Skip to content
I Build MVPs
The full process, step by step

How a rough idea becomes a working first product.

Six stages, two of them optional. Most founders go from first contact to live product in under three weeks. Here's exactly what each stage looks like.

Check if your idea fits See pricing
  1. 01

    Fit Check

    5 minutes Free

    You fill out a short form describing your idea, who it's for, and what the first useful version would do. You get a written answer within two working days: looks like a fit, doesn't fit, or worth a longer conversation. No sales call, no pitch.

    I say no to most ideas at this stage. That's the point.

  2. 02

    Blueprint

    24–48 hours £950

    If the idea looks like a fit, I run the Blueprint. I translate the messy idea into a one-user, one-workflow product brief: target user, core problem, the smallest useful version, what's in, what's cut, the data model, the technical approach, and a fixed-scope sprint proposal with a price and a date.

    The Blueprint ends with one of three answers: proceed, validate first, or not a fit.

  3. 03

    Scope lock

    Same day Included

    If you're proceeding, the scope is locked in writing before any code is written. What's in the sprint is in. What's out is out. This is the moment that protects the price, the date, and your sanity. It's also where most founders try to add things, and where I say no.

    Scope cannot be expanded mid-sprint. It can only be cut.

  4. 04

    Build

    5–10 working days £5,950–£14,950 (set in Blueprint)

    I build the core workflow on a proven application foundation: a small library of reusable architectures and patterns assembled over years of building first products. Authentication, basic admin, billing, AWS deployment, transactional emails — none of those start from a blank page. The novel work is your workflow, not the scaffolding around it. You see progress as it goes. You don't see surprises at the end.

    One core workflow, built properly. Not a polished SaaS, not a throwaway prototype: the first useful version.

  5. 05

    Handover

    Final day Included

    You get the source code in a repository you own, handover notes, known limitations, and a next-step backlog so you can keep building. With me, with someone else, or yourself. No lock-in, no proprietary glue. Hosting arrangements (you operate it, I operate it for you, or we transition over time) are a separate conversation, not a hostage situation.

    You own everything. I just leave you with something that works.

  6. 06

    Iterate (optional)

    5 days £3,950

    After the first product is in front of real users, you'll know more than you knew before. The Iteration Sprint is for the changes that follow: tweaks based on user feedback, the next obvious workflow, a small new section. Same model: fixed scope, fixed price, in days.

    Most founders take at least one Iteration Sprint within 60 days of handover.

The build / do-not-build call

The first decision isn't how to build it. It's whether to build it now.

Most money wasted on first products is spent on something that didn't need to be software yet. The Blueprint runs a sequence of questions before recommending a build. Any “no” sends me to a different recommendation, not the same recommendation with a discount.

  • Can the user, problem, and outcome be described in one sentence each?
  • Is there real evidence the problem hurts, not just sounds plausible?
  • Can the first version be reduced to one workflow, one user, one outcome?
  • Does this genuinely need software now, or would a manual test be cheaper and faster?
  • Can you, the founder, make decisions in hours rather than weeks?

The questions founders ask between stage two and stage three

The moment the price gets serious, these are what come up.

What does “Validate first” look like in practice?

It varies by idea. For some founders it's ten conversations with the people they think will pay. For others it's a landing page with a real waitlist, a manual concierge service done by hand, or a one-page test that runs without a database. The point is to spend hundreds, not thousands, before committing to a build. If the Blueprint says validate first, you'll get the specific test I recommend, not a generic checklist.

What if I disagree with the Blueprint's answer?

That's fine, and not unusual. The Blueprint is honest, not gospel. If you disagree, push back. The reasoning is in the document, and reasoning can be wrong. What I won't do is reverse the answer to win the work. If your idea isn't a fit, I don't take the build money and hope for the best.

How do you decide what makes the cut?

I ask one question: does this feature have to exist on day one for an early user to do the workflow end-to-end? If yes, it's in. If no, it's on the next-step backlog. That's it. “It would be nice”, “investors will expect it”, and “everyone has it” are not yes answers.

Can I bring my own design?

Yes, and I encourage it if you have something solid. A clean Figma file, even a partial one, can shave time off the sprint. If you don't, I use a proven, restrained design system rather than spending sprint days designing from scratch. The first product should look credible, not custom.

What if the build hits a problem I didn't foresee?

Things will surface during the build that the Blueprint didn't predict; they always do. The rule is simple: the price doesn't change, the scope flexes. If something turns out to be more complex than expected, something else gets cut to keep the sprint on rails. You'll be told the moment a trade-off comes up. You make the call, in minutes, not weeks.

Ready to see if your idea fits?

The Fit Check is free, takes five minutes, and ends with a written answer. I don't book sales calls before there's something to talk about.

Start with the Fit Check