Guide
Market validation research: methods, questions and sources
Market validation research is the evidence-gathering half of validation: talking to the people who have the problem, and checking public data for the size and shape of the market around them. Done well it takes days. Done badly it produces a deck full of numbers that all agree with you.
The rule throughout: measure behaviour that already happened, not intentions about the future.
Primary research: the interview
Ten to fifteen conversations with people in one segment beats a hundred survey responses across five. Keep them to twenty minutes, do not describe your product until the last two minutes, and record the exact words people use — that vocabulary becomes your marketing copy.
- Tell me about the last time this happened. What did you do?
- How long did that take, and who else was involved?
- What did it cost you — in money, time or missed work?
- What have you already tried to fix it? Why did you stop?
- If it stayed exactly as it is for another year, what changes for you?
- Who decides on spending for this, and what does approval look like?
Surveys: useful only after interviews
Use surveys to measure how common a pattern you already discovered is — never to discover it. Ask about frequency, current spend and current tools, all in the past tense. Avoid any question containing 'would you'. Report raw counts, not percentages, until your sample is large enough for percentages to mean anything.
Desk research sources worth using
- National statistics offices and central bank data for population, income and sector size in your actual market.
- Industry associations and regulator registers for counts of businesses or licensed practitioners.
- Job postings and procurement notices — they reveal what organisations are actively paying to solve.
- Competitor pricing pages, changelogs and review sites for what customers complain about in public.
- Search demand data for whether people are actively looking for a solution or unaware one exists.
Prefer bottom-up sizing built from these counts over any headline 'global market worth X' figure. A number you assembled yourself can be checked; a report figure cannot.
Reading the results without fooling yourself
Write your pass and fail thresholds before you collect data. Count how many people described the problem unprompted, how many had already spent money trying to solve it, and how many committed something after hearing the price. Those three counts, not overall sentiment, decide the verdict. If the result is ambiguous, the segment is usually too broad — narrow it and repeat.
Run this on your own idea
Samchu-AI does this work for you. Describe your problem in a paragraph and a nine-expert council returns scored solutions, bottom-up costs, S-curve adoption modelling, a customer profile, a competitor radar and a prioritised action plan — in your own region, currency and language. Your first project is free.
Validate my idea free