Market research
Market Research for Startups: A Working Method
Not a list of ten tactics. A method — the order that actually gets you a decision.
Search “market research for startups” and you'll find a lot of checklists: run surveys, do competitor analysis, interview customers, check search volume. All reasonable individually. Almost none of these guides tell you what order to do them in, or why the order matters — which is why founders often burn weeks on the wrong first step.
Why order matters more than the list of tactics
Each research method answers a different question, and some of those questions only make sense once you've already answered an earlier one. Running a customer survey before you know whether the category has real search demand means you're spending real effort validating something that might not have an audience large enough to matter. Sizing the market before checking whether anyone's actually complaining about the problem means building a confident number on top of an unconfirmed need.
The method, in order
| Step | What it answers | Time |
|---|---|---|
| 1. Check if the problem is being searched for | Is this a thing people actively look for solutions to, and is volume stable or growing? | ~1 afternoon |
| 2. Read where people already complain | Specific, repeated frustrations with current options, in their own words | 2–3 hours |
| 3. Map who already competes for this | Open field or locked field — see our competitor analysis guide | 1–2 hours |
| 4. Talk to 10–15 real prospective customers | Test the hypothesis from steps 1–3 against real behavior | 1–2 weeks |
| 5. Size the opportunity | Only meaningful once there's real signal behind it — see our market sizing guide | After 1–4 |
Why this order specifically
Steps 1–3 use data that already exists — search volume, public discussion, public search results. They're fast, mostly free or cheap, and they either confirm the idea has a pulse or tell you to stop before you've spent real time on it. Step 4 is the expensive one — real people's time, including yours — so it belongs after the cheap filters, not before them. Step 5 only matters once there's something real to size.
The mistake most guides encourage
What this method doesn't replace
No amount of pre-launch research replaces actually shipping and watching real usage — research reduces the odds you're building for a market that doesn't exist, it doesn't guarantee your specific execution lands. Treat it as narrowing the odds before you spend the months, not as a substitute for spending them. For the broader question of what's worth doing beyond this five-step core, see our market research guide.
Signal & Demand · $129 per idea
Steps 1–3 of this method — search demand, community evidence, and competitive landscape — are exactly what a Signal & Demand report runs for a specific idea in one pass.