How Google Should Rescue WearOS

Juhani Lehtimäki
Mar 5, 2019 · 11 min read

Android Wear, the first release, was exceptional when it comes to Google software launches. It’s no secret that Google releases early and iterates out in the open (unlike companies like Apple who wait until they believe they are complete before showing any of it to the public). This time though, Android Wear was great from the first launch. At least the idea was good. Naturally there was technical issues to be solved and the partner hardware wasn’t very good.

History of Wear

For context, I want quickly go through the history of Android Wear, or Wear OS like we now know it.

Fundamentals of Wear 1

The fundamental idea of Wear 1 (Android Wear) was that the small screen on your wrist is an extension of your phone. You weren’t supposed to do much on it. The Wear 1 devices improved and extended the already powerful Android notification system. It was easy to see your notifications on glance and react to them quickly. While standalone apps were possible they were discouraged. Most apps didn’t have to change anything to add support for Wear as the system used the standard notifications and actions on them.

The fall, Wear 2

The Wear devices probably did not sell very well. Also, public wasn’t convinced about the need for a wrist device like that. Apple’s first iteration device wasn’t very good either and wearable market was struggling. Companies like Motorola withdrew very publicly from the market after making some of the best and most liked Wear devices and Samsung shifted to their own platform (Gear) after being a launch partner for Android Wear.

Same times clueless tech blogs were spamming headlines slamming Wear devices for “not doing anything” completely missing the point of extending your phone’s already existing capabilities instead of creating new ones.

My yardstick of a joke of a wearable was, from the very beginning, existence of a keyboard on the platform. When Samsung introduced one on their Gear platform I could not stop making fun of them. Needless to say.. Wear now also has one.

If you have to put a keyboard on your wearable something has failed in the design.

To me Wear 2 was the fall. It made me stop using Wear devices altogether for a long time. In fact, I went through almost a year using a Samsung Gear device and thereafter Apple Watch for a while. There is a lot to love on both of these competing device but neither of them are perfect (or really even very good in the long run).

What I want Wear 3 to be

Okay, enough whining! What should the Wear be? This is the list of things I would change if I was the dictator of Wear OS. Google, if you’re looking to hire a Wear OS dictator just email me I’d be happy to dictate. But maybe it’s easier just to use the notes below ;)

Rethinking the role of the watch face

Let’s stop trying to replicate analog watches and rethink what a watch face should be. It is the main screen of your wearable. It should look great, but above all provide the information you need.

Watch faces should be contextual. We should have a specific watch face when media is playing and another one when doing sports and yet another one when commuting. Google is great at detecting our context already, let’s utilise it.

Both Apple Watch and Samsung Gear are trying to do this in a sense but are not quite succeeding. Let’s open the APIs for developers and let people go crazy with providing the best functional and best looking watch faces for all situations. Let user pick which one they want to use in whichever situation and we’re good to go.

Complications must become centrally controlled

Complications is a simple idea. The potential is there in the current Wear iteration but it’s largely underutilised. In the current Wear versions each watch face creates its own complication configuration UI and some do not support them at all.

The complication setup UI must be unified on the platform level! The way you select complications should not be dependent on which watch face you have. Once you decide that you want to see certain info in certain context (see above) that’s what should always be shown!

All apps should expand to the watch

Google should make it much easier for app developers to support Wear. Let’s be honest, there’s not a huge amount of money to be made on building Wear apps at this moment. To justify the investment on supporting Wear either the return must be big (very rarely the case) or the investment small. We should start by minimising the investment.

The APIs to send data to Wear are not very good in the current form. Also, building and distributing standalone apps are too much work (why is it so difficult to deploy this stuff to Play Store).

First step is to take a look how apps could expand to the Wear devices without any code changes. Notifications are already working well but is there something else we could use without forcing app developers to update their apps?

Slices provided by apps or Actions on Assistant should be utilised on Wear. This is content that is not tightly coupled to the way it is show. Exact implementation probably needs a bit more thought but taking everything that can be taken from the phone to Wear should be done. That way the Wear ecosystem isn’t empty even if app developers do not invest into it. This is the first step towards solving the chicken-egg problem.

A better designed and easier API is needed for exposing data to Wear. This is where we should look at the awesome Firebase APIs and build something similar. Apps on phone and Wear should only care about writing data to somewhere which then gets synchronised to the other side automatically. Exactly like writing and reading from Firebase Realtime Database.

There should be a very simple-to-use API for exposing data to complications, directly from phone, without having to build anything for Wear specifically.

App developers should be able to simply push to an opaque API information about their app and what they want to show on complication on the Wear. For example, on our Social Steps app, I should be able to push my step count update whenever it is refreshed and things like your position. The API should define a semantic way for the data to be published so watch face complications can combine and format the data in a way it fits to their style.

Stop pushing the idea of standalone Wear apps. All apps should extend a phone app and only become installed to the Wear device when the phone app is installed to the user’s device. Apps should be able to provide a simple companion app and show data on the Wear if needed. In most cases this would be a simple screen showing a status at glance etc.

With the above mentioned simplified Firebase-like API app data transfer will be easy and app development time small.

Our Social Steps app has a Wear component but building it was too big investment for it to make financial sense.

All apps should be standalone temporarily only. For example sport trackers should work even if you leave your phone home but would mostly require a (Automatic) sync via phone before having the data synced to the corresponding cloud services.

Runtastic on Wear makes sense when exercising.

If I log in to an app on my phone I should automatically be logged in to it on my Wear as well. No prompts, no questions. This should all be transparent to both, app developers and users. Same applies to app permissions. Only ask permissions on my phone and apply the same ones to the Wear app.

I think it’s nowhere as apparent that Wear is a side priority for Google than in the Play Store deployment process. Multi-APK, custom app versions and retaining APKs is a massive pain and very confusing. This shouldn’t be the case. Either go back to the old system where the build tools generate an APK inside the main APK or create a UI where I can drag & drop my Wear app as part of my main app.

Forget iOS support

I know that a big portion of Wear users actually use it to connect to an iPhone. It’s nice but I feel that supporting iOS forces a bit of a lowest-common-denominator approach. There’s just no point trying to support a platform that won’t be implementing any of the required systems in their apps. Simply forget iOS as a supported platform. Few extra sales are not worth ruining the whole platform.

Eliminate support for rectangular screens

Round watch feels right. Most of the Wear ecosystem new devices are round. In fact, I don’t remember seeing any rectangular watches releasing lately. It’s time to eliminate rectangular support. Let’s let developers focus on supporting, and optimising for, circular screens. This way apps and watch faces ship with best optimisation for the devices they are meant to be running on cutting down development and testing time and lowering the threshold on investing on building Wearable support on their apps.

Remove Play Store from the watch

Play Store should never have been in the watch in the first place. Remove it, remove it now!

Make everything depend on the connected device the way Wear 1 worked. When I install an app that has a Wear component simply install it on my Wear. When I update it, update it on my Wear. There’s no simpler way of doing this!

Move all configuration and settings to the phone

Configuration, settings etc should (almost) all be only on the phone. I don’t want to be changing app permissions or something similar on the tiny screen when I have a larger screen right there? Keep things like screen brightness or cinema mode on the Wear but remove everything that is possible to remove from it.

The Wear companion app has always felt like it’s been ignored or been a second priority. It really needs more love. I need to understand what is going on, change settings, set complications etc on my phone, not on my watch.

Simple things like showing new watch faces first are still missing on the Wear companion app after years of it being available.

Watch faces appear in random-feeling order. Finding the one you just installed is nearly impossible.

Refocus on notifications

Our smartphones are communication tools. So much so that we nowadays receive overwhelming amount of notifications. In some reviews wearables got bad rep for becoming yet another device interrupting you constantly.

We’re flooded with notifications. A wearable could help us filter them effortlessly.

A wearable should support you to filter and manage the incoming notifications with ease and with less or lesser interruptions than without. The wearable should allow you quickly and easily filter the low priority notifications and help you to react to the ones that need your immediate attention.

Android’s notification system is very powerful. It comes with a lot of inbuilt features like actions, inline replies, pictures, media mode etc. That is a powerful platform to build on. The best part is that apps don’t have to change anything to participate. As long as an app uses the Android system correctly the extended notification system would just work.

Assistant, Google Home, Auto etc

Google Assistant is becoming an ecosystem on its own. I use it everywhere; from turning on my lights to navigate home in my car. One place I do not use it is in public. I still refuse to talk to my devices outside my home and my car. Wear could change this. Assuming the Assistant is very well and thoughtfully integrated (easier said than done) it could become the primary interface of the wearable device. Or at least the main input method beyond simple touches.

Focus on smoothness

Android 9 on a Pixel 3 is the smoothest mobile system I’ve ever used. By smoothness I don’t mean just framerate or how fast apps open (although that certainly plays a big part in it). By smoothness I mean the general feel of the system. Everything from meaningful and fluent notification animations to fluent app switching. You never have to second guess your gestures or taps. Everything just works, and works quickly. Transitions are coupled with meaningful animations helping you understand what is happening and where. There is no unnecessary waiting times and spinners in the system apps or the launcher. The system also predicts my behaviour beyond anything I’ve ever seen before. In general the system is smoooooth.

I want to see the same to continue on my wearables.

Apple Watch users like their devices. I like Apple Watch too. It’s currently the best smart watch on the market. It’s smooth and pleasant to use. However, I don’t think it is right direction for wearables in general. I think it is liked only because it compares favourably to anything else available out there, not because it is inherently good.

Photo by bady qb on Unsplash

To induce same amount of smoothness to Wear as we experience on our Pixels Google has a long and difficult road ahead of them. A big part of the game is the hardware. The current generation of Wear devices (even with the latest Snapdragon 3100 chip) are laggy, slow and uninspiring. A lot of work is needed on the hardware side but there is hope as Google recently brought a big chunk of the Fossil’s smart watch division under Google. I hope this move indicates a serious investment to hardware.

Tapping on small hit targets is annoying. Especially on a tiny and constantly moving screen. Rethinking interaction to prioritise swiping and hardware controls might be worth the effort. Apple Watch’s digital crown isn’t exactly the most revolutionary innovation but it has been thought through carefully creating a pleasant and accurate interaction. Google’s Wear devices need to add same level of polish to all interaction patterns. In the small-scale, low-end processor-powered devices lack of polish is multitudes more visible and less forgiving than on higher power devices where minute mistakes can be easily compensated.

So, conclusion I guess

Some people want to see the death of smart watches to validate their initial scepticism of the form factor. I would rather see them evolve into something much more useful. I feel that I saw a glimpse of potential in the Wear 1 but it was distinguished in the later iterations. Getting from the current generation of Wear requires more than just polish, it required multi-tier revolution of thinking. But it’s not hopeless. Google has the talent (just look at Android Pie) but do they have the will? It remains to be seen. I’m hopeful but also patient. Changing things will take time.

Anyways.. he’s our puppy:

Leeloo the frenchie puppy

Snapp Mobile

Snapp Mobile Engineering and Design Blog

Juhani Lehtimäki

Written by

Founder and Android lead at Snapp Mobile. http://snappmobile.io/ Co-lead of GDG Android Munich. Author: Smashing Android UI. Android GDE.

Snapp Mobile

Snapp Mobile Engineering and Design Blog

Welcome to a place where words matter. On Medium, smart voices and original ideas take center stage - with no ads in sight. Watch
Follow all the topics you care about, and we’ll deliver the best stories for you to your homepage and inbox. Explore
Get unlimited access to the best stories on Medium — and support writers while you’re at it. Just $5/month. Upgrade