Market Research for a Business Idea: What to Study Before You Build

A founder can spend weeks asking, "Is this a good idea?" Market research gives a better question. What would need to be true for this idea to work, and what evidence can we find today?

That shift matters. An idea is not validated because an industry is growing. It is not validated because people discuss the problem online. It is not validated because competitors raised funding. These are signals. Market research collects and organises them. Direct validation then tests the assumptions that matter most.

TL;DR — Read this first

What it is

Market research for a business idea is the process of studying the people, organisations, alternatives and conditions around a proposed product or service to reduce uncertainty before major decisions are made.

Why it matters

Without structured research, founders may rely on signals that sound positive but do not explain whether an idea can work for a specific customer. Research replaces the question "Is this a good idea?" with a sharper one: "What would need to be true for this to work?"

Use it when

Use it when an idea exists but needs evidence about customers, alternatives and market conditions before major investment in building. It is most useful after the Idea Clarifier and before customer interviews or validation experiments.

What you receive

A structured report covering problem context, customer groups, current behaviour, competitors and substitutes, market size, trends, constraints and evidence gaps, with recommended next phases.

Important limit

Market research shows whether a problem and market appear credible. It cannot validate the idea or replace customer conversations, pricing tests or a pilot programme. Real customer behaviour is always required to test demand and willingness to pay.

What is market research for a business idea?

Market research for a business idea is the process of studying the people, organisations, alternatives and conditions around a proposed product or service. Its purpose is to reduce uncertainty before major decisions are made.

For an early idea, the research should help answer six questions:

• Who may have the problem?
• How do they handle it today?
• How often and how seriously does it occur?
• What alternatives compete for attention and budget?
• What market or regulatory conditions affect the idea?
• Which assumptions still require real-world testing?

The purpose is not to write a positive story about the market. The purpose is to help the user decide what to do next.

Primary research and secondary research

Secondary research

Secondary research uses information that already exists. This can include government statistics, company reports, trade associations, academic studies, public pricing pages, reviews, job postings, search data and industry publications.

It is useful for understanding market structure, terminology, trends and known constraints. It is also faster than starting with interviews.

But public information has limits. Reports may define the market differently. Data may be old. A large global figure may say little about the first group you can realistically serve.

Primary research

Primary research collects new evidence from the people or organisations being studied. Interviews, observation, surveys, pilot programmes and behavioural experiments fall into this group.

Primary research can show how people describe the problem, what they do today and whether they are willing to change. It is especially useful after secondary research has made the questions more specific.

Good market research uses both. Secondary research builds the map. Primary research checks whether the map matches real behaviour.

What should a market research report cover?

1. Problem context

The report should explain when the problem occurs and why it matters. A problem that happens once a year is different from one that interrupts work every day. A painful problem with no budget may still be difficult to turn into a business.

2. Customer groups

The report should identify possible users, buyers and decision-makers. These may be different people.

In a school software product, teachers may use the tool. A school administrator may approve it. A procurement team may control the purchase. A parent may influence whether it is accepted.

3. Current behaviour

The strongest competitor is often the current process. People may combine email, spreadsheets and manual follow-up because the process is familiar and already paid for.

Research should examine the cost of switching as well as the weakness of the current solution.

4. Competitors and substitutes

Direct competitors solve a similar problem for a similar customer. Indirect competitors solve the problem in another way. Substitutes may include internal staff, agencies, generic software or doing nothing.

A detailed comparison belongs in the Competitor Teardown phase. Market Research should still show the main categories and what they suggest about demand.

5. Market size

TAM, SAM and SOM are often presented as large numbers. They are more useful when the definitions and assumptions are visible.

TAM is the broad total market under a defined model. SAM is the part that fits the product and geography. SOM is the part the business may reach within a realistic period and route to market.

A bottom-up estimate is often more useful for an early founder. Count the reachable customer units, estimate a sensible price and state the adoption assumptions. The number may be smaller, but the reasoning is easier to challenge.

6. Trends and timing

Research should explain why the opportunity may exist now. A change in regulation, technology, cost, behaviour or distribution can create a new opening.

A trend is not automatically a reason to start. The report should connect the trend to the customer problem.

7. Constraints and risks

Some ideas depend on licences, access to data, long procurement cycles, physical supply, trust or professional approval. These conditions belong in market research because they affect who can buy and how the product can be delivered.

8. Evidence gaps

The report should end with questions, not certainty. Which facts are well supported? Which numbers are estimates? Which assumptions need customer interviews, pricing tests or a pilot?

The Market Evidence Ladder

Different signals carry different weight. A useful report should not treat every mention, survey response or signup as equal.

The lower levels are useful for forming hypotheses. The higher levels are stronger evidence. Even payment does not prove that a market will scale, but it changes the quality of the discussion.

Example: research for a refillable cleaning product

Consider an idea for refillable household-cleaning products for people living in apartments.

A weak report may say the sustainable-products market is growing and consumers care about plastic waste. This sounds positive, but it does not explain whether the idea can work.

A stronger report would examine where apartment residents currently buy cleaning products, how much storage they have, whether they prefer delivery or local collection, what refill systems already exist, price sensitivity, packaging rules and the economics of repeated delivery.

It may find that environmental concern is not the main buying trigger. Convenience, scent, child safety or avoiding heavy bottles may matter more. That changes positioning and product design.

The next step may be ten interviews with apartment residents and a small preorder test in one neighbourhood. The report has not validated the idea. It has made the validation more precise.

Market research for students and case studies

A student may not be preparing to launch the business. The same structure is still useful.

The report can show how the market is defined, which data sources support the analysis, where estimates were made and which claims would need primary research. This creates a stronger case study than a collection of industry statistics.

The student should still cite the original sources required by the institution. An IdeaClarify report can organise the work, but it should not replace academic source review.

Common market research mistakes

Using one market-size number

Different reports often measure different products, geographies and years. The definition matters more than the headline number.

Researching the solution instead of the problem

Questions such as "Would you use an AI app for this?" push attention towards the proposed answer. Start with current behaviour and cost.

Treating competitors as proof

Competitors may indicate demand. They may also serve a different customer, depend on funding or struggle with retention. Their existence is evidence to study, not a verdict.

Ignoring the cost of change

A new product competes with habit, existing contracts, training and internal politics. A better feature may not be enough.

Hiding weak evidence

An estimate should be labelled as an estimate. A simulation should be labelled as a simulation. A claim from one source should not be written as a settled fact.

How Market Research connects with other IdeaClarify phases

Market research in a chat window vs IdeaClarify

An LLM can summarise an industry, suggest competitors and explain market-sizing methods. It can be very useful during research.

The problem is consistency. A general answer may mix facts, estimates and assumptions. It may not preserve the same definitions across later work. It may also give a confident answer when public evidence is limited.

IdeaClarify should use a fixed research structure. Sources should be visible. Claims should carry dates where needed. Estimates should show assumptions. The report should finish with evidence gaps and recommended next phases.

That process creates a better bridge between research and action.

Frequently asked questions

Can market research prove that my idea will work?

No. It can show whether the problem and market appear credible. Real customer behaviour is needed to test demand, willingness to change and willingness to pay.

How much market research is enough?

Enough to identify the customer, alternatives, market conditions and highest-risk assumptions. Research becomes unhelpful when it delays direct validation without changing the questions.

Can AI conduct market research?

AI can help find patterns, organise sources, compare information and form questions. Source quality, missing data and unsupported inference still need review. AI output should not be treated as primary evidence.

Should market research come before competitor analysis?

Usually yes. Market research defines the market and customer context. The competitor teardown then compares the most relevant alternatives in detail. The work may overlap and can be updated together.

How often should market research be updated?

Update it when the target customer, geography, product category or route to market changes. Time-sensitive facts such as regulation, pricing and market estimates should carry a reviewed date.