Sleep audio for shift workers

Stage 1 of 5

Validate

Is this a real problem for a customer you can reach, who wants it solved?

What the founder did: talked to 13 night-shift workers: nurses, two paramedics, a security guard and a bakery worker. 11 had tried a sleep app and given up, mostly because the content assumed a night-time bedtime. Then a pre-order page went up at $4 a month, shared in three nursing groups. 210 people visited and 31 paid for the first month up front.

What the guide flagged: Calm and Headspace each have some sleep content for shift workers, so "nobody does this" wasn't true. What the interviews kept coming back to was the schedule: sessions sorted by where you are in a rotation, not by mood. The guide also flagged that this is audio, not medical advice, and the page should say so plainly.

Why Go: people didn't just say they'd pay; 31 of them paid, from one page and three posts. The audience gathers in a few places the founder already belongs to. Weak spot noted: at $4 a month it takes hundreds of subscribers to matter, and recording new sessions never stops.

Stage 2 of 5

Define

What's the smallest thing you can build that proves value?

What got cut: the first list had 7 features: a sleep tracker, shift calendar sync, a community feed, alarms, personal plans, breathing exercises and the audio library. Kept: the library, sorted into after-shift, before-shift and switching-back sessions, plus a subscription and an end-of-shift reminder. The tracker and calendar sync went to Later.

Why Go: the library is what the 31 people paid for. Everything cut either needs data the founder can't collect yet or turns a small audio product into a health app. The target is 60 paying subscribers by the end of the first month after launch.

Stage 3 of 5

Build

Can you build it with AI tools without losing control?

Build path chosen: an AI app builder for the web app and a subscription payment service, with the audio files on an ordinary hosting service. No app store: it's a website that saves to the phone's home screen, which also avoids the app stores' cut of subscriptions. Recording the first 18 sessions took longer than the build.

What the guide flagged: the first version let anyone with a session's link play it without paying, and cancelled subscribers kept their access. The guide flagged both before launch. The fix for both was the same: check the subscription each time a session starts, not only at sign-in.

Stage 4 of 5

Launch

Is it safe to put in front of real users, and how will they find it?

What the guide flagged: the readiness check found that the reminder sign-up collected email addresses with no way to unsubscribe, which Canada's anti-spam law requires. It was fixed the day before launch, along with a backup of the audio files outside the hosting account.

What happened at launch: launched to the 31 pre-order customers first, then posted in the same three nursing groups and two paramedic forums. In the first two weeks, 74 people subscribed at $4 a month, about $296 a month. 6 cancelled in the first week, most saying the sessions were too long for a nap before a shift.

Stage 5 of 5

Learn

Is it working, and what should you do next?

What the numbers said: after two weeks: 74 subscribers, 6 cancellations, and the average subscriber played 9 sessions. Before-shift naps were played the most, even though there are the fewest of them. The target of 60 paying subscribers in the first month was passed in week two.

What the guide flagged: the cancellations all pointed the same way: shorter nap sessions. The guide suggested recording three 5-minute ones before adding anything else, then watching whether cancellations drop. The bigger question for Learn is whether subscribers keep paying past month two, once the novelty is gone.