Find out before you spend everything.
An MVP is not a cheap version of the real thing. It is the smallest build that answers one question honestly: will people actually use this, and will they pay?
We build the smallest product that tests whether your idea works — usually eight to twelve weeks. Real software with real users, not a prototype. The scoping conversation matters more than the build: most MVPs fail by being too big.
Updated August 2026
The list is always too long.
What goes into version one, and what waits until you have evidence.
The hard part is deciding what to leave out.
Every founder's first list is too long, ours included. Everything on it feels necessary because you can picture the finished product — but the finished product is not what you are building yet.
The question an MVP answers is narrow: do people want this enough to use it, and to pay? Anything that does not help answer that can wait, including things that will definitely be needed later.
So we spend real time on scoping before quoting. It is the least billable part of the work and by far the most valuable, because a smaller launch means you learn sooner and still have budget to act on what you learn.
What you get.
Real software, not a prototype
Deployed, working, with real users. A clickable mockup cannot tell you whether people come back.
The one thing it exists to do
Built properly. Everything else built cheaply or not at all.
Payment, if that is the question
The only reliable test of willingness to pay is asking people to pay.
Analytics from day one
Where people stop, what they never touch. Without this you launch and learn nothing.
A way to talk to your users
The first fifty are worth more as conversations than as numbers.
Foundations that do not trap you
Standard technology, sensible structure. Fast, without deciding your architecture for the next five years.
Honest scoping first
Including telling you when something does not need building at all yet.
Code you own outright
Yours from the first commit. Continue with us, your own team, or anyone else.
What to cut, and what never to cut.
The decisions that decide whether an MVP launches in ten weeks or twenty-six.
| Version one | Later | Why | |
|---|---|---|---|
| Core action your product exists for | Yes | — | Without it there is nothing to test |
| Payment, if you are testing willingness to pay | Yes | — | Intent is not evidence |
| Sign-up and accounts | Simplest possible | SSO, teams, roles | Necessary, but not the question |
| Admin panel | A database view is fine | Proper tooling | Only your team sees it |
| Settings and preferences | None | Yes | Nobody churns over missing settings |
| Onboarding flow | Minimal | Designed properly | You do not yet know where people stall |
| Mobile app | Rarely | If usage proves it | A responsive web app tests the same idea |
How we scope.
If your list will not fit, we cut it with you rather than quietly extending the timeline.
Launching without measurement is the most common way an MVP teaches nothing.
If you already have paying customers doing this manually, you do not need an MVP.
When an MVP is right.
You have an idea and no evidence
The classic case. Build the smallest thing that produces evidence.
You are raising and need traction
Real usage numbers move investors. A deck describing a product does not.
You do it manually today
The best starting point there is — you already know the process works.
Not if you already have proof
If people are paying you for this manually, build the real thing. We will say so.
How we work.
We work out the question
What one thing must be true for this to work? Everything after this is in service of answering it.
We cut the list, together
The hardest conversation and the most valuable. Expect to remove more than half.
We agree scope and date
Fixed, in writing. An MVP with a moving scope is not an MVP.
We design the core flow
Clickable, so you can feel it before it is built. Days, not weeks.
We build in two-week blocks
Working software you can use, every fortnight. No surprises at the end.
We launch and instrument it
Analytics live from day one, because launching without measurement wastes the whole exercise.
We look at what happened
What people did, not what they said. Then we decide together what version two is — or whether there should be one.
Frequently asked.
6 questions answered. Still have one? Reach out.
Less than you would spend finding out the same thing the hard way. We do not publish figures because the range is genuinely wide and depends almost entirely on scope. What we can say is that the scoping conversation moves the number far more than any technology choice does, and that is a conversation we have before quoting rather than after.