Prove people will pay before you build it.
Three weeks. Fifteen to twenty problem interviews, a live page that takes real money or real waitlist emails, a clickable prototype of the core flow, and a one-page go/no-go memo with the pre-order count on it. You finish with a decision, not a codebase.
Three sprints a quarter and the list decides who gets them. Joining is free and takes no payment — 50% up front only applies once a slot is yours, and the second half is refunded if zero of 20 interviews confirm the problem.
What you actually get
1. Problem interviews
15–20 conversations with the exact people you think have the problem. I never pitch your idea — I ask what they do today, what it costs them, and what they have already paid to fix it. Every call transcribed, quotes pulled, patterns counted.
2. A page that takes money
A live landing page with your offer and a real payment or pre-order path — not a "coming soon" email box. It runs on infrastructure I own and deploy, so it is up the same week, and the analytics are yours at the end.
3. A clickable prototype
The core flow, clickable, in front of a real user. Enough to answer "does this work the way they expect," not enough to be a product. This is the highest-fidelity thing you show before anyone writes production code.
4. The go/no-go memo
One page: what the interviews said, what the page converted at, how many said yes with money or a signed letter of intent, what to build first if it is a go, and what kills it if it is not.
How the three weeks run
| Week | What happens | What I look at |
|---|---|---|
| 1 Problem |
Scope the wedge. Build the interview list from your network plus cold outreach. Run the first 8–10 interviews. No pitching. | Unprompted pain: wasted time, lost money, workarounds already paid for. |
| 2 Demand |
The page goes live with the offer and a payment or pre-order path. Finish the interviews. Point real traffic at it. | Visitor → signup, and signup → checkout. Both, not one. |
| 3 Wallet |
Warmest leads get the clickable prototype and the ask: pre-order now, or sign a letter of intent naming a date and a price. | Cards and signatures. Interest is free; this is not. |
Who this is for
Good fit
- You can describe the buyer in one sentence.
- You have some way to reach 50+ of them.
- You will act on a no. A killed idea is a win here.
- You want evidence, not a build.
Wrong fit
- You need production software built. This is not an agency, and no code ships.
- You want someone to agree the idea is good.
- You cannot talk to customers for 3 weeks, or will not let me.
- You need a technical co-founder.
Price and the one term that matters
Half up front to start, half on delivery. If zero of the 20 interviews confirm the problem, the second half is refunded — you keep the interviews, the page, the prototype and the memo. The risk that the idea is wrong sits with me, not with you.
Why fixed: the deliverable is a decision, and a decision has a date. Below the $7,500 floor of fixed-scope studios because I am not building production software — I am buying you the information that tells you whether to spend $50k on it.
Join the waitlist
Taking three sprints a quarter, and the list decides who gets them — first refusal, in order. Joining is free and takes no payment. Before a slot is yours I will ask three questions about the idea, and if it does not fit I would rather say no than take money for a sprint I cannot run.