Best AI Prompts for Consultants

AI summary

Seven AI prompts for management, strategy, and operations consultants: discovery-to-SOW translation, stakeholder interview synthesis, slide deck audits, client update drafting, proposal pricing pressure tests, implementation risk audits, and engagement post-mortems. Built to scaffold consulting judgment without imposing a template-flavored output.

Prompt 1

Discovery-to-SOW Translator

You had a discovery call. The client wants a proposal. Most consultants jump straight to scoping without translating what the client said into what they actually need. This prompt does the translation.

Discovery call summary:

CLIENT (briefly, role and company stage): [BRIEF]
WHAT THEY SAID THEY WANTED: [LIST]
WHAT THEY SAID WAS HARD: [LIST]
WHAT I HEARD AS THE REAL PROBLEM (different from what they named): [YOUR READ]
MY INITIAL HYPOTHESIS about the engagement shape: [BRIEF]

Translate into a scope:

1. THE STATED PROBLEM restated cleanly.
2. THE ACTUAL PROBLEM as I heard it, with the evidence from the call.
3. THE ENGAGEMENT SHAPE: discovery / diagnostic / implementation / change management / advisory, with the case for each.
4. THE 3-PHASE OUTLINE: what would happen in phases 1, 2, 3, with rough duration each.
5. THE DELIVERABLES: specific outputs they would have at the end.
6. THE SUCCESS CRITERIA: how we would know in 90 days if this worked.
7. THE QUESTIONS I should ask in our follow-up call before drafting the SOW.

Do not jump to a fixed-fee scope. Do not assume their stated problem is the right problem. Surface the gap between what they asked for and what they need.

When to use: Within 24 hours of the discovery call. · Best model: Claude (most disciplined about not jumping to scope).

Prompt 2

Stakeholder-Interview Synthesis

You did 8 stakeholder interviews. Your notes are a mess. The presentation is in 4 days. This prompt does the synthesis without bias, not the version that confirms your hypothesis.

Here are notes from stakeholder interviews (de-identified):

[FOR EACH: role, key quotes, notable affect, what surprised you]

My starting hypothesis going into the interviews: [ONE SENTENCE]
The decision this work informs: [SPECIFIC]

Synthesize:

1. WHAT THE INTERVIEWS CONFIRMED: hypotheses the data supports, with the specific evidence.
2. WHAT THE INTERVIEWS CHALLENGED: where the data pushed back on my starting assumptions.
3. THE FRACTURE LINES: where interviewees disagreed with each other. This is the most important pattern.
4. THE QUIET PATTERN: a recurring theme across multiple interviews that I might be under-weighting.
5. THE LOUDEST VOICE: any single interviewee whose feedback I might be over-weighting.
6. THE 3 QUOTES that should anchor the deliverable, with the speaker role.
7. THE QUESTION the synthesis raises that I need to bring back to the client.

Do not let me confirm my hypothesis if the data does not support it. Surface the disagreements rather than smoothing them.

When to use: Within a week of finishing the interviews. · Best model: Claude. The discipline about not smoothing disagreement matters.

Prompt 3

Slide Deck Audit

Most consulting decks are too long, bury the recommendation, and have too many ideas per slide. This prompt audits your deck against those failure modes.

Here is the deck I am about to send to the client:

[PASTE DECK CONTENT, SLIDE BY SLIDE]

The one recommendation I want them to walk away with: [ONE SENTENCE]
The audience: [WHO IS READING THIS]
The decision it informs: [SPECIFIC]

Audit the deck:

1. THE RECOMMENDATION TEST: read the deck back to me and state the recommendation it actually delivers. If it is not the recommendation I stated, the deck is broken.
2. THE SLIDE-PER-IDEA RATIO: any slide that has more than one idea on it. Flag.
3. THE BURIED LEDE: if the recommendation appears on slide 23, it should be on slide 2. Where would I move it.
4. THE WEAKEST SLIDE: the slide a partner would call "so what?" Why.
5. THE CUT CANDIDATES: 3-4 slides that, removed, would strengthen the deck.
6. THE MISSING ANSWER: the obvious client question the deck does not address.
7. THE EMOTIONAL READ: what the deck makes the audience feel, and whether that matches the situation.

Do not rewrite slides. Surface the issues. I will rewrite.

When to use: Two business days before the client presentation. · Best model: Claude. The discipline about not rewriting matters.

Prompt 4

Client Update Drafter (Engagement)

Weekly client updates either ramble or get auto-templated into nothing. This prompt drafts the update so the client opens it.

End of week on the engagement:

WHAT WE COMMITTED TO THIS WEEK: [BULLETS]
WHAT WE ACTUALLY DELIVERED: [BULLETS]
WHERE WE FOUND SOMETHING UNEXPECTED: [BULLETS]
WHERE WE ARE BEHIND: [BULLETS, BE DIRECT]
WHAT WE NEED FROM THEM NEXT WEEK: [LIST]
NEXT WEEK'S COMMITMENTS: [LIST]

Draft a client update:

1. THE HEADLINE: one sentence on the most important thing that happened this week.
2. WHAT WE DELIVERED: bullets tied to the commitment, not a status update.
3. WHAT WE FOUND: 1-2 sentences on insights worth surfacing.
4. WHERE WE ARE BEHIND: direct, with the recovery plan.
5. WHAT WE NEED: specific asks, with the person and the timing.
6. NEXT WEEK: the commitments, in priority order.
7. THE QUESTION FOR YOU: one question that, if answered, would meaningfully shape next week's work.

Keep under 350 words. Tone: candid, calm, never spinning misses into progress.

When to use: Friday afternoon or end of business of your engagement’s weekly cadence. · Best model: Claude. Tone discipline matters; consulting reputations live on the weekly update.

Prompt 5

Proposal Pricing Pressure Test

You are about to send a fixed-fee proposal. Your gut says the number is right. This prompt structures the pressure test before the proposal goes out.

Proposal I am about to send:

SCOPE: [SUMMARY]
DURATION: [WEEKS]
FEE: [NUMBER]
MY UTILIZATION ASSUMPTIONS: [DAYS PER WEEK FOR EACH ROLE ON THE TEAM]
KNOWN RISK FACTORS: [SCOPE CREEP, CLIENT RESPONSIVENESS, DATA ACCESS, ETC.]
MY EFFECTIVE HOURLY EQUIVALENT at this fee: [DOLLARS / HOUR]
WHAT THE CLIENT IS EXPECTING based on the conversation: [RANGE OR ANCHOR]

Pressure test:

1. UTILIZATION REALITY: at the fee divided by the duration, is the assumed utilization realistic given the scope.
2. THE SCOPE-CREEP TRAP: what specifically is most likely to expand scope. Build in or call out explicitly.
3. THE WALK-AWAY: if the client comes back with a 20 percent lower number, what would I cut.
4. THE ANCHORING: how does this number land relative to the client's expectation. Adjust framing if needed.
5. THE TRUST-BUILD: is this engagement worth doing for less to win the relationship, or is the relationship not worth the discount.
6. THE OPPORTUNITY COST: what else I am NOT doing during this engagement window.
7. THE QUESTION to surface in the proposal call before locking the fee.

Do not let me anchor to my own hopes. Stress-test the assumptions.

When to use: Before sending the proposal. · Best model: Claude. The discipline about pushing back on optimistic utilization matters.

Prompt 6

Implementation Risk Audit

Most consulting recommendations get half-implemented and then quietly abandoned. This prompt audits the implementation risk before the recommendation lands so the client has a real shot.

The recommendation I am delivering:

WHAT WE ARE RECOMMENDING: [ONE PARAGRAPH]
WHO HAS TO DO WHAT: [ROLES]
WHAT HAS TO CHANGE in the client's operating model: [SPECIFIC]
WHAT THEY ARE CURRENTLY DOING that needs to stop: [SPECIFIC]
KEY PLAYERS' STATED SUPPORT LEVELS: [WHO IS IN, OUT, AMBIVALENT]
HISTORICAL BASELINE: prior change initiatives this organization has run and how they went: [BRIEF]

Audit the implementation risk:

1. THE SHOW-STOPPER: the one thing that, if it goes wrong, kills the recommendation. Name it and propose mitigation.
2. THE AMBIVALENT STAKEHOLDER: the person whose mid-implementation withdrawal would tank momentum. Plan for them.
3. THE CHANGE-FATIGUE READ: based on the historical baseline, is this organization ready for another change initiative or are they exhausted.
4. THE 90-DAY MILESTONE: what should be true at day 90 that would signal traction. Name it.
5. THE 90-DAY CANARY: what would signal the change is failing. Name it explicitly so they can pivot rather than slowly fade.
6. THE PRECONDITIONS we should require BEFORE accepting the engagement to deliver this recommendation.
7. THE QUESTIONS THE CLIENT SHOULD ASK before signing off.

Do not over-promise on what implementation will deliver. Surface the risk; the client can decide.

When to use: Before the final recommendation deck is sent. · Best model: Claude. The discipline about surfacing risk matters for engagement quality.

Prompt 7

Engagement Post-Mortem

Engagements end and the team moves on. The lessons go with them. This prompt structures the post-mortem so they survive.

Engagement just ended:

CLIENT (de-identified): [BRIEF]
ORIGINAL SCOPE: [BRIEF]
WHAT WE DELIVERED: [BRIEF]
WHAT THE CLIENT SAID at engagement close: [BRIEF]
INTERNAL TEAM READ on the engagement: [BRIEF, INCLUDING FRICTIONS]
FINANCIAL OUTCOME: [PROFITABLE / BREAKEVEN / LOSS, AT WHAT MARGIN]

Draft a post-mortem:

1. THE OUTCOME: did we deliver what we sold. Specific, not narrative.
2. THE PROCESS: where did the engagement run smoothly, where did it strain.
3. THE CLIENT SIDE: what about the client made the work easier or harder.
4. THE INTERNAL TEAM: what about how we worked made the engagement easier or harder.
5. THE LESSON: 2-3 specific things to do differently on the next engagement of this type.
6. THE LESSON TO INSTITUTIONALIZE: anything worth turning into a firm-level practice change.
7. THE RELATIONSHIP STATUS: is this a client we want to do more work with, given the post-mortem read.

Do not let me protect my ego. Engagements that look successful on the outside often have real lessons inside.

When to use: Within 2 weeks of engagement close. · Best model: Claude or Grok. Both willingly call out team-side issues.

Connectors are now standard

Claude, ChatGPT, and Grok all support connectors that let your AI read live data from your work tools (Gmail, Notion, GitHub, Asana, HubSpot, Stripe, and many more) instead of relying on you to paste context. For consultants this means the AI can read your Notion or Confluence engagement workspace, your Gmail thread with the client, your Calendly bookings, your interview transcripts in Descript or Otter, or your client-facing deck repository in Google Drive.

For consultants, the connectors worth pairing with these prompts:

Get Smarter About AI Every Morning

Free daily newsletter. Built for people who want to use AI well, not chase every model.

Free forever. Unsubscribe anytime.

Two ways to go further