Sustain

Sustain came out of a structured market research pass evaluating solo-buildable app ideas against real App Store data and competitor teardowns. A GLP-1 companion app ranked highest: fast-growing user base, high willingness to pay since people are already spending $200 to $1,000 a month on the medication itself, and every existing tracker built around weight instead of what these patient communities actually worry about most, losing muscle instead of fat.

Built solo, mostly with AI coding tools, SwiftUI and SwiftData, no third-party dependencies until billing and account sync forced two: RevenueCat and Supabase. The core feature is a camera-based protein estimate, point it at a plate, get a grams estimate in a few seconds instead of weighing food.

Here's the part worth being honest about. My first attempt at marketing this was a Reddit post pitching the "your medication dose affects your protein needs" angle, and a user tore it apart, said I was assuming people are idiots for not already knowing that. She wasn't wrong that the logic is obvious, what she was actually reacting to was the post reading like a thin excuse for a plug. I rewrote the messaging around effort reduction instead, the photo estimate, an automated weekly recap, rather than a correlation nobody needed explained to them.

Just shipped a 2.0: mandatory Sign in with Apple plus real account sync, replacing what used to be a fully local, no-account app. That was a real tradeoff, some users specifically liked that nothing left their device, but losing all your data on a phone switch with no way to recover it was a bigger problem in practice.

Free for manual logging. Subscription unlocks the AI estimate and full history. Genuinely curious how other people here think about the local-only versus account-required tradeoff, and whether you've had a piece of user feedback that stung at first but was actually right.