Scattered scope cards settling into a roadmap, then one card turning into a live screen.

Every project starts with a fixed-price discovery sprint. We talk to the people who will use the product, test the riskiest assumption with a prototype or a technical spike, and write down what the first release contains and what it leaves out. You get a written estimate for the build, as fixed milestones or a monthly team. Then we build the MVP in weekly releases you can click.

A good fit if

  • You have an idea, or a spec that keeps growing, and money about to be committed.
  • One technical unknown is blocking the plan.
  • An AI-built prototype now has real users and real data.
  • You need a price you can take to a board or an investor.

Not the right start if

  • You haven't spoken to five people who'd use it. Do that first. It costs nothing and it changes the plan.
  • A tool you can buy already does most of it. Buy it. We'll say so on the call.
  • The product has traction and a clear backlog. Start with SaaS Products

Release 1 is one complete journey

Everything above the line ships first. Move stories across the line and see whether someone can still finish the job.

Find a cleaner

Book a slot

Pay

Manage

Release 1: 4 stories. One complete journey: a customer can find, book, pay and manage a booking.

Every case

Default cut
Release 1: 4 stories. One complete journey: a customer can find, book, pay and manage a booking.
Empty Pay column
Release 1 can't be used: nobody can pay. Every step of the journey needs at least one story.
All 13 stories
Release 1: 13 stories. Still complete, but every extra story delays what you learn from real users.

A sample product. Your map comes out of step 5.

How it runs

  1. Discovery sprint · Steps 1–7

    01 of 08

    Kickoff.

    Your goals, constraints and what success looks like, written down.

    You keep A goals and constraints brief.

  2. 02 of 08

    Talk to users.

    Around five conversations with people who'd use it.

    You keep Interview notes and a problem statement.

  3. 03 of 08

    Map the assumptions.

    What must be true for this to work, sorted by risk and by how little we know.

    You keep A ranked list of the riskiest assumptions.

  4. 04 of 08

    Test the riskiest one.

    A clickable prototype for a question about users, a throwaway technical spike for a question about the build.

    You keep An evidence note, and the spike code.

  5. 05 of 08

    Cut release 1.

    A story map with the line drawn: one complete journey, and a written not-now list.

    You keep The release 1 scope and the not-now list.

  6. 06 of 08

    Sketch the build.

    The stack, the architecture and the top risks, with what we'll do about each.

    You keep An architecture sketch and a risk list.

  7. 07 of 08

    Price it.

    A written estimate for release 1, as fixed milestones or a monthly team.

    You keep The written estimate.

  8. MVP · Step 8

    08 of 08

    Build it weekly.

    Plan, build, release, demo and measure, every week, on a staging site you can click.

    You keep A weekly changelog and demo, with the code in your repository from day one.

Most MVPs fail before the first line of code

  1. Building something nobody needs.

    Startup post-mortems keep ranking "no market need" near the top.

    Prevented in step 2
  2. An MVP that teaches nothing.

    It ships, but it was never designed to answer the question that mattered.

    Prevented in step 3
  3. An AI-built prototype goes live as is.

    Once real data arrives, missing security rules leak it.

    Prevented in step 6

170 of 1,645

A 2025 flaw in apps built with one AI app builder (CVE-2025-48757) exposed user data in 170 of the 1,645 apps a researcher checked.

Matt Palmer, the researcher who reported it, 2025Checked Nov 2026

On AI prototyping tools:

good from afar, but far from good.

Nielsen Norman Group, 2025Checked Nov 2026

What you keep

  • Scope document

    What the first release does, what it leaves out, and why.

  • User flows

    The paths real users take, drawn before anyone writes code.

  • Clickable prototype

    A screen-by-screen model you can put in front of users.

  • Technical spike

    A small build of the riskiest part, to prove it works.

  • Written estimate

    The build priced as fixed milestones or as a monthly team.

  • Working MVP

    Released weekly to a staging site, so you see progress in the product itself.

Done means

  • Discovery: release 1 scope, the riskiest assumptions with their evidence, the top technical risks and a written price, all agreed.
  • MVP: one complete journey live with real users.
  • Code in your repository from day one.
  • Sign-in, roles, analytics, error alerts and backups in place.
  • A written decision on what comes next, based on what users did.

Sample

Release 1 scope

Release 1Search by postcode · Pick a time · Pay by card · Cancel a booking
Not nowReviews, favorites, repeat bookings, discount codes, split bills, ratings
Riskiest assumptionCleaners will accept a booking within an hour
EvidencePrototype test with 6 cleaners: 4 accepted within 20 minutes
DecisionBuild release 1, and measure acceptance time from week 1
Illustrative values, not a client result.

Questions about Discovery & MVP

Still deciding?

Ask the people who would build it. We reply within one business day.

Because it is cheaper to find a wrong assumption on paper than in production. The sprint is fixed-price, so you know its cost up front, and its output is yours even if you build elsewhere.

You see the number before you commit to anything beyond the sprint. The estimate comes as fixed milestones or a monthly team, and what drives it is scope and the number of integrations.

Often yes, for the first clickable version. Once real user data, payments or permissions arrive, someone has to own security: in 2025 a flaw in apps made with one such tool exposed user data in 170 of the 1,645 apps checked. We can audit and harden your prototype, or rebuild the parts that matter.

They usually do, and that's fine. We estimate each change in writing before building it, so you choose whether to swap scope, add time or hold the release.

You do, in your own repository, from the first commit. That includes the throwaway spike from discovery.

Yes. We turn business goals into a technical plan, explain trade-offs in plain language and handle the engineering end to end.

Have an idea or a stuck project? Let's talk.

We reply within one business day. If we're not the right team, we'll say so on the first call.

On the first call

  • The idea, who it's for and what you've tried.
  • The decision the MVP has to help you make.
  • The deadline or budget you're working to.