TL;DR:
A research brief has six sections: question, context, method, findings, open questions and sources, in that order.
Write the question and method before you start, and the findings last, so the brief shows what you decided and why.
Keep it to one page where you can. A reader should know the answer and how sure you are in the first minute.
A teammate asks "can you look into whether people actually use reminders?" Two weeks later you have a folder of notes and they have forgotten the question. A research brief fixes that. It is a short document that states what you were asked, what you did, what you found and what is still unknown, in the same six sections every time.
When a brief is worth writing
Write one whenever the work will be read by someone who was not in the room, or by you in three months. A short question you answered in an afternoon still deserves a half page: the question and the answer in one place is how work gets reused instead of redone.
Set up the page
Make a blank document (any word processor or notes app will do) and paste in the six headings below before you research anything. Filling the first three sections first forces you to agree on the question with whoever asked. Findings and open questions are written last.
Title: <short, specific>
Date / author:
1. Question
2. Context
3. Method
4. Findings
5. Open questions
6. SourcesWhat each section does
Question: One sentence, ending in a question mark, that can be answered. Prefer "Do new users open the reminders screen in their first week?" over "Look into reminders".
Context: Why anyone cares and what decision depends on the answer. Two to four sentences. Include what is already known so nobody repeats it.
Method: What you looked at, how, and what you left out. Enough that someone else could repeat it. Dates, sample sizes and any filters belong here.
Findings: The answer first, then the evidence. Lead with one sentence a busy reader can stop at, then support it with numbers, quotes or short lists.
Open questions: What you could not answer, what surprised you, and what you would check next. This protects you from being quoted as more certain than you were.
Sources: Every document, dataset, link or interview the brief rests on, with dates, so a reader can check your claims.
Worked example: the six sections filled in
The project and the numbers are invented to show the shape. Here is a one-page brief on a small app, written the way you would hand it over.
Section | What the example says |
|---|---|
Question | Do new users open the reminders screen during their first week? |
Context | The team is deciding whether to simplify onboarding. If most new users never see reminders, a tutorial for that screen is not the best use of the next sprint. |
Method | Looked at the activity export for the 120 accounts created in September. Counted accounts with at least one visit to the reminders screen within 7 days of sign-up. Left out accounts created by staff. |
Findings | 30 of 120 accounts opened reminders in week one, which is 30 ÷ 120 = 25%. The other 90 did not. The sample is one month only. |
Open questions | Why did the other 90 skip it? The export does not say. Would the answer differ for accounts that came from a referral? A few short user calls could answer the first one. |
Sources | Activity export, 3 October (saved in the team folder). Staff account list, same date. |
Check the one calculation: 30 ÷ 120 = 0.25, which is 25%. And 120 − 30 = 90 accounts did not open the screen. The finding sentence states both, and the method says which accounts were excluded so the denominator is clear.
Look at what the example avoids. It does not say "users do not care about reminders", because the data only shows that they did not open the screen. The open question says what the data cannot tell you. That distinction is the most valuable thing a brief gives its reader.
Treat the brief as a living document for a short time. If a new fact changes the answer, update the findings and add the date of the change, rather than writing a second brief that contradicts the first.
Mistakes to avoid
A question you cannot answer. "Understand our users" is a topic. Rewrite it until a specific result would answer it.
Findings before method. Readers cannot weigh a number until they know where it came from.
Burying the answer. Put the answer in the first line of the findings section, then the support.
Missing denominators. "30 accounts" means little without "of 120".
Hiding the gaps. If something is unknown, say so in open questions rather than leaving it out.
Doing it faster
Save the six-heading skeleton as a template file or a duplicate-ready document, and start every project by copying it. If you like talking a draft through, a voice assistant that sees your screen, such as DuckyHelper, can listen while you describe your method, but you should still read the finished brief against your sources yourself.
Frequently asked questions
How long should a brief be?
As short as it can be while keeping all six sections. One page is a good target. Longer evidence, such as raw notes or tables, can sit in an attachment that the sources section points to.
Can I write the findings first and fill the rest in later?
You can, but the question and method are easier to get right before you have an answer you like. Writing them first keeps you honest about what you set out to do.
Get started
