Rocket Insights recently shared the lessons they learned when designing an Apple Watch app for Virgin Pulse:

We created a Watch version of each of the features on the iPhone. We built clickable mockups and tested them. What we learned surprised us. Most features were not useful as Watch features...in fact most were actually harder to use there. […]

We discovered that the only features worth replicating were those that were actually better on the Watch. Not just the same, but easier, faster, or more delightful.

A big misconception for smartwatch doubters is that the smartwatch is supposed to replace the phone. But actual smartwatch owners tell a very different story:

When we asked them what apps they use and like on their Apple Watch the answer was invariably apps that delivered intelligent notifications. This was fascinating...that answer was consistent across the board. People are not using Watch apps as much as they are merely receiving quick notifications about their lives. So Watch software is, in a sense, a notification framework for your iPhone app. The Watch is merely the delivery device for it.

Rocket Insights also sums up perfectly how each Watch UI element should be used:

• Complications are for frequently changing data.
• Glances are for data that changes a few times per day.
• Notifications are for real-time updates.
• Apps are the very last thing a user interacts with, therefore all of the longer-term, slower-changing data should be left here.

I still see a lot of smartwatch doubters and Apple naysayers proclaiming that the Apple Watch is already a failure. The reality is, smartwatch app design is still in its infancy, and it takes a lot for people to rewire theirs brains and get it right — smartwatch apps are not just shrunken-down smartphone apps.

We're barely in the "fart app" phase of watch apps. But eventually, someone will figure out the next Instagram/Snapchat/Uber for the wrist.