Before a first client meeting, I would bring the meeting purpose, selected evidence and open questions into a short working brief. For an independent consultant or advisor, this is an internal preparation note. It gives me a place to see what supports the discussion and what I still need to ask.
I would decide separately what, if anything, to share with the prospective client. An assembled brief still needs its important claims checked before I rely on it.
Begin with selected inputs
If I am still deciding what deserves research, I would use the pre-meeting research checklist and its copyable research plan. Here, I am working with the inputs already selected for a particular meeting question.
I would choose relevant primary pages or authorized records. Each source needs an actual title or record identifier and a usable link or location. I would record the information date separately from the date I checked it. A date that is not supplied stays unknown.
Keep each statement beside its source
I would put a source reference beside each factual statement, with a passage or section I can find again. A quotation needs the original wording; a paraphrase needs the same care about what the source supports.
My possible explanation belongs in a separate interpretation field. I can leave that field empty when the inputs offer no useful basis for an explanation. An open question can remain unanswered in a usable preparation brief.
Copy the working brief
I can copy this form into my own note and add only what the meeting needs. It requires no signup, download or submission.
Brief label and version:
Meeting or contact, if known:
Meeting purpose:
Which discussion or decision am I preparing for?
Source-supported context:
For each statement, add the source identifier and passage location.
Source details:
For each identifier, record the actual title or record name,
link or authorized location, information date and date checked.
Keep unknown dates visible.
My interpretation, if useful:
What is my possible explanation, separate from the source?
Open questions:
What could change the conversation? Which answers are still missing?
Assembly and claim-review state:
Assembly: in progress or complete.
Claim review: pending or completed for the identified statements.
Record the actual reviewer, review date and material changes.
Version and refresh note:
What changed from the earlier brief, or needs checking again?I would keep the brief short enough to use before the meeting. A longer source collection can stay separate, with references connecting it to the relevant statements.
A fictional example, before the first meeting
In this illustration, I am the imaginary consultant assembling a preparation brief. The only input is a fictional prospective client’s public services page. Its complete mock text is:
Reporting services are offered.
The type of report is not supplied. This is a page description, not something a client has said in a meeting. No real company, page, source date, conversation or result is represented.
For the example, the consultant has chosen report review as a possible discussion topic. I would first confirm whether that topic is useful to the contact. The page supplies no internal review steps or evidence of a problem.
| Brief field | Completed fictional entry |
|---|---|
| Meeting purpose | Confirm whether report review is a useful topic for the first conversation |
| Source-supported context | Fictional page A describes a reporting service. Internal review details are not supplied. |
| Source details | Fictional page A, the complete mock text above. No actual URL, information date or date checked is supplied. |
| My interpretation | None entered from this limited input. |
| Open questions | Is report review relevant to this conversation? If so, what happens before a report is released, and is a review step part of that? |
| Assembly and claim-review state | Illustrative assembly complete. No real source or client claim has been verified. |
| Version and refresh note | Fictional first preparation brief, before discovery. The questions remain unanswered. |
I would carry that narrow context and the open questions into preparation. Missing detail on a public page supplies no finding about how the client’s operation performs.
Use AI for assembly, if useful
With material authorized for the particular tool and use, I could ask:
Organize only the material I supply into the brief fields above. Keep a source identifier beside each factual statement. Separate source-supported context, my interpretations and open questions. Preserve missing dates and answers as unknown. Leave interpretations empty when the inputs do not support one. Mark claim review as pending. Invent no links, quotations, client goals, problems or recommendations.
I would compare the assembled draft with the supplied passages and check that each identifier points to the source it names. I can also fill the form manually. Before uploading private material, I would confirm authorization for the tool, information and intended use.
Review the important claims before using the brief
For each important statement, I would open the source, check its wording and dates, and retain the limits on what it establishes. A reference that does not support the statement needs a narrower claim or removal. A consequential gap can become a neutral meeting question.
I would record which statements were reviewed, by whom and when. Assembly complete and claim review complete are separate states. Neither settles an unanswered client question.
My handoff would contain the brief, usable source references and the questions still worth asking. The brief supports discovery; it does not establish a diagnosis or agreed work. When the conversation adds information, I would retain the earlier version and record what changed.
For background, the NIST AI RMF Core discusses documentation and human review in AI system teams. This client-brief format is my proposed working aid.