turn-client-research-into-recommendations

name	turn-client-research-into-recommendations
description	Turn client research, interviews, data, and engagement context into evidence-backed options, a defensible recommendation, and a decision-ready memo, deck, workshop artifact, or supporting spreadsheet. Use when a consultant or internal advisory team needs to move from findings to what the client should do.

Turn Client Research into Recommendations

Help the client make a decision from the available evidence. This skill owns the move from findings to credible options and a recommendation; it does not replace the research skills that collect or structure the underlying evidence. Follow an existing consulting method when the user has one.

1. Frame the decision

Identify the client, decision, intended audience, decision owner, stakes, time horizon, constraints, criteria, deadline, and useful output. Clarify what is already known, what would change the decision, and whether the user wants a recommendation, help choosing between options, or a neutral set of options.

If the request is only to collect, map, or compare information, route to the appropriate research skill instead of forcing the work into a recommendation.

2. Choose the right evidence path

Start with the accepted engagement brief, prior research, meeting records, client data, source links, files, visible tabs, connected apps, and relevant client-specific memory. Keep every client's sources, calculations, artifacts, permissions, and learned context separate.

Use existing research skills for the work they already own:

• strawberry/research-analysis/map-a-market for a landscape, taxonomy, or relationship map;
• strawberry/research-analysis/extract-web-data for repeated fields across many rows, including large company-research runs;
• strawberry/research-analysis/research-a-company for a sourced brief on one company;
• strawberry/research-analysis/create-beautiful-slide-deck when accepted research should become a visual presentation;
• strawberry/sales/research-an-account when the company matters as a commercial relationship.

Strawberry can combine this work with material in visible logged-in browser sessions and connected sources while keeping the supporting evidence inspectable. Propose the sources, depth, likely time or cost, and a useful first review point before filling gaps. Do not start a broad research run when the existing evidence already supports the decision.

3. Build a trustworthy evidence base

Organize the material around the decision rather than the order in which it was found. Keep observed facts, client statements, calculations, and consultant inference distinct. Preserve source links and relevant dates, reconcile definitions, and flag stale evidence, contradictions, missing data, and important limitations.

When the classification, research depth, or output structure is subjective, review a representative sample with the user before scaling. Ask only for gaps whose answer could change the options or recommendation.

4. Move from findings to a recommendation

Explain the decision-relevant patterns and implications. Compare credible options against the accepted criteria, then recommend a path when the evidence supports one. Show the rationale, tradeoffs, risks, reversibility, dependencies, and remaining uncertainty. Include the strongest alternative and explain what evidence or event would make it preferable.

Do not turn a client claim into an established fact, hide a logical leap, invent a calculation, or present confidence the evidence does not support. When evidence is thin, provide a partial recommendation, bounded options, or the highest-value next research step rather than manufacturing certainty.

5. Create the decision-ready artifact

Propose the structure before producing the polished version. A useful default includes an executive recommendation, the decision and criteria, key evidence, credible options, rationale, risks and assumptions, next steps, and traceable sources.

Choose the simplest format that fits the decision:

• a memo for a detailed written recommendation;
• a deck for a client readout, using strawberry/research-analysis/create-beautiful-slide-deck for production;
• a workshop artifact when stakeholders need to evaluate options together;
• a supporting spreadsheet when comparisons or calculations need transparent inputs, formulas, assumptions, and source notes, using the internal strawberry/general/spreadsheets capability.

Use a dataset or CSV as the evidence layer for a large research run, not as a substitute for the recommendation the client needs. Do not allow visual polish to obscure weak evidence or unsupported calculations.

6. Review before sharing or action

Test each material claim against its source and check the path from evidence to conclusion. Look for counterevidence, inconsistent definitions, feasibility constraints, client-voice problems, and facts that may have changed. Surface the remaining decisions and recommend answers where possible.

Establish the audience, access model, and destination before recommending a sharing path. Confirm that client confidentiality, personal data, source permissions, and visual examples are appropriate for that audience. Internal review, client delivery, public sharing, sending, publishing, CRM changes, and implementation are different approval states; do not treat approval of the analysis as approval for any of them.

7. Preserve the accepted method

After the approach works, offer to save the accepted questions, sources, evidence standards, decision criteria, artifact structure, voice, and review behavior as a custom skill. For boutique teams, preserve only the reusable method unless the team has explicitly agreed access to the client context and destination.

A draft-first Routine can refresh a trusted recommendation when the source set, cadence, output, and review owner are stable. Define its permissions, notification behavior, and approval boundary. Stop when a material source is unavailable, sources conflict, access changes, calculations cannot be reproduced, or the underlying client decision has changed.

Suggested outcome

Deliver a decision-ready recommendation with inspectable evidence, credible alternatives, explicit risks and assumptions, and a polished memo, deck, workshop artifact, or supporting spreadsheet that fits the audience and decision.