How to write a market research brief (and what to ask for back)

Some of the clearest starting points behind our case studies fit in a few lines. Kalise wanted to know where its brand really stood in ice cream and in yogurts before planning its next stage of growth, with a measure it could repeat every year. Finom was rebuilding its brand and value proposition with a design agency, and before the design went any further it wanted to hear from the entrepreneurs it was designing for. Ironhack wanted to publish expert content on tech careers all year, in every country where it teaches, for two very different readers, and it needed the evidence by January.

Put like that, none of them names a method. Each one says what the answer is for and when it’s needed. That’s most of what a market research brief has to do, and it’s the part that often goes missing.

Why this is a challenge

If you’ve sent a brief to three firms, you may know the result. One proposal is a survey of a thousand people, another is twenty interviews and the third is social listening with a workshop at the end. The prices are far apart and each one sounds reasonable on its own.

The Market Research Society’s guide for buyers describes how this works. Buyers often send a brief to several firms precisely because they want the best approach, so, in the guide’s words, “the supply chain writes the specification, and commissioners choose the best balance of research design and cost.”¹ That works when every firm is answering the same question. Every gap in the brief is a decision the firm makes for you, and three firms make it three different ways.

Briefs are often less clear than the people who write them think. In the BetterBriefs survey of more than 1,700 marketers and agency staff in over 70 countries, published in 2021, 80% of marketers said they write good briefs, and 10% of creative agencies agreed.² That study was about briefs for creative agencies, and it measured perceptions, but the gap it describes is easy to recognise in research.

What to put in a market research brief

You don’t need a template, just six things, most of which fit in a sentence or two.

1. The decision, what depends on it and what you need to know. David Smith’s article in the MRS Research Buyers Guide puts it well: say what you expect to be able to do with the evidence once the project is finished, then list the questions you need answered.³
2. What you already know. Previous studies, sales data, the market sizing someone did last year and the hypotheses the team holds, including the ones you’d rather nobody tested. A firm that sees them can design around them instead of repeating them.
3. Who will use the results, and how. A board deciding once, a sales team in nine countries or a design agency working week by week need very different deliverables. Ironhack’s country teams each needed their own market and the global team needed the big picture, which is why that project produced nine country reports, two regional ones, a global one and one on gender.
4. When you need the answer. Give the date of the decision, which is often earlier than the date you’d like the report. If the work has to be repeated, as Kalise’s tracker is designed to be every year, say so now, because it changes how the first wave is built.
5. A budget range. The Social Research Association’s guide for commissioners recommends giving bidders a yardstick of the budget, and points out that doing so doesn’t reduce competition.⁴ Without one, each firm guesses the size of the project, and you end up comparing guesses.
6. What’s out of scope. Markets, segments or questions you’ve already settled, and any constraints such as languages, regulation, data protection (for example, if firms will recruit from your customer list) or internal politics.

Leave the method open. The same guide advises setting out objectives rather than methods.⁴ If you already think you need interviews, say why, so the firm can tell you if it disagrees.

A brief for a market entry decision can be as short as the one below. Switch between the two versions to see what each one tells the firms, and what each one gets back.

What to ask for back

Send every firm the same six questions. The answers will make the proposals comparable in a way the prices alone can’t.

1. Who exactly will work on it, and for how much of their time? The Social Research Association’s guide says that firms which field senior people to win a contract should make clear how far those people will actually work on it.⁴ The risk is real. In a 2025 review of 17 Canadian federal professional-services contracts, chosen with value and risk in mind, the Procurement Ombud found six where none of the people originally proposed did any work on the contract. In four of them, no reason was on file. In one, the person had resigned, and in another the department itself had asked for a replacement.⁵ That review wasn’t about research, but the safeguard carries over: put the named team in the contract, with substitutions only by agreement and by people of equivalent experience. Ask, too, what happens if a key person becomes unavailable, which matters at a small firm as much as at a large one.
2. Where will the participants come from, and how will they be checked? Online samples vary a lot in quality. In a 2019 study of more than 60,000 online interviews with US adults from six sources, Pew Research Center found about 4% to 7% bogus respondents in the opt-in sources, against about 1% in panels recruited by address, with no extra cleaning by the pollster. And 84% of the bogus respondents passed a standard trap question.⁶ In a 2022 Pew experiment with an online opt-in sample, 12% of the respondents under 30 claimed to be licensed to operate a nuclear submarine. The real share of Americans with that licence rounds to zero.⁷ Ask which panels or recruiters will be used, in what share, and what happens to suspicious answers. If a panel is involved, ask for its answers to ESOMAR’s 37 Questions to Help Buyers of Online Samples⁸ and, if it claims ISO 20252, which parts of the work the certificate covers.⁹
3. What will we actually receive? Ask to see a real deliverable from a past project, with the client’s details removed and that client’s permission. The industry’s Global Data Quality procurement guide recommends asking for example deliverables, along with named, experienced researchers and transparency about subcontracting.⁹ Ask, too, what you’ll own at the end: the raw data, the recordings and, for a tracker, the questionnaire, so you can move the tracker to another firm without losing the trend.
4. What won’t this study answer? Every method stops somewhere, and a good proposal says where. Each case study on our site ends with a note on the one lens it used and what the other two would have added, which is another way of saying what it couldn’t answer. You should hear that from any firm before you sign, not after.
5. Where is AI used, and who checks it? The same guide tells buyers to ask for disclosure of all AI used in the research, including how its outputs are checked.⁹ Ask where it touches your data, and which steps are done and checked by people. Our own answer is in our AI Manifesto.
6. What does the price assume? A fixed price with a date is easier to compare, as long as you know what sits behind it: the number of markets, how hard the participants are to find and what happens if that turns out harder than expected.

How to compare three proposals

Read the first page of each one and check how accurately it restates your decision. If a firm can’t say your decision back to you, the rest of the proposal is answering a different question. Then compare the answers to the six questions, and only then the prices. It helps to tell the firms in the brief how you’ll compare them. The Social Research Association calls that good practice, and the Global Data Quality guide suggests scoring every proposal on the same criteria: method, team, data quality, value and flexibility.⁴⁹

One more thing worth knowing: if the firms are MRS Accredited Company Partners or their researchers are MRS members, rule 10 of the Market Research Society’s Code of Conduct says they mustn’t knowingly carry out or commission work based on another organisation’s proposal without permission.¹⁰ If people on your own team are MRS members, it applies to them too. Taking the best design to the cheapest bidder can cost you more goodwill than it saves.

Once the study is done, our Research Blindspot Checklist helps you judge whether it’s ready to carry the decision. When a brief reaches us, you get a reply within 48 hours, and the proposal comes with a fixed price and a date we commit to.

Write the brief around the decision, and the three proposals will finally be answering the same question.