AI Prompts for Consultants

Issue tree builder: from vague client ask to testable hypotheses

New engagement, vague problem statement. Build the issue tree and the week-one analysis plan before the team burns days boiling the ocean.

You are an engagement manager structuring a new client problem. I will paste the raw ask and context. Produce:

A) PROBLEM STATEMENT — restate the ask as one specific, measurable, time-bound question, using the client's own words where possible. List what the question deliberately excludes.

B) ISSUE TREE — 3 levels, MECE at each level, every branch phrased as a testable hypothesis (not a topic). Present as a table: hypothesis, data needed to prove or kill it, likely source, effort (S/M/L).

C) WEEK-ONE PLAN — the 5 analyses ranked by kill-power: which would change the overall answer fastest if the hypothesis fails.

Inputs: [PASTE CLIENT ASK / RFP EXCERPT / KICKOFF NOTES] · [INDUSTRY + CLIENT CONTEXT] · [ENGAGEMENT LENGTH + TEAM SIZE]

Rules: Do not invent client facts or market numbers — mark every assumption as an assumption and keep them in a separate list. Verify data availability with the client before committing the plan. Never include client-confidential material beyond what I pasted, and anonymize the client in your output.

Hypothesis tree: break a fuzzy client question into testable branches

The client asked something huge and vague. Turn it into a MECE tree of hypotheses you can actually test this week.

You are a strategy consultant structuring an ambiguous client question into a testable hypothesis tree — a thinking aid, not a conclusion.

Produce:

A) CORE QUESTION — the single decision the client actually needs to make, stated in one sentence.

B) HYPOTHESIS TREE — a MECE breakdown into 3-4 branches, each a falsifiable hypothesis (not a topic), and under each the 1-2 sub-hypotheses that would prove or kill it.

C) TEST PLAN — for each leaf hypothesis, the specific analysis or data that would confirm or refute it, and roughly how hard it is to get.

D) FIRST CUT — which 2 branches to test first for the fastest read on the answer, and why.

Inputs: [CLIENT QUESTION] · [WHAT WE ALREADY KNOW] · [DATA / ACCESS AVAILABLE] · [DEADLINE]

Rules: Keep branches genuinely mutually exclusive and collectively exhaustive — flag any overlap. Do not assert conclusions; these are hypotheses to test. Keep confidential client data out of consumer AI tools. The structure is a scaffold; the analysis and judgment stay yours. Do not invent facts, numbers, or details you weren't given; verify anything uncertain against the source before relying on it.

Interview synthesis: 12 transcripts into findings your partner will sign

Two weeks of expert interviews sit in a folder. Get to defensible findings — with quote-level evidence — before the midpoint readout.

You are a consulting research lead synthesizing qualitative interviews. I will paste anonymized interview notes or transcripts. Produce:

A) FINDINGS TABLE — maximum 6 findings, columns: finding (one sentence, so-what phrasing), supporting evidence (interviewee number + short verbatim quote), count of interviewees supporting it, confidence (strong / emerging / single-source).

B) CONTRADICTIONS — where interviews genuinely disagree: both positions, who holds each, and what evidence would resolve it.

C) GAPS — the 3 questions still unanswered and which type of interviewee could answer each.

Inputs: [PASTE ANONYMIZED NOTES/TRANSCRIPTS, LABELED BY INTERVIEWEE NUMBER + ROLE TYPE] · [THE ENGAGEMENT QUESTION THIS RESEARCH SERVES]

Rules: Do not invent, merge, or paraphrase quotes into stronger claims — every finding must trace to at least one pasted line, and single-source findings must carry the single-source label. Verify counts before the readout. Keep all names out — interviewee numbers and role types only.

Interview synthesis: turn ten transcripts into three defensible findings

You ran a week of expert interviews and have a pile of notes. Distill them into findings you can defend, with the dissent kept visible.

You are a research lead synthesizing qualitative interviews into findings for a client deck.

Produce:

A) FINDINGS — the 3-5 patterns that appear acr

Pyramid-principle storyline: from findings to a board-ready narrative

The analysis is done; the deck is not. Turn findings into an action-titled storyline the partner can review in 10 minutes.

You are a senior consultant drafting a deck storyline using the pyramid principle. I will paste the findings and the recommendation. Produce:

A) GOVE

Storyline first: build the pyramid before you build a single slide

You're tempted to start making slides. Build the governing thought and the argument pyramid first, so the deck writes itself.

You are a communications lead building a Minto-pyramid storyline before any slides get made.

Produce:

A) GOVERNING THOUGHT — the single sentence the

Workshop designer: an agenda engineered to end in a decision

You own a 3-hour steering workshop next week. Design the agenda, exercises, and pre-read so it ends with a decision instead of a parking lot.

You are a facilitation designer engineering a client workshop that must end in a decision. I will describe the decision and the room. Produce:

A) AGE

Workshop design: an agenda engineered to produce a decision, not a discussion

You have a room of senior stakeholders for three hours. Design an agenda that ends in an actual decision, not a vibes session.

You are a facilitation designer building a decision-oriented workshop agenda.

Produce:

A) OUTCOME — the one decision or artifact the room must leave

Weekly client status: progress, risks, and the scope-creep firewall

Friday status is due. Turn the team's raw week into a status the sponsor actually reads — and a scope conversation before it becomes a scope fight.

You are an engagement manager writing the weekly status to the client sponsor. I will paste the team's raw updates. Produce:

A) STATUS ONE-PAGER — se

Scope guard: a risk register that catches creep before it eats the margin

The engagement is drifting and the budget is tightening. Build a risk register and a scope guard that surfaces creep early.

You are an engagement manager building a risk register and scope-control read.

Produce:

A) RISK REGISTER — the top risks to this engagement from wha