user interface? It’s both art and science, with a dose of business and politics thrown in. Interface designs start out looking fresh and clean, but over time, they lose their luster and eventually feel stale, like that bathroom remodel from 1995. People buy new devices because they look vibrant and convey style. This is why car companies overhaul a car’s design every 5–8 years, even if the old design worked well. When a new model looks different, suddenly your old device, which was fine yesterday, now seems dated, encouraging replacement. Apple is in the business of selling new hardware, but there’s much less room to change the look and feel of hardware than software.
Occasionally, interface design changes because someone new takes charge. iOS 7 included a major interface update that was as much political as technical. Jonny Ive, who had led Apple’s hardware design team for years, took over software design too and wanted to leave his mark. He ripped out the old skeuomorphic design in favor of a clean, modern aesthetic. Unfortunately, he went too far in preferring form over function. Users had trouble deciphering which interface elements were controls and which were just labels. Limited contrast hindered people with low vision. iOS 7.1 cleaned up the most egregious interface missteps.
Apple usually tests new interface ideas internally. When a new feature is proposed, designers sketch out a half dozen different interface approaches. Product managers, engineering, marketing, and management give their opinions, and a few interface ideas are coded up to try out. Employees on the product team try out the new features on their devices. They report bugs and discuss whether the interfaces are working well. The designs are refined, the least successful ones are dropped, and employees try the updated versions. It’s an iterative process, with more employees added in each round. Ultimately, hundreds, if not thousands, of employees use a new feature before it ships to customers.
Apple knows it inconveniences users by changing an established workflow. The fact that Apple rarely does this lets you know how seriously the company takes it. Nevertheless, for the most part, Apple truly believes that the new design is—or at least will eventually be—better than what it replaces.
For instance, when Apple released Final Cut Pro X with significant changes from the workflow in Final Cut Pro 7, angry users complained about the new approach and the lack of features from the previous version. However, after Apple added back those features and smoothed the rough edges in the first six months after release, most of the complaints died down. Many video editors now agree the Final Cut Pro X workflow is better. (But some still prefer the old workflow!)
Similarly, when Apple’s Swift programming language was new, its developers regularly added features to improve it. But after a few years, the Swift developers decided it needed some rethinking, so Swift 3 incorporated significant changes. Most old code wouldn’t compile anymore. My team spent weeks rewriting our code to work with Swift 3. There was plenty of grumbling. But over time, I realized the changes were all good. Swift 3 code was easier to write and easier to read.
Apple is willing to take—and inflict—short-term pain for a longer-term gain. Whenever any change makes it to users, you can be sure that employees have been living with the changes internally for months (“eating your own dog food,” in the industry parlance), and the product managers have decided that the new way really is better, once you reprogram your muscle memory and expectations of how things “should” work.
An iPod Story
Years ago, I worked on the iPod’s Voice Memo feature. The user had to click through three screens to start a recording. Steve Jobs didn’t like it. He thought it was too complicated. But try as we might, no one could come up with a more straightforward design that wasn’t missing key features. Steve didn’t know how to fix the design either; he just knew it wasn’t good enough. Even though the voice recorder worked, it didn’t ship in that year’s iPod model.
Several months later, someone on the team had a brainstorm to simplify the interface while preserving all the features. The user had to click through only two screens to start a recording. The Voice Memo feature shipped a year late, but the interface was much better. In my opinion, that was the right tradeoff.
Users never knew how much effort went into simplifying that interface. A good interface feels obvious, but it usually takes intensive work on the part of many people. In a quote commonly misattributed to Mark Twain, the 17th-century French mathematician and philosopher Blaise Pascal said, “I have made this longer than usual because I have not had time to make it shorter.” The same applies to user interfaces.