Product venture · iOS and watchOS
DO: instructions you follow without picking your phone back up
A native iPhone and Apple Watch app that turns a written procedure into a run you execute from your wrist. The phone is where a workflow gets prepared. The watch is where it gets performed.
- Role
- Product Designer · Builder
- Platform
- Native iOS and watchOS
- Year
- 2026
- State
- Working build, not shipped
30 Second Summary
Product
An iPhone app for writing a procedure and an Apple Watch app for executing it one step at a time, including counted steps, timers and safety cautions.
Design Move
Treat the watch as the execution surface rather than a smaller phone: one instruction on screen, and one action pinned where scrolling cannot move it.
Built
Both native apps, workflow authoring and editing, phone to watch synchronisation, timed and counted steps, pause, resume, session restore after termination, and completed runs synced back.
State
A working build stabilized through simulator QA across representative iPhone and Apple Watch configurations. Not shipped, and not validated on physical hardware.
The instructions live on the one device your hands are least free to hold.
Flat pack furniture, an espresso machine backflush, a camera rig. The steps are on a phone you have just put down to use both hands.
The phone holds everything the wrist should not have to.
Writing a workflow, reordering steps, and attaching counts, timers and cautions all happen on the phone, where there is room to think.
One instruction. One action. Nothing else competing for the screen.
The watch shows the current step and the single control that moves it forward. The step counter sits beside the system clock so content never collides with it.
A safety note that can only be half read is barely a safety note.
Some steps carry a caution. On the smallest watch, an instruction, a full caution and a control genuinely do not fit together.
A step is not always simply done.
Some steps are a count. Some are a wait. The pinned control changes to match the step rather than staying a generic button.
The interface was the easy part.
A deeper simulator QA pass exposed product defects that static screen review did not. The most important ones were not visual.
What is built, what was tested, and what is not being claimed.
DO is a working build and a frozen concept. It is not on the App Store, it has no users, and it has never run on a physical Apple Watch.
The constraint did the designing.
Almost every decision in DO was forced by a 42mm screen and an operating system that suspends apps within seconds. Pinning the action, giving a caution its own moment, anchoring timers to absolute time, and writing through on every change were not stylistic choices. They were what survived contact with real content on a real screen size, and the build only became trustworthy once I stopped reading the code and started running it.
