The Mistake Most People Make First
They build. They spend three months creating a product, a website, a brand, a logo — then they try to find customers. By that point they've sunk time, money, and identity into something the market hasn't confirmed it wants.
Validation flips that order. You find out if people will pay *before* you build anything substantial. It's not glamorous, but it's the only method that actually protects your time.
Here's how to do it properly.
---
Step 1: Write Down the Problem, Not the Solution
Most people start with their idea — the app, the course, the product. Start with the problem instead.
Finish this sentence: *"People who [describe your target person] struggle with [specific problem] because [root cause]."*
Example: *"Freelance designers struggle to price their services consistently because they have no repeatable framework for scoping projects."*
If you can't write that sentence clearly, you don't have a validated problem yet. You have a guess. That's fine — it's the starting point, not the finish line.
Keep the problem statement to one sentence. Vagueness here will contaminate everything downstream.
---
Step 2: Identify 10 Specific People Who Have That Problem
Not a demographic. Not "small business owners." Specific people you could message today.
- Former colleagues
- People in niche Facebook groups or subreddits
- LinkedIn connections in a relevant role
- People who have publicly complained about this problem online
You need ten. Not a hundred. Ten people you can actually reach and have a real conversation with.
This forces you to move from abstract market research to concrete human beings, which is where real validation happens.
---
Step 3: Run Problem Interviews — Not Pitch Sessions
This is the core of fast validation and most people skip it or do it wrong.
A problem interview is a 20-minute conversation. Your only job is to understand how this person currently experiences the problem — not to sell them anything.
- *"Walk me through the last time this was a problem for you."*
- *"What have you already tried to fix it?"*
- *"What did that cost you — in time, money, or stress?"*
- *"How important is solving this to you right now, on a scale of 1 to 10?"*
Don't mention your solution. Don't hint at it. Just listen.
After five to seven interviews you'll start hearing patterns. If the same pain points come up unprompted across different people, that's a real signal. If everyone describes the problem differently or most people rate it a 4 out of 10, that tells you something too.
What you're looking for: People who are already spending time or money trying to solve this — imperfectly. That's the gap your solution fits into.
---
Step 4: Test Willingness to Pay — Not Just Interest
Here's the most important line in this article: people lie about interest, but they don't lie with money.
Asking someone "would you use this?" is nearly worthless. Almost everyone will say yes to be polite. You need to manufacture a moment of decision.
A few ways to do this at zero build cost:
Pre-sell or waitlist with intent. Tell five people: "I'm building a solution for exactly what you described. It'll be £X. Want to be first in?" A genuine yes comes with an email address, a deposit, or a commitment to a follow-up call. Vague enthusiasm is a no.
The concierge test. Offer to do the solution manually — by hand, using your own time — for a small fee. If someone pays you £50 to do manually what your product would eventually automate, that's real validation. It also teaches you exactly how the solution needs to work.
A landing page with a sign-up. Build a single-page site describing the problem and the solution in plain language. Drive 50–100 targeted people to it via a relevant Reddit post, LinkedIn article, or paid ad spend of £20–£30. A sign-up rate above 15–20% on a cold audience suggests real demand. Under 5% suggests a messaging or positioning problem at minimum.
---
Step 5: Score What You Learned
After your interviews and your payment test, score four things honestly:
1. Problem sharpness — Can you describe the problem in one sentence that makes the right person say *"yes, that's exactly it"?* 2. Frequency — How often does this problem occur? Daily problems are more valuable than annual ones. 3. Spend signal — Are people already paying for imperfect solutions? 4. Reachability — Can you actually get in front of these people without needing a massive budget?
You don't need all four to be green. But if three of the four are weak, you haven't found your idea yet — you've found a direction to refine.
---
What Fast Validation Actually Looks Like on a Timeline
This entire process — done seriously — takes two to three weeks, not months.
- Days 1–3: Clarify your problem statement and identify ten specific people
- Days 4–10: Run five to seven problem interviews
- Days 11–14: Test willingness to pay with a concierge offer or a landing page
- Day 15: Score what you learned and decide: iterate, pivot, or build
If you spend longer than three weeks validating, you're probably overthinking it or avoiding the discomfort of real conversations. Both are procrastination with better-sounding names.
---
The Honest Truth About What You Might Find
Most first ideas need to be refined, not abandoned. Validation rarely gives you a clean green light — it gives you better information. You'll often discover the real problem is slightly different from what you assumed, or that a specific sub-group cares far more than the general audience you imagined.
That's not failure. That's the process working exactly as it should.
The goal of validation isn't to confirm that you were right. It's to give yourself the best possible starting position before you commit serious time and money to building.
---
Practical Takeaway
Before you write a single line of code, record a single lesson, or design a single product — go have five real conversations with real people who might have the problem you're solving. Not surveys. Not forums. Actual conversations where you ask and they talk.
If you can't get five people to spend 20 minutes talking about a problem they supposedly have, that's your first data point.
