share/artifacts
PricingSign inConnect your agent
← All guides

From research to a decision

Make the client decision clear.

A client decision brief helps a client choose between options by making the recommendation, costs, tradeoffs, and evidence clear.

Make this with my agent

Bring your notes. Review a draft before sharing.

Explore the method
  1. 01
    The ask comes first.

    Who decides, by when, and what is still conditional.

  2. 02
    Compare on equal terms.

    Costs, timing, and effort sit side by side.

  3. 03
    Keep the unknowns visible.

    A recommendation is not an approval.

THE REAL HTML · FICTIONAL EXAMPLEView on its own

A worked example, not a customer result. September 2026 dates are sample facts.

Now make it about your work

Your context. Your agent. Your brief.

Bring the decision, the options, and the evidence. Let your agent help with what’s missing.

Start with three things

  1. The choice your client needs to make.
  2. Your notes, estimates, and credible alternatives.
  3. The constraints and unknowns that matter.
Make this with my agent

You review the facts and access. Nothing is published without your approval.

The method behind the briefFrom evidence to a defensible choice

How do I turn research into a client decision brief?

Define the choice, who can authorize it, and the client's criteria. Give your agent the evidence and viable options; compare them before writing a recommendation. Lead the brief with the decision requested, the proposed answer, and its critical condition. If the evidence cannot support a choice, ask for the next investigation instead. Review the analysis, then check that a fresh reader understands the tradeoff and next action.

What to give your agent

  • The exact choice, decision maker, requested timing and reason for it, and what approval would authorize. Mark missing authority or dates as unresolved.
  • Client-approved outcomes, preferences, and hard constraints. Include disputed priorities rather than letting the agent choose them silently.
  • Dated notes, measurements, and quotes, with source links and sharing permission. Include currency, cost period, scope, exclusions, and client effort where known.
  • Credible alternatives and the consequences of continuing current arrangements. Include existing recommendations, counterarguments, and unknowns that could change the choice.

  1. Name the decision and the authority

    Write a concrete request, not a topic: 'Robin, choose A or B by September 11; if B, name the routing owner.' State the scope of that choice. Selecting an approach is not automatically approval of a final quote, contract, or extra spend. A decision brief supports a pending choice; a decision record documents one actually made.

  2. Agree on constraints before ranking preferences

    Use the client's stated outcomes. Separate requirements an option must satisfy from preferences that can be traded. An attractive benefit cannot cancel a breached hard constraint. If priorities conflict, show the disagreement and the decision it requires. Do not invent weights, force a third option, or treat continuing the current arrangement as cost-free.

  3. Compare like with like

    Give credible options the same criteria, currency, time horizon, and scope. Separate one-time fees, ongoing costs, exclusions, and client effort; unknown is not zero. In Cedar Studio's example, $4,500 is below $5,000 before exclusions, not proof of an affordable total. Preserve conflicting estimates with their sources rather than selecting the tidiest number.

  4. Challenge and verify the recommendation

    Analyse the options before writing the conclusion. Trace decisive facts to original inputs, recalculate totals, and check the proposed option against hard constraints. Explain the strongest case against it and what would change the recommendation. Use a numerical threshold only when supported by a model; otherwise name the condition. If evidence is insufficient, propose a bounded next investigation.

  5. Write for the reader, then review the handoff

    Open with the decision, proposed answer, and critical caveat. Follow with comparable options, reasoning, and linked evidence. Ask a fresh reader to explain the tradeoff, uncertainty, and next action—not simply whether they agree. Name owners and dates only when supplied. Generate readable static HTML; keep approval replies in the established email or project workflow, and review audience access before sharing.

Use the full promptReady for your agent and your notes

Use the conversation where you already did the work. Only attach material you are allowed to share.

Get the ready-to-copy creation request →
Create a client decision brief for [audience] answering [specific decision], using my authorized materials. Establish who decides, what approval would authorize, and any supplied timing. Use the client's stated outcomes and hard constraints. Analyse credible alternatives, including continuing current arrangements when relevant, before writing the recommendation. Compare the same currency, time horizon, scope, and cost inclusions; show ongoing costs and client effort where known. Keep unknowns distinct from zero and preserve conflicting estimates with their sources. Recalculate material totals and check feasibility against hard constraints. Distinguish evidence, assumptions, professional judgment, and unknowns. Explain the main tradeoff, strongest counterargument, and what would change the recommendation. If evidence cannot support a choice, ask for the next evidence-gathering step. Do not invent approval, an owner, a deadline, or scoring weights. Lead with the requested decision, proposed answer, and critical caveat; follow with the comparison and dated supporting evidence. Make the next action explicit and keep approval replies in the existing email or project workflow. Build readable static HTML in my saved style if available, without scripts or forms. Check source accuracy, calculations, and whether a fresh reader can explain the choice and its limits. Show me the draft and flag confidential content for recipient-access review. Do not publish or change access without separate explicit approval. If this is a strategy choice brief, preserve the stated objective, alternatives, explicit non-goals, assumptions and reconsideration triggers. Do not infer strategy approval or claim growth outcomes from interview evidence. Keep the recommendation conditional on the stated priorities.
Check before you shareAccuracy, clarity, and permission
  • Is the requested choice within the client's authority, with selection, budget approval, and contract commitment kept distinct?
  • Recalculate material totals and check decisive facts against original inputs. Does the proposed option satisfy the hard constraints, and does the conclusion follow?
  • Are costs compared on matching terms, with exclusions, ongoing costs, conflicting estimates, and material unknowns visible?
  • Can a fresh reader explain the choice, tradeoff, uncertainty, change condition, and next action without agreeing with the recommendation?
  • Are proposed and approved work distinct, and do you have permission to share the findings, quotes, and linked evidence with every intended recipient?

When HTML fits

Use an HTML brief for an asynchronous decision that benefits from a short opening, a structured comparison, and a deeper evidence section. Readers can choose how much detail they need.

When another format fits better

Use a discussion or workshop when the decision or criteria are disputed. Respect a client's required Word or PDF format. Keep signatures and approval records in the established email, contract, or project workflow: reading or publishing this static page does not capture sign-off.

What we started withThe same facts, as an email

Fictional client email

Cedar Studio — enquiry workflow options for September 11

Hi Robin,

Here are the two options for reducing manual handling of inbound project enquiries, as of September 7, 2026. Your estimate is about 5 staff hours a week today, though that is not time-tracked and we have no conversion baseline.

Option A repairs the existing enquiry form: an estimated $2,000 one-time cost and 1 week after approval. It keeps the familiar workflow, but manual triage remains. Option B adds a guided enquiry flow and routing rules: an estimated $4,500 and 3 weeks after approval. It gives us better structured inputs, but requires staff training and validation of the routing rules.

Both fit the $5,000 budget ceiling before exclusions. Neither is a signed quote; both exclude tax and ongoing maintenance. I recommend B if you can assign a routing owner, because it addresses manual triage. We do not yet have measured savings or conversion lift.

Please choose A or B by September 11, 2026, and name the routing owner if you choose B. No option has been approved. Actual time saved, maintenance cost, tax, ownership and conversion impact remain unknown. After approval, we should confirm scope and exclusions, benchmark handling time, and validate routing with staff before launch.

Alex

Inspect the HTMLOne self-contained file

The example is sanitized and isolated from this site. No scripts, forms, or live data.

<!doctype html>
<html lang="en">
<head>
  <meta charset="utf-8">
  <meta name="viewport" content="width=device-width, initial-scale=1">
  <title>Cedar Studio · Enquiry workflow decision</title>
  <style>
    :root { color-scheme: light; }
    * { box-sizing: border-box; }
    body { margin: 0; background: #ecebe3; color: #24372e; font-family: -apple-system, BlinkMacSystemFont, "Segoe UI", sans-serif; font-size: 15px; line-height: 1.45; }
    main { width: 100%; max-width: 1080px; margin: 0 auto; padding: 24px 30px 20px; background: #faf8f1; }
    p, h1, h2, h3 { margin: 0; }
    .masthead { display: flex; align-items: baseline; justify-content: space-between; gap: 8px 20px; flex-wrap: wrap; padding-bottom: 15px; border-bottom: 1px solid #a7afa3; }
    .brand { font-family: Georgia, "Times New Roman", serif; font-size: 25px; font-weight: 700; letter-spacing: -.7px; }
    .meta { color: #526052; font-size: 14px; }
    .eyebrow { font-size: 12px; font-weight: 700; letter-spacing: 1.1px; text-transform: uppercase; }
    .decision { display: grid; grid-template-columns: 1.55fr 1fr; margin: 18px 0 20px; }
    .recommendation { padding: 21px 23px 22px; color: #faf8f1; background: #183f32; }
    .recommendation .eyebrow { color: #d0ddbc; }
    h1 { max-width: 17em; margin-top: 9px; font-family: Georgia, "Times New Roman", serif; font-size: 32px; font-weight: 400; line-height: 1.08; letter-spacing: -.6px; }
    .recommendation p:last-child { max-width: 37em; margin-top: 13px; color: #e7ebdf; }
    .request { padding: 21px 22px 20px; background: #e8eddf; border-left: 3px solid #bbca9c; }
    .request h2 { margin-top: 9px; font-size: 20px; font-weight: 600; line-height: 1.25; letter-spacing: -.35px; }
    .request .due { display: block; }
    .request p { margin-top: 12px; }
    .pending { font-size: 14px; font-weight: 600; color: #40533e; }
    .section-heading { margin-bottom: 10px; font-size: 19px; font-weight: 650; letter-spacing: -.35px; }
    .comparison-intro { display: flex; align-items: baseline; justify-content: space-between; gap: 4px 15px; flex-wrap: wrap; }
    .comparison-intro .section-heading { margin-bottom: 3px; }
    .ceiling { font-size: 14px; color: #40533e; }
    .ceiling strong { font-size: 17px; }
    .exclusions { margin: 6px 0 13px; color: #526052; font-size: 14px; }
    table { width: 100%; border-collapse: collapse; table-layout: fixed; }
    th, td { width: 50%; padding: 10px 18px; text-align: left; vertical-align: top; }
    th { padding-top: 14px; padding-bottom: 14px; border-top: 2px solid #294d3e; border-bottom: 1px solid #a7afa3; }
    th + th, td + td { border-left: 1px solid #bdc6b5; background: #eff2e7; }
    th .eyebrow { display: block; color: #526052; }
    th h3 { margin-top: 5px; font-family: Georgia, "Times New Roman", serif; font-size: 23px; font-weight: 400; line-height: 1.14; letter-spacing: -.25px; }
    td { border-bottom: 1px solid #d3d8cb; }
    td .field { display: block; margin-bottom: 3px; color: #526052; font-size: 12px; }
    .amount { display: block; font-size: 30px; font-weight: 550; line-height: 1.12; letter-spacing: -1px; font-variant-numeric: tabular-nums; }
    .timing { font-weight: 650; }
    .evidence { display: grid; grid-template-columns: 1fr 1fr; gap: 25px; margin-top: 19px; padding-bottom: 17px; border-bottom: 1px solid #a7afa3; }
    .evidence h2 { margin-bottom: 6px; font-size: 15px; font-weight: 700; }
    .evidence p { font-size: 14px; }
    .baseline { color: #183f32; }
    .next { padding-top: 14px; }
    .next h2 { font-size: 15px; font-weight: 700; }
    ol { display: grid; grid-template-columns: 1fr 1fr 1.25fr; gap: 13px 26px; margin: 9px 0 0; padding-left: 20px; font-size: 14px; }
    li { padding-left: 2px; }
    li::marker { color: #405b34; font-weight: 700; }
    footer { margin-top: 17px; color: #526052; font-size: 12px; }
    @media (max-width: 580px) {
      body { font-size: 14px; }
      main { padding: 19px 18px 18px; }
      .masthead { display: block; padding-bottom: 12px; }
      .brand { font-size: 27px; }
      .meta { margin-top: 3px; }
      .eyebrow { font-size: 14px; letter-spacing: .65px; }
      .decision { grid-template-columns: 1fr; margin-top: 15px; margin-bottom: 20px; }
      .recommendation { padding: 20px; }
      h1 { margin-top: 10px; font-size: 31px; line-height: 1.08; }
      .recommendation p:last-child { margin-top: 12px; }
      .request { padding: 16px 20px 17px; border-top: 3px solid #bbca9c; border-left: 0; }
      .request h2 { margin-top: 7px; font-size: 20px; }
      .request p { margin-top: 9px; }
      .section-heading { font-size: 21px; }
      .comparison-intro { display: block; }
      .ceiling { margin-top: 5px; }
      .exclusions { margin-top: 7px; }
      th, td { padding: 11px 11px; }
      th { padding-top: 12px; padding-bottom: 12px; }
      th h3 { font-size: 20px; }
      td .field { font-size: 14px; line-height: 1.25; }
      .amount { font-size: 27px; }
      .evidence { grid-template-columns: 1fr; gap: 15px; margin-top: 18px; padding-bottom: 18px; }
      .evidence h2, .next h2 { font-size: 16px; }
      ol { display: block; margin-top: 8px; }
      li + li { margin-top: 7px; }
      footer { margin-top: 16px; font-size: 14px; }
    }
    @media print {
      @page { size: A4; margin: 12mm; }
      body { background: #fff; font-size: 10pt; }
      main { max-width: none; padding: 0; background: #fff; }
      .decision { break-inside: avoid; }
      .recommendation { color: #183f32; background: #fff; border: 1px solid #a7afa3; }
      .recommendation .eyebrow, .recommendation p:last-child { color: #24372e; }
      .request, th + th, td + td { background: #f0f2eb; }
      h1 { font-size: 24pt; }
      th, td { padding: 7px 13px; }
      tr, .evidence, .next { break-inside: avoid; }
      .exclusions, .evidence p, ol, .meta, .ceiling, .pending { font-size: 9pt; }
      footer { font-size: 8pt; }
    }
  </style>
</head>
<body>
  <main>
    <header class="masthead">
      <p class="brand">Cedar Studio</p>
      <p class="meta">Prepared by Alex · September 7, 2026</p>
    </header>

    <section class="decision" aria-label="Recommendation and requested decision">
      <div class="recommendation">
        <p class="eyebrow">Conditional recommendation · Option B</p>
        <h1>Choose the guided flow—if someone owns routing.</h1>
        <p>To reduce manual handling of inbound project enquiries, B addresses manual triage and fits the stated budget before exclusions.</p>
      </div>
      <aside class="request">
        <span class="eyebrow">Decision pending</span>
        <h2>Robin, project sponsor <span class="due">By September 11, 2026</span></h2>
        <p>Choose A or B.<br>If B, name the routing owner.</p>
        <p class="pending">No option has been approved.</p>
      </aside>
    </section>

    <section aria-labelledby="options-heading">
      <div class="comparison-intro">
        <h2 class="section-heading" id="options-heading">Two paths, different tradeoffs</h2>
        <p class="ceiling">Budget ceiling <strong>$5,000</strong></p>
      </div>
      <p class="exclusions">Both one-time estimates fit before tax and ongoing maintenance, which are excluded. Neither is a signed quote.</p>
      <table aria-label="Compare the enquiry workflow options">
        <thead>
          <tr>
            <th scope="col"><span class="eyebrow">Option A</span><h3>Repair the existing enquiry form</h3></th>
            <th scope="col"><span class="eyebrow">Option B</span><h3>Guided enquiry flow + routing rules</h3></th>
          </tr>
        </thead>
        <tbody>
          <tr>
            <td><span class="field">Estimated one-time cost</span><strong class="amount">$2,000</strong></td>
            <td><span class="field">Estimated one-time cost</span><strong class="amount">$4,500</strong></td>
          </tr>
          <tr>
            <td><span class="field">Estimated delivery</span><span class="timing">1 week</span> after approval</td>
            <td><span class="field">Estimated delivery</span><span class="timing">3 weeks</span> after approval</td>
          </tr>
          <tr>
            <td><span class="field">Benefit</span>Familiar workflow</td>
            <td><span class="field">Benefit</span>Better structured inputs</td>
          </tr>
          <tr>
            <td><span class="field">Client effort / tradeoff</span>Manual triage remains.</td>
            <td><span class="field">Client effort / tradeoff</span>Staff training and validation of routing rules required.</td>
          </tr>
        </tbody>
      </table>
    </section>

    <section class="evidence" aria-label="Evidence and uncertainty">
      <div>
        <h2>Starting point, not a savings claim</h2>
        <p><strong class="baseline">About 5 staff hours per week</strong> handling enquiries: Robin’s estimate, not time-tracked. No conversion baseline or proven savings or conversion lift.</p>
      </div>
      <div>
        <h2>Still unknown</h2>
        <p>Actual time saved, maintenance cost, tax, routing owner and conversion impact. Without a routing owner, revisit the recommendation.</p>
      </div>
    </section>

    <section class="next" aria-labelledby="next-heading">
      <h2 id="next-heading">After approval</h2>
      <ol>
        <li>Confirm scope and exclusions.</li>
        <li>Benchmark handling time.</li>
        <li>Validate routing with staff before launch.</li>
      </ol>
    </section>
    <footer>Fictional worked example · Source: the accompanying client notes.</footer>
  </main>
</body>
</html>
Questions, adaptations & sourcesGo deeper when you need to

Does every decision brief need a recommendation?

Usually a clear recommendation helps, but insufficient evidence is a legitimate outcome. Recommend a specific evidence-gathering step when the information cannot support a choice. Explain what it should establish and how it affects the decision.

Should the agent score the options?

Only use an aggregate score with defined scales, a defensible weighting rationale, and the client's preference input. Check whether reasonable changes in assumptions or weights reverse the result. Never let a high total hide a failed hard constraint. A qualitative comparison is often more honest than invented numerical precision. These are adapted appraisal practices, not proof of one universally best brief format.

How long should a client decision brief be?

Keep the opening short enough to identify the decision, proposed answer, and critical caveat. Put the option comparison next and detailed evidence below it. Do not shrink type or hide important uncertainty to hit a page count; scale the supporting analysis to the stakes and the client's required format.

Can a brief replace the final client presentation?

It can replace some read-ahead or recap material. Use a presentation when a live conversation is part of the work. The brief can then preserve the argument and evidence after the meeting without requiring readers to reconstruct your spoken explanation.

A strategy choice brief

When the job is choosing a direction, use the existing decision structure: evidence, alternatives, tradeoffs, assumptions, non-goals and what would trigger reconsideration. A proposed focus is not an approved strategy or a promise of growth.

Fictional inputs: Fictional team has capacity for one pilot. Its stated priority is learning from existing teams. Supplied interviews identify onboarding friction in that group; there is no evidence of willingness to pay in a new segment. Leadership has not approved a focus.

A clearer result: Propose an existing-team onboarding pilot because it fits the stated learning priority. Name expansion into a new segment as outside this pilot, not permanently ruled out. Reconsider if the priority changes or credible new-segment evidence arrives. Approval remains pending.

Further reading

By share/artifacts · Published · Updated