Guide

The market validation process, step by step

Most ideas are not killed by competitors — they are killed by an assumption nobody checked. This is the process we use inside Samchu-AI to pressure-test an idea before any money is committed: define the problem, size demand, cost it bottom-up, model adoption on an S-curve, and end with a decision you can defend.

1. 01

   Write the problem, not the product

   State who has the problem, how often it happens, what it costs them today, and what they currently do instead. If you can only describe your solution, you have an idea, not a validated problem. A good problem statement names the user, the frequency, the current workaround and the constraint (budget, device, regulation, geography).

   Dr. Amara Okonkwo — Lead Scientist tests whether the premise holds up to evidence.

2. 02

   Check whether the problem is worth solving

   Estimate the addressable population in your actual market, not globally, then multiply by what they realistically pay per year. If the honest number is small, the fix is usually a narrower, more painful segment rather than a bigger claim.

   Prof. Yusuf Al-Rashid — Economist sizes demand and willingness to pay in your currency.

3. 03

   Cost it bottom-up, never top-down

   Validation fails most often on economics, not desire. Build the cost from parts you can name: engineering weeks, infrastructure per active user, support hours, payment fees, compliance. Bottom-up costing exposes the unit economics that a percentage-of-revenue guess hides.

   Rafael Duarte — Systems Engineer scopes the build; the Economist prices each line item.

4. 04

   Model adoption as an S-curve, not a straight line

   Real adoption is slow, then steep, then flat. Assume a small early cohort, a growth phase driven by referral and distribution, and a ceiling set by your addressable market. A straight-line forecast almost always overstates year one and understates the ceiling, which is why cash runs out earlier than planned.

   Mei-Ling Zhao — Business Strategist maps the distribution route into the curve.

5. 05

   Pressure-test the risks that kill launches

   Legal exposure, data protection duties, security surface, and environmental or operational load are not later problems. Each one can invalidate the plan after the money is spent, so they belong in the validation stage.

   Ananya Raghunathan (Legal), Tomas Novak (Cybersecurity) and Dr. Nomvula Dlamini (Environmental) review exposure before you build.

6. 06

   Design the smallest honest test

   Define the one experiment that would change your mind: a landing page with a price, ten structured interviews, a manual concierge version, or a paid pilot. Write the pass/fail threshold before you run it so the result cannot be reinterpreted afterwards.

   Ingrid Lindqvist — UX Architect turns the test into a usable first experience.

7. 07

   Decide: proceed, reshape, or stop

   Validation ends in a decision, not a document. Score the opportunity, name the top three actions, and set the next review date. Stopping early is a positive outcome when the evidence says so.

   Ruby Callaghan — Lead Investor gives the funding-grade verdict and the action list.