Key takeaways
Write your riskiest assumption as a sentence that can be disproved
Ask about past behaviour, never "would you use this?" — everyone says yes
A waitlist only counts if people reply and ask when it launches
The concierge test — delivering manually for five customers — is the most under-used and most informative test
Pre-sales are the strongest signal; £500 paid beats a hundred compliments
A clickable prototype costs £2,000–£6,000 and tests the solution, not the problem
Full validation takes 4–8 weeks; a misjudged MVP costs £15,000+ and three months
I turn down MVP work fairly often. Usually because the founder in front of me has not yet found evidence that anyone wants the thing, and spending £30,000 to find out is the most expensive possible way to learn it.
Validation is not a stage you skip because you are confident. It is the cheapest research you will ever do. Here is how to do it properly in a few weeks, and what a genuine green light looks like.
Write the assumption down first
Every idea rests on one assumption that, if wrong, makes everything else pointless. Write it as a sentence someone could disprove:
- Weak: "Small landlords need better software."
- Strong: "Landlords with 2–10 properties will pay £19/month for a tool that chases rent arrears automatically."
The strong version names who, what they pay, and what they get. It can be tested. The weak version can only be agreed with politely, which is why founders like it.
The five cheapest tests, in order
1. Problem interviews (£0, one to two weeks)
Talk to ten to fifteen people in your target group. Not about your idea — about their current behaviour. Good questions:
- Walk me through the last time you dealt with this. What did you actually do?
- What are you using now? What does it cost you?
- What have you already tried and abandoned?
- How often does this happen — weekly, monthly, twice a year?
Bad question: "Would you use an app that does X?" Everyone says yes. It costs them nothing to be nice to you.
Listen for two things: an existing workaround (a spreadsheet, a WhatsApp group, a part-time admin), and a number (hours lost, money leaked, deals missed). Both mean the problem is real enough to have cost someone something already.
2. Landing page and waitlist (£500–£2,000, one to two weeks)
One page stating the promise plainly, an email capture, and £150–£400 of paid traffic or three well-chosen community posts. What you are measuring:
| Signal | Weak | Promising |
|---|---|---|
| Visitor to email sign-up | under 5% | over 10% |
| Replies to your follow-up email | almost none | 20%+ engage |
| "When can I use this?" messages | zero | several unprompted |
A waitlist with 400 emails and no replies is not validation. Forty emails where twelve people ask when it launches is.
3. The concierge test (£0–£1,000, two to four weeks)
Deliver the outcome manually for five customers before you automate anything. Do the matching yourself. Send the report by hand from a spreadsheet. Run the bookings through your own calendar.
This is the single most under-used test in software, and the most informative. You learn the real workflow, the exceptions, and whether people value the outcome enough to keep coming back. Several products I have built started as a founder doing the work by hand in Google Sheets — and the MVP scope was obvious because they already knew which parts hurt.
4. Pre-sales or a paid pilot (the strongest signal there is)
Ask for money before the product exists. A deposit, a discounted annual plan, a paid pilot with three businesses, or a signed letter of intent with a price on it.
Nothing else comes close. Ten people saying "great idea" is noise; two people paying £500 for something that does not exist yet is a business. If nobody will pay in advance, ask why — the honest answer is usually that the pain is smaller than you thought, or you are talking to the wrong buyer.
5. Clickable prototype (£2,000–£6,000, two to three weeks)
Once you have real interest, a Figma prototype lets you test the solution rather than the problem: does the flow make sense, does the pricing land, do people understand it without you explaining? It is also what you put in front of investors and pilot customers.
A prototype is not an MVP. It looks real and does nothing. That is the point — it costs a tenth as much to change.
Signals that justify building
Build when you have most of these, not one of them:
- Ten or more interviews where people describe the same problem unprompted
- An existing workaround you can point at
- A number attached to the pain — hours, pounds, lost deals
- Someone who has paid, pre-paid, or committed in writing
- A named list of first users you can email on launch day
- A single core workflow you can describe in one sentence
Signals that you are not ready: friends and family enthusiasm, a large but silent waitlist, competitor funding rounds used as proof of demand, and "we just need to build it so people can see it".
Validation does not mean going slowly
This whole sequence takes four to eight weeks and costs a few thousand pounds at most. A misjudged MVP costs £15,000 to £60,000 and three months, and the worst outcome is not failure — it is a working product nobody needs, which is much harder to walk away from.
For what happens after a green light, read the MVP development process and my UK MVP cost guide.
Where I fit
I build MVPs for UK founders, but the first conversation is always about whether you should build yet. Sometimes the answer is a prototype or a waitlist page instead, and I would rather tell you that than take a project that cannot succeed.
Book a free 30-minute call and bring whatever evidence you have. See also MVP development UK and SaaS MVP development.
