How to brief a research firm
so the answer is useful
A good research brief states the decision the research must inform, the claims it must test, the evidence standard it must meet and the date by which it must arrive. This guide explains how to write one in a page, what to ask providers before signing, and how to judge the proposal you get back.
- One page: decision, claims to test, evidence bar, timetable, budget
- Write the decision as a sentence with a verb: enter, buy, price, launch, stop
- Ask every provider how participants will be verified and how AI will be used
- Judge proposals on who will do the work and who will sign it
- Theory replies to any brief within one business day with a fixed price, a start date or a recommendation not to proceed
Cite this: Theory Intelligence, "How to Brief a Research Firm", theoryintelligence.com/guides/how-to-brief-a-research-firm/, 2026-09-09.
State the decision
Most research briefs describe a topic: the European mid-market payments sector, customer attitudes to our new tier. A topic produces a report about a topic. A decision produces an answer. Write the decision as a sentence with a verb and a date: we will decide by 30 November whether to raise list prices by 15 percent for the enterprise tier. Everything in the brief follows from it.
Name the claims
List the things you currently believe that, if wrong, would change the decision. Enterprise customers have no credible alternative. Procurement will not escalate a 15 percent rise. Competitors will not undercut in response. Five to eight claims is normal. These become the research questions, and the report should return a verdict on each.
Set the evidence bar
Say what would convince you. Verified interviews with fifteen procurement leads across three segments. Documents rather than recollection where possible. No online panels. Verbatim quotations usable in the board pack. If you do not set the bar, the provider will set it at whatever is easiest to deliver.
Fix the timetable and budget
Give the date the decision will be taken and the date the research must land to be useful, which is usually two weeks earlier. Give a budget range. Providers who know the budget propose within it; providers who do not propose the maximum they think you will bear.
Ask five questions of every provider
Who will run the interviews, by name. How will participants be recruited and verified, and will the method be recorded. Where will AI be used, and where not, in writing. Who will sign the recommendation. What is the total price and what would change it.
Judge the proposal
A good proposal restates your decision in its own words, lists the claims it will test, names the people who will do the work, describes verification, discloses AI use, gives a fixed price and a date, and tells you if any part of the brief cannot be delivered as asked. A proposal that offers a comprehensive view of the market and a team of experienced professionals has not read your brief.
How Theory responds to a brief
Send the decision to info@theoryintel.com. A researcher replies within one business day with a fixed price and a start date, or with a recommendation that you do not need us. Our formats are published: the Signal Brief for one question, the Decision Study for one consequential decision, the Intelligence Retainer for standing questions. See also the FAQ.
Questions
How long should a research brief be?
One page. If it needs more, the decision has not been stated clearly enough.
Should I share the budget?
Yes, as a range. It produces proposals you can compare. Theory's prices are published, so with us the question is which format fits.
What if I do not know which claims to test?
Send the decision anyway. A good provider will help you list the claims at scoping. Theory does this at no charge before quoting.
Want this done for you?
Write to info@theoryintel.com with the decision. A researcher replies within one business day with a fixed price and a start date.
Start a brief