Case study 01 · App Factory
7 apps live on the App Store in 58 days. One operator.
We treated app building as an operations problem, not a creative one. The result is a pipeline where AI agents do the repetitive work, a kill test decides what gets built, and every release goes through the same gates.
Building one app is a project. Building many is an operations problem.
Most small studios lose time in the same places a client's operations team does: ideas nobody validated, work retyped between tools, releases that depend on one person remembering every step, and no clear signal for what to stop.
We wanted to know whether the Reiiel method holds up when the "business" is a product studio. So we ran it on ourselves.
A kill test before any code
Every idea is scored against the same market criteria before it earns a build slot: who switches to it, why they switch, what the acquisition channel is, and what could kill it. Most ideas die here, cheaply.
In the second batch, 30 ideas went in. Of the seven finalists that reached validation, five were killed on evidence before a single screen was designed. Some ideas never make the list at all. A mushroom identifier, for example, was rejected outright because a wrong answer could hurt someone.
A pipeline with gates, not heroics
A local control dashboard runs the factory (see the dashboards case study). It tracks every app and build, runs one-click builds, records build history and agent costs, and monitors the store listings. Specialised agents handle defined jobs, and a person approves anything that ships.
The rule we apply to client systems applied here too: reuse the infrastructure, never the product. Release tooling, analytics and security are shared. Each app's brand, navigation and experience are deliberately its own.
The output, dated and public
Every one of these is live on the App Store today. Tap any of them to check.
Same machinery, pointed at your operations
- Kill criteria before build. Your automation ideas get the same test, so you spend on the workflows that pay back.
- Gates, not memory. Every release in your business follows a checklist the system enforces, not one a person has to remember.
- Agents with defined jobs, people who approve. AI does the repetitive work; a human signs off wherever money or customers are involved.
- A control view. One screen that shows what's running, what it costs and what to cut.
Related reading: What shipping 7 apps in 58 days taught us about operations