In this guide
How to Write a Research Brief for a Technical Search (With Template)
· ForesIQ
A research brief turns a job description into something you can search for. It names the problem the new colleague will own, the research topics behind it, the evidence that would make someone worth a conversation and the adjacent fields worth exploring. One page, written with the research lead, saves weeks of misdirected sourcing.
Job descriptions are written to attract applicants. They list responsibilities and requirements in general terms, which is exactly what makes them hard to search from. “Experience with large language models” matches an enormous number of people and distinguishes none of them.
A research brief is an internal working document. Its only job is to make the search precise. This guide walks through the seven fields that matter and gives you a template to reuse.
When you need one
Write a research brief whenever the role depends on specialist knowledge that a title does not capture: a research scientist for a new group, a specialist for a problem the team has not solved or a leader for an area the organisation is entering. For well-understood roles with clear titles, a normal intake is enough.
The seven fields of a research brief
1. The problem. What is the team trying to do that it cannot do today? Write it as a problem, in a sentence or two, in the research lead’s own words.
2. What the person will own. The part of that problem this colleague will be responsible for. It separates someone who will set direction from someone who will strengthen an existing effort.
3. Research topics. The specific topics and methods connected to the problem, in the vocabulary researchers use. Aim for three to six. These become your search terms.
4. Evidence worth a conversation. What would make the research lead want to talk to someone? It might be published work on a named topic, experience at a certain scale or a record of leading projects. Be concrete.
5. Adjacent fields. Where else might the right person come from? Strong candidates often sit one field over. Someone in computational imaging may suit a 3D vision problem.
6. Standing and trajectory. Is the team looking for an established name, a rising researcher or either? The answer changes the search completely, so ask it outright.
7. Constraints. Location, working pattern, level and timing. Keep these last, so they narrow a well-defined search instead of defining it.
A template you can copy
| Field | Prompt | Illustrative answer |
|---|---|---|
| Problem | What can the team not do today? | Policies trained in simulation fail on the real robot. |
| Ownership | What will this person be responsible for? | Closing the gap between simulated and real-world performance. |
| Topics | Which research topics connect to it? | Sim-to-real transfer, domain randomisation, robot learning. |
| Evidence | What would earn a conversation? | Published work deploying learned policies on physical hardware. |
| Adjacent fields | Where else might they come from? | Reinforcement learning, control, computer vision for robotics. |
| Standing | Established, rising or either? | Rising researcher with recent first-author work. |
| Constraints | Location, level, timing? | Hybrid, one of two sites, start within six months. |
The answers above are illustrative. Yours should be written in your research lead’s vocabulary, because that vocabulary is what you will search with.
Running the briefing conversation
Book thirty minutes with the person accountable for the research. Four questions do most of the work:
- “What is the team stuck on that this person would help unblock?”
- “If you could read one paper by the ideal person, what would it be about?”
- “Whose work do you admire in this area, and what is it about their work?”
- “Would you talk to someone from a neighbouring field who had solved a similar problem?”
The third question calibrates the search. You are asking what good looks like, and the names are examples of a pattern. The pattern is what you take away.
Common mistakes
- Starting with constraints. A search defined by location and level before the work is understood returns people who fit the logistics and miss the point.
- Accepting broad topics. “Machine learning” is a field. “Preference optimisation for language models” is a topic. Push for the second kind.
- Writing it alone. A brief written without the research lead encodes your guess about the role.
- Never revising it. The first few profiles you share will sharpen the brief. Update it.
Frequently asked questions
What is a research brief?
A one-page internal document that translates a specialist role into a searchable definition: the problem, what the person will own, the research topics, the evidence that matters, adjacent fields, standing and constraints.
How is it different from a job description?
A job description is written to attract applicants and stays general. A research brief is written to guide a search and is as specific as possible.
Who should write it?
The recruiter or sourcer, together with the person accountable for the research. It needs the research lead’s vocabulary and the recruiter’s structure.
How long should a research brief be?
One page. If it runs longer, the role probably contains two jobs and is worth splitting.
How specific should the research topics be?
Specific enough that a researcher would recognise them as their area. Three to six named topics is a good target.
Explore your research brief in ForesIQ
ForesIQ is a Research Talent Intelligence platform for frontier technology teams. Its specialist database brings expertise, field-relative standing, Momentum, Activity and supported career evidence into one workspace. Current areas include computer vision, machine learning, natural language processing, systems and infrastructure, and robotics.
Bring your brief’s topics into ForesIQ and see the researchers whose work matches them. Request A Demo and we’ll send you a personalised demo showing how ForesIQ can support specialist research talent discovery.