How much should I build for my MVP? Cut to one loop

A feature list aims at being complete. A scope, meaning the decision about what goes in and what stays out, aims at learning fast. Those pull in opposite directions. Every feature you add is another week before you find out whether anyone cares.

In our view, building has become easier for people who do not code. Faster building is good news. It also makes it tempting to skip the question of what to build and just ship everything.

Build only the loop, or even fake parts of it by hand, instead of the whole app. Ask which journey, done start to finish by a real person, would make your main guess true or false. Then build only that.

Here is an illustrative example, not a real founder. Say you are building a booking app for physiotherapists. Your list has ten features: booking, calendar sync, intake forms, treatment notes, invoicing, payments, text reminders, a client portal, an exercise library and video calls.

The loop might need only one of them. The physio enters the week's appointments. Each client gets a text reminder with a one-tap confirm. The physio sees who confirmed without chasing anyone. Your test could be a handful of physios running that loop on real appointments for a few weeks. Everything else waits. It still matters, but proving the core idea matters first.

The riskiest assumption is the belief your idea quietly depends on. It is hard to see because you assumed it before you started.

In the physio example, interviews might show that missed appointments are a real pain. That feels like proof. The real bet is different. Will a busy physio actually type appointments into a new tool every week? Agreeing in an interview that a problem exists is not the same as changing a habit.

In our view, most ideas lean on one of three beliefs: the market exists, people will pay, or people will change how they work. Find the one carrying the most weight. Then build the shortest path that tests exactly that. If you have not checked the idea itself yet, start with how to tell if your idea is good or you are just excited.

Anything the loop can run without. Notifications, search, account settings, payments: all reasonable, and all cut if the loop works without them. You can often fake a step by hand at first, for example sending reminders yourself.

For a solo founder, think in weeks, not months. Every week spent on a supporting feature is a week your real question stays unanswered. Once the loop is clear, picking tools gets simpler too, see which tools a non-technical founder should build with.

The strongest case against it: some products cannot be tested with one thin loop. A marketplace, a site that connects two groups such as buyers and sellers, needs both sides present at once. Products that handle health data, money or children may need safety and privacy work you cannot cut. In those cases, the smallest honest version is bigger.

The missing fact that changes the answer is what you already know. If you have paying customers, how they actually use the product beats any scoping theory, including this one. And a scope does not replace talking to people. It tells you what to build so those conversations test the right thing. For that part, see how to find real users to talk to.

MVP Scoper asks about your idea and suggests one loop to test and what to leave out for now. Treat its read as a starting point to argue with, not a final plan.

Write your loop in one sentence: "A user does X, then Y, and gets Z." Next, cross out every feature on your list that this sentence does not mention. Then name the one belief the loop tests and the result that would prove it wrong. If you cannot write that result down, the loop is not ready to build.

What if the smallest version still feels like too much work?

That is a signal, not a failure. Either the idea needs cutting again or you need help with the build. Do not shrink the test below the point where it proves anything just to hit a date.

Does the scope change if I get a cofounder?

Yes. More hands make it faster and also make it tempting to build more. Redo the scope whenever the team or the main guess changes.