Market research is the work of systematically gathering evidence about your market—how big it is, who's in it, what they want, and who you're up against—so you build on facts instead of hope. For a founder, learning how to do market research isn't about producing a 50-page report; it's about answering a few high-stakes questions cheaply: is this market big enough, reachable, growing, and underserved? This guide covers the types of research, how to size a market with TAM/SAM/SOM, how to read demand signals, and a lean process you can run in days, not months.
What market research is and what it's for
Market research is structured information-gathering about the people and forces in a market. For a large company it might mean expensive panels and reports; for an early founder, it's a focused effort to de-risk decisions before pouring in time and money.
It answers questions guessing can't: How many people have this problem? Is that number growing or shrinking? Who already serves them, and how well? What would it cost to reach them? Crucially, market research is the quantitative and landscape counterpart to the deep, one-on-one learning you get from running customer interviews. Interviews tell you why a handful of people behave as they do; market research tells you how many of them exist and whether the opportunity is worth pursuing. Together they form the backbone of validating a startup idea, and a large, reachable market is precisely the "good market" half of reaching product-market fit.
The goal isn't certainty—it's enough evidence to make a confident go/no-go call.
Secondary vs primary research
Research data comes in two forms, and the smart order is to start with the cheap one.
Secondary research: start here
Secondary research uses data that already exists, collected by someone else. It's fast, cheap, and the right first move. Sources worth mining include industry reports (Statista, IBISWorld, analyst firms like Gartner), government and census data, trade associations, public company filings, and—often most useful for founders—your competitors' own websites, pricing pages, and customer reviews on sites like G2 and Capterra. App store reviews and community forums are goldmines of unfiltered complaints about existing solutions.
Secondary research quickly tells you the rough shape of a market: its size, growth rate, major players, and trends. Its limitation is that it's generic and sometimes stale, and it can't answer questions specific to your idea.
Primary research: gather your own
Primary research is data you collect yourself, directly from the market. The deepest form is customer interviews; others include surveys (good for quantifying what interviews surface qualitatively) and live demand tests like running a small ad campaign to a landing page. Primary research is more effort but answers the specific questions secondary data can't—and it's where you confirm that a problem you read about in a report actually matters to real people.
A parallel distinction runs through both: qualitative research (open-ended, the why—interviews, reviews) versus quantitative research (numerical, the how many—surveys, search volume, market size). You want both. Numbers without stories are hollow; stories without numbers don't scale.
Sizing your market: TAM, SAM, and SOM
The central quantitative question is how big is this opportunity? The standard framework breaks it into three nested circles:
- TAM (Total Addressable Market) — total demand if every possible customer bought. The whole pie.
- SAM (Serviceable Addressable Market) — the slice your product and business model could actually serve, narrowed by segment and geography.
- SOM (Serviceable Obtainable Market) — the realistic portion you can capture in the near term given competition and your resources.
There are two ways to calculate these. Top-down starts from a big industry figure and slices it with percentages ("the market is $50B, and we'll get 1%"). It's easy and almost always overoptimistic and hand-wavy. Bottom-up builds from real units—number of target customers times what each would pay—and is far more credible.
A bottom-up example. Say you're building project-management software for freelance graphic designers:
| Layer | Calculation | Result |
|---|---|---|
| TAM | Entire project-management software market | Billions (whole pie) |
| SAM | ~500,000 freelance designers in your target regions × $120/yr | ~$60M/yr |
| SOM | 2% of SAM realistically reached in 3 years | ~$1.2M/yr |
Those numbers are illustrative, but the method is the point: starting from a countable customer base and a real price gives you a defensible figure you can sanity-check, rather than a fantasy slice of a giant number. Investors and your own judgment both trust bottom-up math far more.
Sizing up competitors and demand
A market with no competitors usually isn't an untapped goldmine—it's a sign there's no demand. So mapping competitors and confirming real interest are essential.
For competitor analysis, identify both direct competitors (solving the same problem the same way) and indirect ones (solving it differently, including the spreadsheet or manual process people use today—your most common real competitor). Study their pricing, positioning, features, and especially their reviews, where customers spell out exactly what's missing. Gaps in what competitors do well are your openings.
For demand signals, you want evidence that people are actively seeking a solution:
- Search volume. Keyword tools (Google Keyword Planner, Ahrefs, Semrush) reveal how many people search for terms related to your problem each month. Real search demand is a strong signal.
- Trend direction. Google Trends shows whether interest is rising or fading—you want a market growing, not dying.
- Community activity. Active forums, subreddits, and groups full of people complaining about the problem prove it's felt and unsolved.
- Ad smoke tests. A small ad budget pointed at a simple landing page measures whether people will actually click and sign up—demand you can observe, not just infer.
A lean market research process, and mistakes to avoid
You can run useful research in a week. Here's a practical sequence:
- Define the decision. Write the specific question you need answered—"Is there a big enough, reachable market of freelance designers for a niche PM tool?" Research without a question becomes endless.
- Do secondary research first. Spend a day or two on reports, competitor sites, reviews, and search data to map the landscape cheaply.
- Size the market bottom-up. Estimate TAM, SAM, and SOM from real customer counts and prices.
- Analyze competitors and demand. Map who's serving the market and confirm people are actively looking for a solution.
- Go primary. Validate your findings with real people through interviews and a demand test—this is where research connects to building, informing what to put in your MVP.
- Synthesize into a decision. Pull it together into a clear go, no-go, or pivot. The output is a decision, not a document.
The recurring mistakes to avoid:
- Researching to confirm, not to test. Going in hoping for a yes guarantees you'll find supporting data. Try to disprove your idea.
- Top-down sizing only. A "1% of a huge market" estimate is a fantasy. Build bottom-up.
- Confusing market size with reachable demand. A huge TAM means nothing if you can't actually reach and convert those customers affordably.
- The "no competitors" delusion. Usually it means no market, or you haven't looked hard enough. Find the indirect ones.
- Analysis paralysis. Endless research is procrastination. Time-box it and move to building.
- Skipping primary research. Secondary data alone never tells you whether your specific solution will land. Talk to real people.
Frequently asked questions
What's the difference between primary and secondary market research? Secondary research uses existing data others collected—industry reports, competitor sites, census data—and is fast and cheap, so start there. Primary research is data you gather yourself through interviews, surveys, or demand tests, answering questions specific to your idea that generic sources can't.
How do you calculate market size? Use the TAM, SAM, SOM framework: total possible demand, the portion your business can serve, and the share you can realistically capture. Calculate bottom-up—multiplying your number of target customers by what they'd pay—rather than top-down from a big industry figure, which tends to be wildly optimistic.
How much market research should a startup do? Enough to make a confident go/no-go decision, not more. A focused effort over a week—secondary research, bottom-up sizing, competitor and demand analysis, then a round of primary validation—usually suffices. Endless research is often procrastination dressed up as diligence.
Is it bad if my idea has no competitors? Usually yes. A complete absence of competitors typically signals there's no real demand, or that you've defined the space too narrowly. Look for indirect competitors and the workarounds people use today—those reveal the real market.
What tools can I use for market research? For demand and trends, keyword tools like Google Keyword Planner, Ahrefs, or Semrush, plus Google Trends. For market data, sources like Statista and industry reports. For competitor insight, their own sites and review platforms like G2 and Capterra. For primary research, simple survey tools and a landing page for ad tests.
The takeaway
Knowing how to do market research as a founder means gathering just enough evidence to decide with confidence: size the market bottom-up, map the competition, confirm real demand, and validate it all with real people. Your next step is to write down the single decision your research needs to inform, spend a day on secondary sources and a bottom-up size estimate, then talk to a handful of actual prospects to pressure-test what you found. Research is only valuable when it ends in a decision—so set out to reach one, not to fill a report.