How to write a market research agency brief

State the decision and its owner

Replace a broad request such as “understand our customers” with a decision someone will make. Examples include choosing between product propositions, identifying which market to investigate further or deciding how to improve onboarding. Name the person or group responsible and the point at which they need evidence.

Explain what is open to change. If the product name is fixed but the message is still flexible, say so. If leadership has already approved a launch, the study may need to refine execution rather than revisit whether to launch. Clear boundaries prevent a thoughtful proposal from solving the wrong problem.

Share existing evidence and competing explanations

Summarize relevant previous research, customer feedback, product analytics and commercial information. Distinguish observations from interpretations. “Trial users stop at setup” is an observation; “the setup is too complicated” is an explanation to test.

Include the explanations your team disagrees about. Ask the agency how it would distinguish them and what evidence could change the decision. Share useful source material through an agreed process instead of sending an unfiltered archive and expecting the supplier to discover the brief.

Define the audience and the boundaries

Describe who needs to participate, where they are and what relevant experience they must have. Specify essential differences such as customer status, professional responsibility, purchase recency or product usage. Mark requirements as essential or preferred so the agency can identify recruitment tradeoffs.

State whether you can supply contacts, introductions, a prototype or access to an environment. Ask the agency to explain what it will verify, what you must provide and what happens if the intended audience cannot be reached. Leave the proposed sample size open unless there is a clear analytical reason for fixing it.

Set the scope, budget and deadline

Give a working budget range or spending boundary when you can. Explain whether it covers the whole project, including recruitment, incentives, translation, travel, tools and reporting. If the budget is still being set, request a recommended design and a smaller alternative, with the implications of each.

Use milestones instead of a single delivery date. Include proposal selection, kickoff, materials approval, recruitment, fieldwork, analysis and the decision meeting. Name the client-side approver and expected review time. Ask suppliers to flag dependencies rather than absorb uncertainty into an unexplained estimate.

• Separate core work from optional markets, audiences or additional study waves.
• Request recruitment and incentive assumptions, including cancellations and replacements.
• Ask what would trigger a change in scope or fee and how that change would be approved.
• Compare the same deliverables and responsibilities before comparing proposal totals.

Specify deliverables and how they will be used

Ask for outputs that support the decision. A senior leadership discussion may need a concise argument and clearly stated uncertainty. A product team may also need prioritized issues, examples and access to supporting evidence. Specify formats and audiences rather than requesting an impressive-looking report by default.

Agree which project materials should be delivered or accessible: instruments, analysis tables, transcripts, recordings or a repository handover, as appropriate to the study. Ask the agency to explain practical restrictions, access arrangements and retention. Clarify who will review automated analysis and own the final interpretation.

Example brief: improving product onboarding

Decision: Our product lead must decide which of two onboarding approaches to develop next. We can change the sequence and guidance, but the account requirements are fixed. The decision meeting is already scheduled.

What we know: Existing analytics identify where users leave the process, but do not explain why. Internal teams disagree about whether users misunderstand the value, struggle with the workflow or encounter a requirement they cannot meet.

Audience and scope: Include people with recent relevant experience, covering both successful and incomplete onboarding. We can supply prototypes and some customer contacts. Please recommend the study design and recruitment mix, explain the limitations and identify what our team must provide.

Outputs and proposal: Recommend changes with supporting evidence, distinguish observations from interpretation and include a working session with product and design. Price the recommended scope and a reduced alternative. State assumptions, milestones, named delivery roles and exclusions.

Name the research tools your team will use

If your team already uses Versive, include it in the brief. The onboarding example above could combine real-participant Figma or website tasks, survey questions and AI-moderated interviews in Versive. State whether your team or the agency will build the study, recruit participants, review the responses and produce recommendations.

Your own customer contacts can receive study invitations through Versive, and integrated panel providers offer another sourcing route. Separate AI persona tests can help explore early design questions; specify when the project needs observation and feedback from real participants. Sharing the intended tools helps agencies scope the work and handover accurately.

Review assumptions and clarify the proposal

Assess proposals against criteria you set before seeing them: understanding of the decision, audience feasibility, suitability of the method, clarity of analysis, delivery-team experience and commercial scope. A supplier that identifies a weakness in your brief may be adding value; ask it to explain the alternative.

Hold a focused clarification discussion with the likely delivery team. Resolve different assumptions in writing and update the agreed scope before work starts. The aim is a common understanding of what will be learned, what will remain uncertain and who will turn the findings into action.

Common questions