To conduct a market analysis, study a market’s size, segments, competition, and demand to answer one specific business decision, then test that decision against real buyers before you commit money to it. Most analyses skip the step that decides the outcome, which is checking whether buyers will actually pay your price. Desk research shows the market is big. Only primary research shows the market will buy.

Key takeaways

• A market analysis answers six questions in order: what decision are we making, how big is the market, who is the target segment, who do we compete with, will buyers actually pay, and what do we do next.
• Secondary data sizes the market and maps the competition, but it cannot tell you whether buyers will pay your price. Primary research with real respondents closes that gap, and most generic market-analysis guides skip it.
• Size the market with TAM, SAM, and SOM so the analysis separates the whole opportunity from the slice your business can realistically win.
• Building something with no real market need remains one of the top reasons companies fail (CB Insights, analysis of startup post-mortems). A market analysis that tests demand directly is the cheapest insurance against that outcome.

For founders, product teams, and insights or marketing leads deciding whether a market is worth entering, expanding in, or repricing.

What is a market analysis, and why run one?

A market analysis is a structured assessment of a market’s size, segments, competitors, and demand, run to inform one specific decision. It is not a research project for its own sake. It is the evidence you gather before you commit capital to a launch, an expansion, or a price.

You run one to replace assumptions with evidence on the few questions that decide the outcome: how big is the opportunity, who exactly should you sell to, who will they compare you against, and will they pay your price. Skip the analysis and you are betting the budget on a guess. Run it well and you de-risk the decision before it gets expensive.

What you need before you start

You do not need much to begin, but you do need two practical capabilities in place. The decision itself is Step 1, not a prerequisite.

• A source of industry and competitor data. Analyst reports, company filings, and trade data give you market size and competitor moves.
• A way to collect primary data from real buyers. A survey platform with access to a quality-controlled respondent panel is what lets you test demand, rather than stopping at desk research. Most teams already have the first and skip the second, which is exactly why so many analyses describe the market well and predict it badly. With those two capabilities ready, the work starts with the decision.

Step 1. Lock the decision before you open a spreadsheet

Before you touch a spreadsheet, lock down the exact move you are trying to make. If your objective is “to understand the market”, stop. You are about to build a 50-slide graveyard of useless data.

Anchor the project to one ruthless, one-sentence question instead. Something like, “Should we launch our premium tier at $49 or $69 in Australia?” That single sentence dictates your geography, filters your audience, and tells you exactly when to stop digging. If a data point does not help answer that question, ignore it.

The decision sets the geography, the time horizon, and the segments worth studying, so a sharp question narrows the work and a vague one expands it without end. With the decision fixed, the first thing it demands is a sense of scale.

Step 2. Size the market and sanity-check the number

Size the market in three layers so the analysis separates the total opportunity from the slice you can realistically capture. TAM, SAM, and SOM is the standard framework, and each layer answers a different question for the decision.

• TAM (total addressable market). Total demand if you had zero competitors. Good for a pitch deck, useless for daily operations.
• SAM (serviceable available market). The geographic and commercial ceiling of who you can actually reach.
• SOM (serviceable obtainable market). The realistic share you can win from competitors over the next 24 months. This is the number that goes in the forecast. Once you have a figure, calculate it two ways. Run it top-down from macro industry reports, then bottom-up from your own pricing multiplied by reachable local buyers. If the two numbers do not match, you have buried a bad assumption somewhere. Find it before it reaches a board deck. With a credible number in hand, the next question is who inside that market you actually serve.

Step 3. Segment on behaviour, not demographics

Treating a market as one giant audience is the fastest way to build something acceptable to everyone and compelling to no one. You need a target segment.

Stop segmenting by lazy, easy-to-measure metrics like age or postcode, though. Demographics rarely predict buying behaviour. Slice your market by needs, usage occasions, and willingness to pay instead. A usage and attitude study quantifies how segments differ on behaviour and need, and a MaxDiff analysis ranks what each segment actually prioritises. Pick one high-value segment to win first, then leave the rest for later. A clear target then tells you which competitors you are really up against.

Step 4. Map the real competitors, including doing nothing

List the alternatives your target segment actually turns to, then map your direct and indirect rivals on a price-and-feature grid to find the underserved gaps. That is the standard move, and it is necessary, but it is not enough.

The biggest competitor most analyses ignore is the status quo. Most buyers do not choose your rival. They choose to do nothing, because switching is too much effort. If your competitive map does not account for that inertia, your launch will stall against a wall you never plotted. Desk research takes the map only this far. To learn whether buyers will switch, and at what price, you have to ask them.

Step 5. Test demand before you guess at a price

This is where most market analyses go to die. Teams size a huge market, map rivals perfectly, then guess at the price tag. Secondary data only tells you what buyers did yesterday. It cannot tell you what they will pay tomorrow.

Before you commit capital, run primary research on real respondents to test your exact proposition.

• Testing price on its own. Use the Van Westendorp Price Sensitivity Meter to find your acceptable price floor and ceiling, or Gabor-Granger to pinpoint where demand drops off a cliff.
• Testing features and price together. Use conjoint analysis. It forces buyers to make real trade-offs and shows you exactly which features justify a premium price.
• Getting a reliable signal. A managed respondent panel gives you buyers who match your target segment, and a sample size calculator sets how many you need for a read you can trust. A quick primary study costs a fraction of a failed, mispriced launch. For the full method set, the pricing research overview maps each technique to the decision it suits. Once demand is measured, the analysis is ready to become a verdict.

Step 6. Turn the data into a one-page verdict

Nobody wants to read a 60-slide research deck. A market analysis is judged on one thing, whether it drives an action.

Close with a one-page executive summary that leads with the verdict. State what the data recommends, how confident you are, the revenue forecast (built on your Step 5 willingness-to-pay evidence, not on desk assumptions), and the exact metrics that would change your mind. Put the spreadsheets, SWOT grids, and charts in the appendix where they belong. A forecast grounded in tested willingness to pay is defensible in a way a forecast built on guesswork never is, which is the whole reason Step 5 earns its place.

Secondary versus primary data in a market analysis

Use secondary data to frame the market and primary data to decide, because each answers a different kind of question. Secondary data, drawn from existing reports and records, is fast and cheap but describes the past and is shared with every competitor reading the same reports. Primary data, collected from your own respondents, is the only source that speaks to your specific decision.

The two are sequential, not either-or. Secondary data narrows the question cheaply, then primary research answers the part the decision actually rests on. That sequence is what keeps a market analysis both fast and decision-grade.

Common challenges and solutions

Four problems account for most market analyses that never change a decision. Each has a practical fix.

Challenge: the analysis has no decision behind it.
An open brief produces endless data gathering and a report that sits unused.
Solution: Write the one decision and the date it must be made before collecting anything. Cut any data that would not change the decision.

Challenge: market size rests on one unchecked number.
A single top-down estimate inherits whatever assumption sat underneath it.
Solution: Size top-down and bottom-up, then investigate the gap. The discrepancy is usually where the flawed assumption hides.

Challenge: the competitive map ignores the real alternative.
Analyses compare named rivals and forget that most buyers default to doing nothing.
Solution: Include the status quo and indirect substitutes as competitors, and test demand against them, not only against direct rivals.

Challenge: the price is assumed, not tested.
The plan names a price no real buyer has ever evaluated.
Solution: Run a short willingness-to-pay study with (Van Westendorp)[https://conjointly.com/products/van-westendorp/], (Gabor-Granger)[https://conjointly.com/products/gabor-granger/], or (conjoint)[https://conjointly.com/solutions/conjoint-analysis/] before the price reaches the forecast.

Frequently asked questions

What is the difference between a market analysis and market research?

A market analysis is the broader exercise of assessing a market’s size, segments, competition, and demand to inform a decision. Market research is the activity of collecting the data, both secondary and primary, that the analysis draws on.

How long does a market analysis take?

A focused market analysis takes one to three weeks. Allow a few days to size and segment the market from secondary data, and one to two weeks to run and read a primary study on a research platform, which is far faster than the months a traditional project once required.

Do I need primary research, or is secondary data enough?

Secondary data is enough to size a market and map competitors. Any decision that turns on price, feature trade-offs, or whether buyers will switch needs primary research, because no existing report measures how your specific segment will respond to your specific offer.

Which method measures willingness to pay in a market analysis?

Van Westendorp maps an acceptable price range, Gabor-Granger estimates demand at set price points, and conjoint analysis measures how price trades off against features, with conjoint the right choice when the decision involves more than price alone.

Conclusion and next steps

A market analysis is only as good as the decision it changes. Secondary data tells you the shape and size of the market, but the part that de-risks the call, whether real buyers will choose your product at your price, comes from asking them. Work the six steps in order.

1. Lock the single decision your analysis must inform, and the date it is due, in one sentence.
2. Size the market with TAM, SAM, and SOM, calculated top-down and bottom-up, and investigate any gap.
3. Segment on behaviour and willingness to pay, then pick one target segment to win first.
4. Map the competitors your target actually considers, including the status quo of doing nothing.
5. Test demand and price on real respondents with Van Westendorp, Gabor-Granger, or conjoint before any number is locked.
6. Close with a one-page verdict that leads with the recommendation, the confidence level, and the forecast your demand evidence supports.