Claude for Consulting: Frameworks, Slides, and Deliverables That Ship
20 Claude prompts for case analysis, MECE frameworks, pyramid principle storylines, client deliverables, and the consulting work where Claude's reasoning pays off.
In short: This page contains 20 copy-paste ready prompts, organized into 4 categories with a description and pro tip for each. The first 5 prompts are free instantly, no signup needed. Drafted with AI writing assistance and checked against our editorial policy.
Case Analysis
5 promptsMECE Problem Decomposition
1/20✨ What it does
Claude breaks the [paste problem statement] into a MECE tree with assumptions, how to validate each branch, and what to prioritize first. Add [industry, company, situation], then start with the highest-leverage branch on your list.
<task>Break down this business problem using MECE principles</task> <problem>[paste problem statement]</problem> <context>[industry, company, situation]</context> <output> 1. Restate the core question precisely 2. MECE tree decomposition (3-5 levels deep) 3. Per branch: assumption being made, how to validate 4. Which branches are highest leverage for impact 5. Which branches can be deprioritized 6. Synthesized "so what" hypothesis per major branch </output>
Pro tip: MECE trees force clarity. Branches that overlap or leave gaps reveal sloppy thinking. Claude is excellent at this kind of structured decomposition: feed it a problem and insist on MECE rigor.
Hypothesis Tree + Validation Plan
2/20✨ What it does
Claude builds a hypothesis tree for the problem you [describe], with data that would confirm or refute each bet, sources, and effort. Sequence the cheap tests first, then drop hypotheses your data kills.
<task>Build a hypothesis tree and validation plan</task> <problem>[describe]</problem> <output> 1. Root hypothesis (the big bet) 2. Supporting sub-hypotheses (3-5) 3. Per sub-hypothesis: what data/evidence would confirm or refute 4. Data sources (internal data, external research, expert interviews, surveys) 5. Effort estimate per data-gathering activity 6. Prioritized investigation sequence 7. Quick-wins vs deep-dives </output>
Pro tip: Good consulting isn't answering questions: it's asking the right ones. A hypothesis tree forces you to say "here's what I think is true, and here's how I'd prove it wrong": which is the only way to test strategy rigorously.
Quick Case Crack
3/20✨ What it does
Claude cracks the [paste case description] in a 30-minute structure: restatement, clarifying questions, MECE branches, data needs, and a recommendation. Ask the clarifying questions first, then walk your interviewer through the structure.
<task>Crack this case in 30 minutes</task> <case>[paste case description]</case> <output> 1. 60-second restatement of the problem 2. Clarifying questions you'd ask (3-5) 3. Structure (MECE) for the analysis 4. Per branch: 2-3 data points needed 5. Likely conclusion given typical case answers 6. Recommendation with 2nd-order implications </output>
Pro tip: Case interviews teach structured thinking that generalizes to real consulting. Using Claude as a practice partner: feed it cases, get structure feedback: dramatically accelerates case prep.
Market Sizing (Top-Down + Bottom-Up)
4/20✨ What it does
Claude sizes [market] with both top-down and bottom-up math, then triangulates a range you can defend. Check the unit and price assumptions, then keep the range you will put on a slide.
<task>Size [market] using both top-down and bottom-up methodology</task> <market>[describe]</market> <output> Top-down: - Global/national total relevant spending - % of that going to this category - Growth rate - Our reachable portion Bottom-up: - Unit count × price × frequency - Per-segment calculation - Sum + validation Triangulation: - Compare results - Explain discrepancies - Final estimate with confidence range - Key assumptions that could shift estimate </output>
Pro tip: Market sizing that uses only one methodology is fragile. Triangulation (top-down + bottom-up) reveals where assumptions are weak. If the two methods disagree by 10×, you have an assumption problem to solve before using either.
Root Cause Analysis (5 Whys + Fishbone)
5/20✨ What it does
Claude runs 5 Whys plus a Fishbone on the problem you [describe], then names the top root causes and where to apply leverage. Choose one cause you can actually fix, then brief the team on that lever.
<task>Apply both 5 Whys and Fishbone to this problem</task> <problem>[describe]</problem> <output> 5 Whys: - Sequential cascade of why questions - Identify the root cause(s) Fishbone (6M categories): - Man (people) - Method (process) - Machine (tools/tech) - Material (inputs) - Measurement (metrics) - Milieu (environment) Synthesis: - Top 3 root causes - Interdependencies - Highest leverage interventions </output>
Pro tip: 5 Whys gives depth; Fishbone gives breadth. Combining both catches what either misses alone. Claude's structured thinking makes this traditionally tedious exercise fast and rigorous.
Deliverables & Communication
5 promptsPyramid Principle Storyline
6/20✨ What it does
Claude builds a pyramid storyline from [the answer we want to deliver] with three MECE arguments, evidence, and anticipated objections. Drop in [paste key findings], then cut any argument your evidence does not support.
<task>Build a pyramid principle storyline</task> <conclusion>[the answer we want to deliver]</conclusion> <supporting_data>[paste key findings]</supporting_data> <output> 1. Governing thought (1 sentence: the "so what") 2. 3 supporting arguments (each MECE) 3. Per argument: 2-3 evidence points 4. Anticipated objections + responses 5. Storyline as a sequence of slides (not a wall of text) 6. Appendix content (what to move off the main deck) </output>
Pro tip: The Pyramid Principle is the hidden OS of consulting communication. Once you internalize it, you can't unsee its absence in decks. Claude helps structure unstructured findings into Pyramid-compliant arguments.
Executive Summary Writer
7/20✨ What it does
Claude writes a one-page executive summary of the work you [paste or describe], opening with the so-what, plus evidence, decisions, and risks. Name who will read it, then trim the page until your reader can finish it standing up.
<task>Write an executive summary for this work product</task> <full_content>[paste or describe]</full_content> <audience>[who will read: CEO, board, ops leader]</audience> <output> - 1-page max - Opening: the core "so what" in 2 sentences - 3 supporting points with evidence - Top 2 recommendations with decision required - Risks / open questions - Suggested next step Tone: plain business English. No jargon. No hedging.
Pro tip: Execs read the summary, skim the deck, and forward the conclusion. A poorly written exec summary buries your work. Claude is strong at compressing findings into the top-line story without losing nuance.
Slide-by-Slide Outline
8/20✨ What it does
Claude outlines a slide-by-slide deck for [deck purpose] with action titles, visual types, and narrative flow for [who]. Set the [target] slide count, then drop any slide that does not earn its place in your story.
<task>Build a slide-by-slide outline for [deck purpose]</task> <audience>[who]</audience> <key_findings>[describe]</key_findings> <slide_count>[target]</slide_count> <output> Per slide: 1. Headline (action title: tells the point) 2. Supporting visual type (chart, matrix, framework, image) 3. Body content outline 4. Connection to previous slide (narrative flow) 5. Speaker notes Plus: overall storyline arc and pacing review.
Pro tip: Slide titles should BE the point, not label it. "Revenue Analysis" is a label; "Revenue fell 23% because of churn in SMB segment" is the point. Action-title every slide and the deck reads like an argument, not a report.
Client Presentation Script
9/20✨ What it does
Claude writes a spoken script for [deck] with per-slide lines, transitions, and answers to likely questions, paced to [minutes]. Rehearse the transitions aloud, then leave the script beside you, not on the slide.
<task>Write a presentation script for [deck]</task> <deck_outline>[paste or describe]</deck_outline> <time>[minutes]</time> <audience>[describe]</audience> <output> - Per slide: what to say (conversational, not read-aloud of slide) - Transitions between slides - Questions to anticipate (3-5) and prepared answers - Where to pause for client input - Opening hook + closing call-to-action </output>
Pro tip: Reading slides to clients is the fastest way to lose them. Good presenters SPEAK to the point of the slide, not read it. Claude can draft speaker notes that sound like a human thinking, not a script being recited.
Client-Ready Proposal
10/20✨ What it does
Claude drafts a consulting proposal covering problem understanding, phased workstreams, deliverables, and credentials for the client you [describe]. Confirm the timeline and fees, then send the version that sounds like you listened.
<task>Draft a consulting proposal</task> <client>[describe]</client> <problem>[describe]</problem> <our_approach>[describe]</our_approach> <output> 1. Executive summary 2. Our understanding of the problem (demonstrates listening) 3. Proposed approach (phased workstreams) 4. Deliverables per phase 5. Timeline + milestones 6. Team + roles 7. Fees + payment structure 8. Why us (brief credentials) 9. Next steps Tone: confident, client-centric.
Pro tip: Proposals win when the "our understanding" section is genuinely insightful. Generic understanding ("you want to grow revenue") gets you filed with competitors. Specific understanding ("you're losing mid-market customers to vertical competitors") builds trust.
Research & Analytics
5 promptsExpert Interview Guide
11/20✨ What it does
Claude builds an expert interview guide for [industry veteran, executive, customer, etc.] with themed questions, probes, and a time budget of [minutes]. Mark the questions you will actually ask, then keep the guide on the call.
<task>Build an expert interview guide</task> <topic>[describe]</topic> <expert_type>[industry veteran, executive, customer, etc.]</expert_type> <time>[minutes]</time> <output> - Opening + warmup - Background/credentialing questions - Core questions by theme (3-4 themes) - Follow-up probes per question - Challenge questions (test conflicting hypotheses) - Closing (ask for referrals, remaining resources) - Time budget per section </output>
Pro tip: Expert interviews are where real consulting insight lives: if you ask good questions. Most consultants are too polite. Good interviews politely push back on conventional wisdom. Have challenge questions ready.
Survey Design
12/20✨ What it does
Claude designs a survey for [purpose] with question order, types, branching, and a 5-7 minute length target. Confirm [what decisions this informs], then test the survey on one colleague before you send it.
<task>Design a survey for [purpose]</task> <audience>[describe]</audience> <decisions_supported>[what decisions this informs]</decisions_supported> <output> 1. Survey instrument (specific questions in order) 2. Question types (multiple choice, scale, open-ended) 3. Logic branching rules 4. Length target (optimal: 5-7 min) 5. Required sample size for statistical validity 6. Distribution channels 7. Expected response rate + incentive plan 8. Data analysis plan BEFORE launching </output>
Pro tip: Design your analysis BEFORE launching the survey. If you can't explain what you'll do with each question's data, remove the question. Surveys that collect unused data annoy respondents and dilute response rates.
Competitive Benchmarking
13/20✨ What it does
Claude builds a competitive benchmarking table for the company you [describe] against the competitors you [list], with sources, gaps, and targets. Verify each metric against a source, then pick one gap your team will close.
<task>Build a competitive benchmarking analysis</task> <our_company>[describe]</our_company> <competitors>[list]</competitors> <dimensions>[what to benchmark: revenue per employee, margins, time-to-market, etc.]</dimensions> <output> 1. Comparable metrics table across competitors 2. Data sources per metric 3. Our position relative to each competitor 4. Performance gaps (strengths + weaknesses) 5. Root causes of significant gaps 6. Realistic performance targets 7. Strategic implications </output>
Pro tip: Benchmarking reveals blind spots. "We're industry-standard" is often "we're 40% behind the leaders." Claude helps organize benchmarking output across 5-10 competitors quickly.
SWOT + PESTLE Combined
14/20✨ What it does
Claude runs a combined SWOT and PESTLE on [business], folding external factors into opportunities and threats with ranked implications. Challenge any factor that is guesswork, then take the top implication into your plan.
<task>Run combined SWOT + PESTLE on [business]</task> <business>[describe]</business> <output> PESTLE: - Political, Economic, Social, Technological, Legal, Environmental - External factors affecting business SWOT (internal): - Strengths, Weaknesses, Opportunities (external from PESTLE), Threats (external from PESTLE) Integration: - Which PESTLE factors create biggest opportunities - Which create biggest threats - How our internal S/W position us for these - Top 3 strategic priorities from integrated view </output>
Pro tip: SWOT alone is internally focused; PESTLE alone is externally focused. Combined, they show where external shifts create strategic urgency. Particularly valuable for industries in flux (AI, energy, healthcare).
Interview Synthesis
15/20✨ What it does
Claude synthesizes the interviews you [paste or describe] into themes, contradictions, consensus, minority views, and quotable lines. Star the tensions that reframe the case, then put those quotes on your slide.
<task>Synthesize insights from multiple expert interviews</task> <transcripts>[paste or describe]</transcripts> <output> 1. Themes that emerged across interviews (with frequency) 2. Contradictions between interviewees (interesting tensions) 3. Consensus views vs minority opinions 4. Surprising insights that reframe the problem 5. Quotable excerpts for client deliverables 6. Remaining questions to explore 7. Synthesized top-line insights </output>
Pro tip: Interview synthesis is where most consulting teams lose value. Individual interviews are read carefully; cross-interview synthesis gets rushed. Claude's long-context handling makes this dramatically better.
These prompts give you the what. Tutorials give you the why.
Learn when to use extended thinking, how to build Claude Projects, and workflows that compound. The library keeps growing.
Team & Operations
5 promptsWork Plan Design
16/20✨ What it does
Claude builds a week-by-week work plan for [engagement] covering workstreams, deliverables, assignments, and the critical path over [weeks]. Set [number] on the team, then lock the first week's checkpoints with your client.
<task>Build a work plan for [engagement]</task> <scope>[describe]</scope> <team_size>[number]</team_size> <duration>[weeks]</duration> <output> Per week: - Workstreams in flight - Deliverables due - Milestones + checkpoints - Hypotheses to validate - Team member assignments - Client touchpoints - Risks to watch Plus: dependencies between workstreams and critical path.
Pro tip: Work plans that don't have weekly hypotheses waste motion. Every week should validate or kill a hypothesis. Without that, consultants drift toward work that feels productive but doesn't advance the answer.
Client Kickoff Meeting Agenda
17/20✨ What it does
Claude builds a kickoff agenda for the engagement you [describe], covering objectives, not-scope, norms, and risks, timed to [hours]. Assign a lead per section, then send the agenda the day before your meeting.
<task>Build a client kickoff agenda</task> <engagement>[describe]</engagement> <duration>[hours]</duration> <output> - Opening + introductions (10 min) - Engagement objectives alignment - Scope confirmation (including NOT-scope) - Project plan walkthrough - Working relationship norms (cadence, communication, escalation) - Data access + stakeholder plan - Risks and dependencies - First deliverables preview - Q&A + next steps Per section: lead, key questions, expected outcomes.
Pro tip: Kickoffs that don't explicitly agree on "what's NOT in scope" create scope creep. Always name what you're not doing: it protects both sides and sharpens the work.
Consultant Career Coaching
18/20✨ What it does
Claude coaches you at the [analyst, associate, senior, partner track] stage toward [promotion, specialization, exit], with a readiness read and a 90-day plan. Keep the skills you will actually practice, then put week one on your calendar.
<task>Coach me on my consulting career</task> <current_stage>[analyst, associate, senior, partner track]</current_stage> <strengths>[describe]</strengths> <growth_areas>[describe]</growth_areas> <goal>[promotion, specialization, exit]</goal> <output> - Assessment of my readiness for next step - Specific skills to build - Visibility strategies within firm - Relationships to cultivate - Exit opportunities if not pursuing partnership - 90-day action plan - How to handle common career setbacks (bad reviews, tough managers) </output>
Pro tip: Consulting career progress is 50% work quality + 50% relationship capital. Most consultants over-invest in work and under-invest in internal relationships. Claude helps think through both dimensions.
Slide Feedback (Critical Review)
19/20✨ What it does
Claude reviews the slide you [describe or paste headline + body content] for a clear headline, MECE content, visual fit, and unsupported claims. Apply the headline fix first, then cut any number you cannot source.
<task>Review this slide for Pyramid Principle and visual clarity</task> <slide_description>[describe or paste headline + body content]</slide_description> <output> 1. Does the headline state the point? (yes/no + suggested fix) 2. Is the content MECE? 3. Does the visual support the point? 4. Any data gaps or unsupported claims? 5. How this slide connects to the surrounding narrative 6. Specific redlines 7. Rewrite suggestion </output>
Pro tip: Slide review by Claude catches what senior reviewers catch but at 1/10th the time cost. Use it for early self-review before showing decks to partners. Saves them time; makes you look sharper.
Post-Engagement Retro
20/20✨ What it does
Claude runs a post-engagement retro on the work you [describe], covering keep, stop, client lessons, team gaps, and what the firm should capture. Share the keep list with your team, then file one lesson in the knowledge base.
<task>Run a post-engagement retrospective</task> <engagement>[describe]</engagement> <outcome>[describe]</outcome> <output> Format the retro with: 1. What went well (keep doing) 2. What didn't go well (stop/change) 3. What we learned about the client 4. What we learned about ourselves (team dynamics, skill gaps) 5. Actions for next engagement 6. Knowledge to preserve in firm KM 7. Lessons to share with the firm </output>
Pro tip: Retros that aren't captured die. Turning retros into actionable knowledge (playbooks, templates, case studies) is what separates firms that compound vs those that relearn the same lessons.
Free tool
Prompt Optimizer
Turn a rough idea into a structured, professional AI prompt.
Frequently Asked Questions
Prompts are the starting line. Tutorials are the finish.
A growing library of hands-on tutorials on ChatGPT, Claude, Midjourney, and many other AI tools. New tutorials are added regularly.
7-day free trial. Cancel anytime.