Play

Play was an iOS-native prototyping tool for product designers. Six people built it in four months. It grew to over ten thousand designers through word of mouth, won an Apple Design Award in 2025, and was acquired by Apple in June 2026.

I was part of the design team. I worked on the product direction, the interaction design, and what the tool felt like to use. That last part was the hardest one.

Every designer who has ever handed off a prototype knows the problem. It looks right, it clicks through fine, and then engineering builds it and it feels completely different. A picture of an app is not an app, no matter how right the specs were.

Figma's prototyping mode fakes native behavior with animation curves over static screens. Building a throwaway Swift prototype often took longer than shipping the real feature. Neither option let you test what the thing would actually feel like before a single line of production code existed. That gap was the whole opportunity.

Six people, four months. No long research phase, no room for it. The team had a strong bias toward shipping fast, so the operating assumption from day one was: build the smallest version that could earn trust, then learn from real use.

The harder constraint was the audience. Our users were professional designers. They would reject the tool the moment the tool itself felt unconsidered. You cannot build a design tool for designers and ship anything that looks like a placeholder. The product had to be good on its first impression or it was done.

Most prototyping tools layer animation curves over static screens and call it native. Play actually ran on device. Real gestures, real spring physics, real haptics, real scroll behavior. It felt like iOS, not an approximation of it.

That distinction matters more than it sounds. A swipe-to-dismiss that is a millisecond off, a keyboard that does not push the view the right way, a scroll that loses momentum too fast. These things break the illusion immediately for anyone who has spent real time on the platform. Our bar was not "good enough for a demo." It was indistinguishable from the real thing.

Designers clock your easing curve from across the room. They feel scroll inertia and have an opinion about it. They notice when an import flow is lazy. They notice when the empty state got no love. You cannot get away with "most users won't notice" when your users are the exact people who notice everything.

That raised the bar on every screen, including the ones most tools treat as plumbing: import flows, project settings, export options. We treated those with the same care as the core canvas. The bigger product bet was restraint. A smaller set of things that worked beautifully beats a larger set that works adequately. We cut scope aggressively and made the things that shipped actually feel good.

We launched with a deliberately narrow scope: import a design, define interactions, preview natively, share a link. That was it. More advanced authoring came after the core loop held up under real daily use. Ten thousand users, mostly through word of mouth inside the design community.

That is the metric I trust most from this project. Designers recommend things to other designers only when the thing actually works. You cannot buy that with marketing. My working hypothesis was that word of mouth would only compound if people kept coming back after the novelty wore off, so week-four retention was the number I actually watched, not signups. It held at 63%, well above what a typical early-stage pro tool sees in that window, which is what told me the native feel was landing and not just impressing people once. In 2025, Apple gave Play a Design Award. That felt like the community's read made official.

Apple acquired Play in June 2026. The fit makes sense. Apple has spent years trying to lower the barrier to building for their platforms, approaching it from the developer side. Play was the answer built from the designer side. When your product already speaks the language of a platform at that level of fidelity, the conversation tends to happen.

The product is offline now. What I think happens next is that some version of what Play was doing ends up inside Apple's own tools. The problem it solved is real and it is not going away. If anything it is more relevant now than when we shipped it.

Building a tool for your own profession is the hardest design brief there is. You lose every excuse. "Most users won't notice" stops working when your users are the people who notice everything. Every screen has to be the work, including the ones nobody sees in a demo.

Shipping small on purpose is a real strategy, not a compromise. We cut scope aggressively and the core loop was tight because of it. That tightness is what made the word of mouth happen.