How to write a useful research brief

1. Start with the decision, not the topic

A research brief is a short explanation of what you want to understand and why. It gives research a direction without deciding the answer in advance. You do not need specialist language or a long document; you need a question that another person can interpret consistently.

“Research artificial intelligence” is a topic. “Help us compare whether these tools meet our customer support needs” describes a decision. Write down what you expect to do differently after reading the work. If you are exploring rather than deciding, say what you want to understand better.

2. Name the reader and their starting point

Explain who will use the answer and what they already know. A founder considering a new market may need different context from a technical team comparing products, even when both ask about the same industry.

Include the reader’s priorities and any vocabulary that needs explaining. Avoid assuming that a familiar internal abbreviation will mean the same thing to everyone. A sentence about the audience can prevent several pages of unnecessary background.

3. Give the question workable boundaries

Set the place, period, options and exclusions that matter. “Compare three tools for a small team in Singapore” is easier to scope than “find the best software.” Explain why those boundaries are useful, and flag the ones you are willing to revise.

Distinguish the deadline for making your decision from your preferred research delivery date. State constraints honestly. An answer based on public information may not resolve a question that depends on confidential pricing, a product trial or direct interviews.

4. Say what evidence would help

List the criteria you want to compare before choosing a favourite option. Criteria might include documented capabilities, relevant costs, practical limitations or compatibility with an existing process. Separate requirements from preferences.

Share useful starting sources and previous work, but do not require a conclusion that confirms them. Ask the researcher to identify information gaps, conflicting accounts and assumptions. “We do not yet know” can be a more useful answer than an unsupported ranking.

5. Describe a useful output

Explain what you want to be able to read or use at the end. A comparison table, a short evidence summary and a list of unresolved questions may serve your decision better than a long general report.

Specify where source references and limitations should appear. If you want a recommendation, say which criteria it should follow and whether the researcher may recommend further investigation. This makes room for an answer that is careful rather than falsely definitive.

A fictional example: comparing support tools

The following brief is fictional. It illustrates how to frame a question; it contains no product findings, client work or recommendation.

• Decision: Should our fictional five-person support team trial one of three shortlisted tools?
• Audience: The operations lead and support team. They know the current workflow but have not compared these products systematically.
• Scope: Compare the three named options using current public documentation. Focus on shared inbox features, integration requirements and published pricing. Exclude a full security audit or any claim about performance that would require testing.
• Criteria: Must support the team’s existing email workflow. Treat automation features as a preference, not a requirement. Flag any pricing that needs a supplier quote.
• Output: A comparison table with references, a short explanation of trade-offs and questions to test during a trial. We will set the deadline before agreeing the work.

6. Check the brief before sharing it

Read your question as someone who has not attended your meetings. Can they identify the decision, reader, boundaries and expected output? Remove any sentence that assumes the result, and clarify terms that could have several meanings.

Leave out passwords, unnecessary personal information and confidential documents from an initial enquiry. A useful first brief explains the need without exposing material that is not required to scope it. You can revise it when the question becomes clearer.

Prepare your own starting point

Use the New Eruditions brief builder to turn these ideas into a draft you can copy or download. Drafting is a planning step: it does not submit information, place an order or commission research. Our company launch is planned for November 2027; reports and commissions are not available yet.