A research workflow

From five saved sources to one useful project brief

How I would turn a small reading pile into a decision without smoothing away the gaps.

A person using a laptop while writing in a paper notebook.
Photo by MJ Duford · Unsplash License. Illustrative photograph.
Contents

Five useful sources can still leave me with an empty page. The missing piece is often a decision about what the brief is meant to help someone do. I would settle that before turning highlights into polished sentences. Otherwise, a neat summary of the reading can become a very long detour around the actual question.

Consider a constructed project: a volunteer group is deciding whether to trial a monthly repair meet-up. Imagine it has a venue email, a local timetable, a materials list, an organizer interview and a booking policy. These are invented source types, not findings about a real group. They give us a concrete pile to work through.

Give the brief one job

I would write its job at the top: “Help the group choose whether to investigate a trial session.” That is narrower than proving the event will succeed. It suggests an appropriate ending: a short recommendation about further checks, with clear conditions. It does not require a false certainty about attendance or costs.

The intended reader changes the brief. A venue manager may need a practical outline of space, access and responsibility. A volunteer may need to know the role being proposed. I would choose one main reader and make other needs explicit rather than trying to produce a document that answers every possible question.

Give each saved source a role

I would label the venue email as evidence about what that venue has offered, not about venues in general. The timetable can inform possible dates, not tell me who will attend. The interview might suggest useful questions, but one organizer's account does not establish a universal outcome. These small limits are where a brief becomes trustworthy.

Zotero collections can group an item into more than one project without making separate underlying records. That is useful if the booking policy already belongs to another project. I would reuse the record while writing a new note about its role here. Reuse should preserve the source, not quietly reuse an old conclusion.

Build a small evidence map

Before prose, I would make a table that pairs each decision with its current support. The demonstration below has no invented numbers. Its gaps are intentional. Filling those gaps with estimates would require an explicit basis and a clear label, rather than a confident tone.

This table is a planning device, not a scoring method. A brief with five weakly relevant sources can still need one important conversation. The useful question is whether the right issue is supported, not whether every row has a link.

Decision questionPossible supportStill unknown
Is the room available?Venue emailConditions of the offer
Can volunteers attend?Proposed timetableConfirmed commitments
What is permitted?Booking policyAny event-specific approval
What should we prepare?Materials list and interviewWhich activities are in scope

Separate the source note from the recommendation

Zotero supports both source-attached notes and standalone notes. I would use that distinction to keep the venue's words close to its record, then draft a project brief that points back to them. A change in the recommendation should not require rewriting the underlying evidence as though the venue said something different.

My draft might say: “Investigate one trial date, subject to venue confirmation and volunteer commitments.” That sentence is a proposed action under stated conditions. It is not an event booking, a promise that the policy permits the activity or an inference that volunteers have agreed. I would make those boundaries plain to the reader.

Write the short version first

I would start with four parts: the decision, what currently supports it, the unresolved conditions and the next person to ask. A long background section can wait. In this example, a paragraph explaining the idea is useful; a history of every repair event ever mentioned in the research is unlikely to help the group choose its next step.

The brief should also say what was left out. Perhaps the group has not selected activities or estimated materials. Recording those omissions prevents a reader from treating a preliminary note as a complete operating plan. Scope is information, not an admission that the writing has failed.

Check every confident sentence on the way out

I would read the brief once as the person expected to act on it. For each sentence that sounds settled, I would ask where its support lives. “Available,” “approved” and “confirmed” deserve particular attention. If the source uses conditional wording, the brief should not remove it just to sound decisive.

Finally, I would keep the source register with the document. If an email changes, the next editor can identify the affected part without reconstructing the whole reading pile. The aim is a brief that can be revised responsibly, not a beautifully formatted page that becomes mysterious as soon as someone asks a question.

Run the example from source intake to first draft

I would begin the invented repair-meet-up project with a five-row register. Each row identifies the document, its origin and the question it can answer. The venue email concerns availability. The booking policy concerns permitted use. The timetable compares possible dates. The materials list and organizer interview help frame practical questions. This reveals a gap immediately: none of those sources, as described, confirms volunteers' commitments. Five relevant documents can support investigation while leaving the most important next agreement unresolved. The register is useful because it lets me see that without drafting several pages first.

That gap affects the opening. I would avoid “the team will run the event” simply because several sources are attached. A more accurate preliminary sentence says a possible trial is being investigated and responsibilities remain unconfirmed. The document can advance the discussion without making commitments for other people. I want it to provide an agenda, rather than become apparent evidence of an agreement nobody has made. This distinction also helps a new reader understand whether they are reviewing an idea, approving a plan or being asked to take a specific action.

I would draft the recommendation before the background. Here it could be to clarify venue conditions and invite explicit commitments before choosing a date. The supporting paragraph identifies the current offer and open questions. Background can explain why the idea interests the group, without claiming demand, savings or community outcomes the fictional source set does not establish. Those require separate support. A useful proposal can describe the question it wants to investigate without manufacturing an answer in order to justify the investigation. The strength of the brief comes from that sequence of decisions.

Only then would I add logistics. I would distinguish a materials list under consideration from purchases authorized, a suggested date from a booking and a possible activity from permitted room use. Those differences may look small, but matter to the people expected to act. The brief becomes more useful when it identifies descriptive information, proposals and conditions awaiting confirmation. I would keep the wording consistent throughout the document, because a careful opening loses its value if a later paragraph quietly treats the same possibility as an established commitment.

Make the source register usable by the next editor

I would choose stable labels such as venue-offer, booking-policy and organizer-interview, rather than identify sources by the order of browser tabs. These local identifiers supplement titles and source details. Their purpose is to keep references understandable when the brief changes order or another document is added. A plain table can do this. I would not introduce a complex numbering system unless the collection grows enough to require one; the identifiers need only be clear to the people who will inspect and revise this particular document.

For each source, I would record a supported point and a limit. The email establishes what its sender offered, while policy conditions may live elsewhere. The interview describes one organizer's experience. The materials list may be a proposed inventory, not a priced estimate. Keeping those limits beside the records means the next editor does not need to rediscover them from the entire background section. A source can be useful and limited at the same time. Describing that limit is part of making the material easier to reuse responsibly.

I would make change-sensitive information easy to spot. Availability, a contact person's instructions and an event-specific offer may need reconfirmation sooner than general background. That does not make them poor sources; it changes the check required before action. I would name who should reconfirm the offer and at what decision point. “Checked once” is not a permanent status for a changing arrangement. If the proposal is paused for several weeks, the register should help identify what needs refreshing before anyone treats the old draft as a current operating plan.

When others contribute links, I would ask for one sentence explaining which question each source helps answer. A link in a chat is useful input, but not yet a fully integrated part of the brief. The writer needs relevance and scope. That sentence can improve the collection more than a long naming rule does. It also welcomes useful contributions from people who are not maintaining the reference system: they can explain why they sent the material, while the person integrating it handles the consistent record and citation.

Resolve disagreements without smoothing them away

Suppose the fictional venue email sounds flexible while the booking policy appears restrictive. I would not pick the friendlier wording and proceed. I would check whether the email concerns the proposed activity, whether the policy is current and who can clarify the apparent exception. Until then, the disagreement belongs in the brief as an open question. A writer's confidence cannot supply permission. Nor does citing both documents make the inconsistency disappear; the action still depends on understanding which terms apply to this proposed use.

Another difference may concern circumstances. The interview could describe an experienced team while this proposal concerns beginners. Those statements need not conflict. They describe different conditions. I would identify the relevant difference rather than combine them into “these events are easy to run.” An example can suggest good questions without proving a similar result here. I would use the interview to examine preparation and responsibilities, then ask the group which of those conditions they can actually meet, instead of importing the interviewee's experience into the proposed team's biography.

My comparison note would state what each source concerns, the difference and the next check. If contacting the venue can resolve it, I would say that. If the group must choose its scope, I would call it a decision rather than imply another search will produce the answer. Some gaps are missing facts; others are unmade choices. Naming the distinction prevents research becoming a way to postpone an ordinary commitment. It also keeps the brief from growing indefinitely in response to a question that requires a conversation.

When the issue is resolved, I would update the recommendation and source note together. Leaving an old caveat in one place and new certainty elsewhere creates confusion. If the change affected the proposal substantially, a brief explanation should remain. That gives a useful revision record without making the document a transcript. The reader needs the present decision and enough context to understand why it differs from an earlier version. A short change note can preserve that relationship more clearly than retaining several nearly identical drafts with unexplained names.

Finish with a handover and a stopping point

The next action should be obvious. I would name the question for the venue, the person responsible and the condition for returning to the date decision. If nobody has accepted responsibility, the brief should say so. “Someone should check” identifies a need but does not create a handover. A practical document allows the group to see that difference and assign the work explicitly. It should also make clear which actions are suggestions and which have already been accepted, because a written task beside a name is not evidence of agreement.

I would set a stopping point for investigation. Perhaps one response about room use is needed before deciding whether to continue. The group need not collect every possible supplier quote first. The brief can sequence later effort behind an earlier condition. This is a planning choice for the illustrative project, not a feature promised by reference software. The software can preserve the materials; the brief's author remains responsible for explaining the dependency and avoiding work that no longer serves the current decision.

I would read the document as a new volunteer. Can they distinguish the idea from an agreed event? Can they find conditions, open questions and the origin of a practical instruction? Is the requested action visible, or buried under the history of the reading? If necessary, I would shorten the opening and move background into references. A longer document should contain more useful reasoning, not more obstacles between the reader and the choice they have been asked to assess. That is the test I would apply before adding another summary.

After discussion, the brief can become a decision record. I would preserve what was agreed, what remains conditional and the sources that matter to those points. Other material can stay in the research collection without equal prominence on the final page. Retaining context does not mean presenting everything at once. The result I want is a document that carries the current decision and a dependable route to its basis, with the group's actual commitments unmistakable. That makes it useful both immediately and when someone returns after the original conversation has faded.

Keep the final brief connected to its revision point

Before handing over the illustrative brief, I would identify what event should trigger a revision: a venue response, a group decision or a change in the proposed scope. That gives the document a practical life after the first draft without making every reader responsible for monitoring every source indefinitely.

I would also distinguish the reference copy from the active working document. The group should know where current changes belong and where the decision already made can be read. A clear revision point prevents a useful preliminary brief becoming an apparently settled plan simply because an old attachment continues circulating after the conversation has moved on.

Frequently Asked Questions

Q: Must a project brief have five sources?

A: No. Five is simply the size of this worked example. The appropriate sources depend on the decision. Several general articles cannot replace a specific confirmation from the person or organization responsible.

Q: Can one source support several points?

A: Yes. Record which part supports each point and keep the scope accurate. Reusing a source is reasonable; presenting one source as several independent confirmations is misleading.

Q: Where should I put conflicting information?

A: State the conflict beside the affected decision. Note the sources, dates and what would resolve it. I would avoid choosing the more convenient version merely to make the draft feel complete.

Q: How long should the brief be?

A: Long enough for its reader to understand the proposed action and conditions. Draft the short decision first, then add only the background, evidence and practical detail needed to assess it.

Q: Should the source register appear in the final document?

A: A usable route to the sources should remain. Depending on the audience, that might be references, linked notes or an attached register. Do not rely on a private browser session that only the writer can access.

One thing to take forward

I would stop collecting long enough to ask what the reader must decide next. Once that is clear, five saved sources become a set of roles, limits and checks. The brief can then make a useful recommendation without pretending the research has answered everything.

Sources & further reading

Something needs correcting? Write to the editors with the article link.