How to do market research for a business idea

You have a new idea. Maybe it is a product line extension, maybe it is a feature your sales team keeps asking for. Before you build anything, you need to know whether the demand is real or just a hunch wearing a confident voice.

This guide walks through a focused way to test a specific idea: write down your assumptions, rank them by risk, pick the right audience, ask about the problem before the solution, and read your results with a clear head.

It is not a tutorial on running a full market research project from scratch. For that broader process, see our market research process guide or the market research hub.

Every idea you are excited about is really a stack of assumptions wearing a trench coat. Before you write a single survey question, pull those assumptions out into the open where you can actually look at them.

Go through your idea and list every claim you are treating as fact. An existing business testing a new subscription tier, for example, might be assuming that current customers want it, that they would pay a specific price, and that it would not cannibalize an existing plan.

Write these down in plain language. An assumption you can say out loud without flinching is one you are ready to test.

A testable assumption is specific enough to fail. Instead of "customers will like this," try "at least 30% of surveyed customers would switch to this option within six months."

Vague assumptions produce vague survey data, which is really just research for a business idea in name only. Specific assumptions give you something a question can actually answer.

You cannot test every assumption at once, and you should not try. Rank them so your first study targets the one that could sink the idea, not the one that is simply interesting to know.

For each assumption, ask a direct question: if this turned out to be false, would the idea still work? Assumptions where the answer is no are your highest-risk assumptions, and they go first.

Score every assumption on two dimensions: how likely you think it is to be wrong, and how much damage it would do if it is wrong. Multiply roughly, high times high, and you have your priority list.

This two-factor approach to researching a business idea keeps you from spending your first round of responses on a low-stakes question just because it felt easier to write.

Once your assumptions are ranked, put them into a table. This becomes your working plan and keeps your team aligned on what you are actually testing, and why.

Keep this table short enough to act on. Three to five rows is usually plenty for a first round of market research for a business idea, and it is the same working structure teams use inside SurveyMonkey LaunchPad to run a test like this at scale.

Your table is only as good as the people you send it to. Before you launch anything, decide who can actually tell you something useful.

Start with these three groups, in roughly this order:

A business that already sells project management software, for example, would likely start with current users before testing a net-new audience for a completely different product category. Testing strangers first, before testing the people who already trust you, wastes your most informative respondents.

It is tempting to describe your idea and ask "would you buy this." That question flatters the idea and tells you almost nothing about whether the underlying problem is real.

Open with questions about the problem itself, not the fix:

Only after those answers come in should you introduce the concept and ask reaction questions. This order keeps respondents from simply agreeing with whatever you show them.

You do not need a 40-question survey to validate one idea. A short, focused set respects your audience's time and gets you cleaner answers, in line with the broader market research use cases teams run on SurveyMonkey.

A workable starting set looks like this:

Keep the whole set under 10 questions where possible. If your idea is a new product or feature, our product testing template is a faster starting point than a blank page, and a shorter set tends to see higher completion rates than broad coverage ever will.

A completed survey can feel like an answer. Treat it as a signal instead, especially on a first-round study with a modest sample.

Keep three limits in mind before you act on the data:

A survey tests demand signals. It does not guarantee the idea will succeed once it is in market, and no honest research process would claim otherwise.

Once your highest-risk assumption has been tested, make an actual decision instead of running another round of research to avoid making one.

Use this rough logic:

Write the decision down along with the evidence behind it. That record is what turns one study into a repeatable way of researching a business idea the next time you have one.

Writing assumptions on a whiteboard only gets you so far. SurveyMonkey features let you turn a ranked assumption list into a real question set, send it to the audience you defined, and see results come in with AI-powered analysis surfacing the themes in open-ended answers automatically.

Once you are ready to run the fuller research process, from setting objectives through reporting, our market research process guide walks through each stage in depth.