For most of software history, custom apps had to justify themselves across a multi-year horizon. Nothing short of we will use this every day for five years could pencil out. So we built generic platforms — SaaS — that amortized the cost across a million users. Anything too specific to one company, one event, one family, one trip simply didn't get built. People made do with spreadsheets, group texts, and a shared note nobody could find later.
That wasn't a law of nature. It was a consequence of the cost curve. And the cost curve has moved.
The default assumption is out of date.
When a small, purpose-built app can be designed, built, and deployed in days instead of months, the scoping question shifts. It stops being is this worth committing to forever? and becomes is this useful enough for the next six weeks?
Very few things are useful enough to feed and maintain for five years. A lot of things are useful for six weeks. The gap between those two thresholds is where the interesting work now lives.
What gets unlocked.
An incomplete list of things we've been asked about in the past year — each unthinkable to commission five years ago, each reasonable today:
- A single wedding. RSVPs, day-of schedule, venue map, shared photo dump. Lives four months. Dark after the thank-yous.
- A renovation tracker for one project. Punch list, photo log, paint and finish selections in one place. Useful for the build. Archived at handoff.
- A conference companion for one three-day event. Schedule, attendee directory, side-meeting requests, the actual room map. Deployed two weeks before, sunset the Monday after.
- A pop-up shop POS for a holiday weekend. Tailored to the inventory, the tax setup, the one vendor's checkout flow. No subscription that outlives the booth.
- A summer-camp parent app for one season. Daily photos, pickup logistics, the dining hall menu. One summer. Done.
- A field-research data collection app for one study. The exact form fields the protocol calls for, offline mode, the export the analyst wants.
- A move tracker. Box inventory, room-by-room, who-has-what after the move. Useful for six weeks. Not a forever service.
- A reunion app. Family, school, one-off gathering — directory, schedule, shared photo dump. A weekend of utility, then nothing.
None of these are big enough to be a real product. None durable enough to be worth a subscription. Most almost solve themselves with group texts and a shared doc — uncomfortably, with the host quietly doing manual coordination work. A purpose-built app would fit each one perfectly, except that purpose-built apps have historically cost too much.
Why this works now.
A few things had to change at once, and they did.
The build got fast. A small purpose-built app is now days of work. AI handles the boilerplate, scaffolding, auth wiring, deploy config — the parts identical across every project. The bespoke part — design judgment, workflow, taste — is still us. The labor that used to dominate the build dropped sharply.
Distribution stopped requiring an app store. A web app that lives at a URL, adds to a home screen with one tap, works offline, pushes notifications — that does the job for nearly every short-lived purpose.
Hosting got essentially free. A small app for a hundred users costs almost nothing to run. The fixed monthly cost is no longer the barrier.
Disposability is a feature.
There's a temptation, when you can build something cheap, to keep adding to it — to turn the wedding app into a startup, to grow the renovation tracker into a platform. That's almost always a bad idea. It's why most cheap-to-build software ends up expensive: nobody actually retires it.
The freedom of a disposable app is the disposability. You build it for a known purpose, use it for a known window, turn it off when the purpose ends. No upgrade treadmill. No support burden. No what should we do with this thing meeting eighteen months later.
Software is normally something you own. A disposable app is something you rent for a defined window — except the rental is from yourself, and the price is the build.
What this changes for us.
The work at the studio has split into two clear shapes. There's the durable stuff — software that needs to live for years, often inherited, often quietly load-bearing. And there's the disposable stuff — small, custom, fit for a specific moment in a specific life or a specific business's calendar.
The disposable category is, right now, the more interesting one. It's where the cost curve has changed most dramatically. The only thing standing between most people and a perfectly-tailored app for the thing they actually care about is the habit of assuming custom software is too expensive. That habit is out of date.