Ask five agencies to quote the same MVP and you'll get a range wide enough to buy a house with. Founders read that as dishonesty. It usually isn't — it's that each vendor imagined a different product, because the brief left room to.
Start from the loop, not the feature list
Every product has one core loop: the sequence a user repeats that creates the value. For a marketplace it's list, find, transact. For a SaaS tool it's import, act, report. Write yours in a single sentence. Anything that isn't part of that sentence is not MVP scope, however obviously useful it seems.
Price the four cost centres separately
- 01
Core loop
The screens and logic that deliver the value. This should be the largest line and the one you protect.
- 02
Account & admin
Signup, roles, billing, and the internal panel your team runs the product from. Consistently underestimated, and consistently needed on day one.
- 03
Integrations
Payments, email, maps, third-party APIs. Each one is small alone and significant collectively — count them honestly.
- 04
Production readiness
QA, security, monitoring, deployment. Roughly a fifth of a serious build. Vendors who omit it are quoting a demo.
The three questions that expose a bad quote
- What is explicitly out of scope? A quote with no exclusions hasn't been thought through.
- What happens if we change our mind in week six? The answer should be a process, not a shrug or a threat.
- What do we own at the end? If the answer isn't 'everything, documented', keep looking.
Buy the plan before you buy the build
The single highest-return spend in early product work is a paid discovery sprint: one to two weeks producing a requirements document, clickable wireframes, a stack decision, a timeline and a costed plan. It typically costs a small fraction of the build, and it converts a range into a number.
It also gives you leverage. With a real specification in hand you can take the same document to three vendors and get three comparable quotes — which is the only way a quote comparison has ever meant anything.
A fixed price is only honest when the scope behind it is fixed too. Everything else is a guess wearing a suit.
A reasonable shape to expect
For a focused MVP with one user type, one core loop and a handful of integrations, eight to fourteen weeks of a small senior team is a realistic frame. Shorter usually means the production-readiness work has quietly been dropped. Much longer usually means nobody cut anything — and that is a scope conversation, not a budget one.
Want this applied to your situation?
A discovery call is 30–45 minutes, costs nothing, and ends with an honest answer — including if that answer is that you don't need us.
Book a Discovery Call