ChatGPT Case Interview Prompts: 15 That Work (2026)

Author: Taylor Warfield, Former Bain Manager and interviewer.

Last Updated: June 10, 2026

 

ChatGPT case interview prompts turn a free chatbot into a strict consulting interviewer that runs full cases, drills your math, and grades every answer against MBB standards. This article gives you 15 copy-paste prompts, a 5-part formula for writing your own, and a 25-minute daily routine that ties them together.

 

Before reading on:

 

Most candidates struggle to land interviews and even fewer turn them into offers. Watch my free training to learn how to triple your chances of landing interviews and increase your chances of receiving an offer by 8x.

 

👉 Watch for free

 

Key Takeaways

 

The best ChatGPT case interview prompts cast ChatGPT as a strict MBB-style interviewer, anchor it in a specific business scenario, and force it to grade your answers against real interview standards.

 

  • Generic prompts fail because ChatGPT defaults to agreeable, which is the opposite of how real consulting interviewers behave

 

  • The 15 prompts in this article cover full cases, structuring, math, market sizing, exhibits, brainstorming, synthesis, and behavioral stories

 

  • The CASTS formula (Cast, Anchor, Set grading, Time-box, Suppress flattery) lets you write your own prompt for any skill

 

  • A 25-minute daily routine using 3 of these prompts covers math, one full mini case, and synthesis

 

  • ChatGPT prompts handle practice volume, but you still need human mock interviews for realistic pressure and honest feedback

 

Why Do Most ChatGPT Case Interview Prompts Fail?

 

Most ChatGPT case interview prompts fail because ChatGPT is trained to be agreeable, and agreeable is the opposite of how a real consulting interviewer behaves. Ask it to give you a case and it produces a vague scenario, accepts whatever framework you offer, and tells you that you did great. That kind of practice builds false confidence, not interview skills.

 

The scale of the problem is real. OpenAI reported in late 2025 that ChatGPT passed 800 million weekly active users, and a large share of consulting candidates now use it for case practice with prompts that train nothing.

 

In my time interviewing at Bain, candidates rarely failed cases on knowledge. They failed on structure under time pressure, and only timed, graded, adversarial practice fixes that. Every prompt below is built to recreate those exact conditions.

 

How Do You Write a ChatGPT Case Interview Prompt That Works?

 

A ChatGPT case interview prompt works when it does five things, which I call the CASTS formula: it casts a role, anchors a scenario, sets grading rules, time-boxes the work, and suppresses flattery. Every one of the 15 prompts below follows this structure, and you can use it to build your own.

 

  • Cast the role: name the firm, the interviewer's seniority, and the case style, since "act as a strict McKinsey partner" produces sharper behavior than "act as an interviewer"

 

  • Anchor the scenario: give an industry, a client, and a specific number, like a retailer with a 15% profit decline, so the case has real stakes

 

  • Set the grading rules: tell ChatGPT exactly what to score, such as structure, math, and communication on a 1 to 5 scale, and make it name your weakest area

 

  • Time-box everything: 45 seconds per math problem, 90 seconds per market sizing estimate, and 30 seconds per final synthesis

 

  • Suppress the flattery: ban praise unless the answer would genuinely pass at McKinsey, BCG, or Bain

 

Miss any one of the five and the quality of the practice drops fast. The most common omission is the last one, and it is the most important.

 

What Are the 15 Best ChatGPT Case Interview Prompts?

 

The 15 best ChatGPT case interview prompts cover six skill groups: full case simulations, case fundamentals, quantitative drills, judgment drills, knowledge and feedback, and behavioral preparation. Copy each prompt as written, swap in your target industry or firm, and run it in voice mode whenever you can so you practice speaking, not typing.

 

Prompt group

Prompts

Skill trained

Full case simulations

#1 to #2

Running a complete case under interview conditions

Case fundamentals

#3 to #4

Clarifying questions and framework building

Quantitative drills

#5 to #7

Market sizing, case math, and exhibit reading

Judgment drills

#8 to #10

Brainstorming, hypothesis, and synthesis

Knowledge and feedback

#11 to #12

Industry economics and framework critique

Behavioral and AI rounds

#13 to #15

PEI stories, intro answers, and AI collaboration

 

Prompt #1: Run an Interviewer-Led Profitability Case

 

Profitability is the most common case type you will face, showing up in roughly 80% of first-round interviews in some form. Start here, since a profitability case trains the core mechanics every other case type builds on.

 

Prompt: "Act as a strict McKinsey interviewer running an interviewer-led profitability case about a regional grocery chain whose profits fell 15% last year. Drive the case yourself, ask me one structuring or math question at a time, and share data only when I ask for it. After each answer, grade my structure, math, and communication from 1 to 5 and name my single biggest gap."

 

Prompt #2: Run a Candidate-Led Market Entry Case

 

BCG and Bain run candidate-led cases, where you drive the structure and decide what to analyze next. A market entry case is the best format for practicing this because it forces you to prioritize among market, competition, capabilities, and financials.

 

Prompt: "Act as a Bain interviewer running a candidate-led market entry case about a US coffee chain considering entering Japan. Let me drive the case, but ask one skeptical follow-up after every framework bucket I present and give me realistic data only when I request it. At the end, ask for my final recommendation and grade it on answer-first structure, supporting evidence, and risks."

 

Prompt #3: Drill Clarifying Questions

 

The first 2 minutes of a case decide whether you solve the right problem. Strong clarifying questions pin down the objective, the client's definition of success, and the scope before you touch a framework.

 

Prompt: "Give me a deliberately vague one-sentence case prompt, such as a client whose revenue is flat. Before I am allowed to structure anything, make me ask 3 clarifying questions and grade each one on whether it narrows the objective, defines the client's goal, or rules out a major branch. Reject any question that just stalls for time."

 

Prompt #4: Build Frameworks Under Time Pressure

 

Memorized frameworks fail because interviewers can spot them in seconds. The skill is building custom case interview frameworks tailored to the exact client and question, and that skill only develops through timed reps with critique.

 

Prompt: "Give me 5 case prompts one at a time, spanning profitability, market entry, pricing, M&A, and operations. For each, give me 2 minutes to lay out a framework with 3 to 4 buckets, then critique whether my buckets are MECE, tailored to the case, and deep enough to drive analysis. Show me a stronger version after each attempt so I can compare."

 

Prompt #5: Practice Market Sizing in 90 Seconds

 

Interviewers care more about your logic than your final number. This prompt drills market sizing the way it gets tested: stated assumptions, staged calculations, and a sanity check at the end.

 

Prompt: "Generate 10 market sizing questions across consumer goods, healthcare, software, and industrial markets. Make me state my assumptions out loud, calculate in stages, and sanity check the final number against a known benchmark like the 330 million US population. Time me at 90 seconds each and penalize skipped units or unstated assumptions."

 

Prompt #6: Drill Case Math With Messy Numbers

 

Round numbers in practice create a false sense of security, because real cases use figures like $87 average order value and 4.7 million customers. Daily reps on case interview math with messy inputs are the fastest fix.

 

Prompt: "Act as an interviewer drilling case math with 15 problems mixing percentages, growth rates, breakeven, and weighted averages. Use messy numbers like 4.7 million customers and $87 average order value, and give me 45 seconds per problem. After each one, tell me whether I set up the calculation correctly, kept clean stages, and translated the answer into a business takeaway."

 

Prompt #7: Interpret Charts and Exhibits

 

Nearly every McKinsey case includes at least one exhibit, and the test is whether you can state the so-what in one sentence. ChatGPT cannot show you a real chart, but it can describe one precisely enough to drill the interpretation skill.

 

Prompt: "Describe a chart or table in words, such as revenue by segment over 5 years with one segment quietly declining, and ask me to interpret it. Make me state the so-what in one sentence, identify the most important trend, and link it to a hypothesis about the client's problem. Then give me a harder exhibit with a deliberate distraction in the data."

 

Prompt #8: Structure Your Brainstorming

 

Listing 10 random ideas scores worse than listing 6 ideas in 2 clean categories. Brainstorming in case interviews is a structure test disguised as a creativity test, and this prompt trains the structure first.

 

Prompt: "Give me a brainstorming question, such as ways a hotel chain could increase revenue per room. Make me organize my ideas into 2 to 3 categories before listing anything, then point out overlapping ideas, missing branches, and any category that is not distinct. Repeat with a cost-side question so I practice both directions."

 

Prompt #9: State and Defend a Hypothesis

 

Top candidates state a case interview hypothesis early and update it as data arrives, which signals consultant-level thinking. Weak candidates either never commit or abandon their structure the moment data contradicts them.

 

Prompt: "Run a short case and stop me after the framework to ask for my initial hypothesis. Force me to state a specific, testable hypothesis tied to one branch of my structure, then challenge it with one piece of contradicting data. Grade whether I update the hypothesis logically instead of abandoning my structure."

 

Prompt #10: Deliver a 30-Second Synthesis

 

The case interview recommendation is the last thing your interviewer hears, so it carries outsized weight in the final evaluation. Having coached hundreds of candidates 1-on-1, I can tell you synthesis is the most undertrained skill in case prep.

 

Prompt: "Give me a case prompt plus 3 data points, then ask for a final recommendation in 30 seconds. Require answer-first ordering: the recommendation, 2 to 3 supporting reasons, one risk, and one next step. Grade me on whether a partner could repeat my recommendation to a client word for word."

 

Prompt #11: Learn an Industry in 10 Minutes

 

You do not need industry expertise to pass a case, but knowing that airlines run on 3% to 5% net margins or that software companies live on retention makes your assumptions sharper. This prompt builds a quick mental model of any industry before you practice cases in it.

 

Prompt: "Teach me the economics of the airline industry in 10 bullet points a consultant would know: revenue drivers, cost structure, typical margins, and 2 current trends with rough numbers. Then quiz me with 3 questions to confirm I can apply the knowledge in a case. Repeat for any industry I name."

 

Prompt #12: Get Brutal Feedback on Your Framework

 

This is the prompt to run after every practice case, whether you did the case with ChatGPT, a partner, or alone. Paste your structure and demand a graded critique with a rewrite of the weakest bucket.

 

Prompt: "Here is the framework I built for this case prompt: [paste your case and framework]. Grade it from 1 to 5 on MECE structure, tailoring to the specific client, and depth, then rewrite my weakest bucket. Do not soften the feedback, and do not call it good unless it would pass at McKinsey, BCG, or Bain."

 

Prompt #13: Pressure-Test Your PEI Stories

 

The McKinsey PEI now tests four traits after the firm's mid-2025 update: Connection, Drive, Leadership, and Growth. Every story needs a clear personal role, real resistance, and a measurable outcome, and most candidates are missing at least one of the three.

 

Prompt: "Act as a McKinsey interviewer running the Personal Experience Interview and ask me for a story about leading through disagreement. After I tell it, ask 3 probing follow-ups about my specific role, the resistance I faced, and the measurable outcome. Then tell me which of those 3 areas was weakest and exactly how to fix it."

 

Prompt #14: Tighten Your Tell Me About Yourself Answer

 

Your answer to tell me about yourself sets the tone for the entire interview, and it should run about 90 seconds. This prompt cuts the fluff and surfaces the follow-up questions your answer invites.

 

Prompt: "Here is my answer to tell me about yourself: [paste your answer]. Cut it to 90 seconds, restructure it as past, present, and future with a clear thread pointing to consulting, and flag any line that sounds generic. Then ask me the 2 follow-up questions my answer most invites."

 

If behavioral prep is your bottleneck, my fit interview course helps you master 98% of consulting fit questions in a few hours.

 

Prompt #15: Train for AI-Assisted Interview Rounds

 

The Guardian reported in January 2026 that McKinsey is piloting AI-assisted interview rounds using its internal tool Lilli, with candidates partly graded on how they collaborate with AI. The skill being tested is judgment: spotting weak AI output, fixing it, and explaining your reasoning out loud.

 

Prompt: "Give me a deliberately flawed framework for a profitability case with one overlapping bucket and one missing driver. My job is to find both flaws, explain why each one matters, and fix the structure out loud. Grade how quickly and precisely I caught the problems."

 

How Do You Turn These Prompts Into a Daily Practice Routine?

 

The most effective routine runs 4 of these prompts in a fixed 25-minute daily rotation, then adds one human mock interview per week. Consistency beats intensity: 25 minutes daily for 4 weeks outperforms two 6-hour weekend sessions.

 

  1. Warm up with math (5 minutes): run Prompt #6 for five problems, since daily case interview math drills compound faster than any other practice

  2. Run one mini case (10 minutes): alternate Prompt #1 and Prompt #2 so you train both interviewer-led and candidate-led formats

  3. Drill one weak skill (5 minutes): rotate Prompts #5, #7, and #8 based on whatever scored lowest yesterday

  4. Close with synthesis (5 minutes): run Prompt #10 once, since the recommendation is the highest-stakes 30 seconds of any case

  5. Add one human mock per week: a live partner supplies the pressure and unpredictability no prompt can recreate

 

This routine also solves the partner problem, since most of it works when you practice case interviews by yourself. Keep a written log of every score below 4 so your drills target real weaknesses instead of random topics.

 

If you want to learn case interviews quickly and systematically, my case interview course walks you through proven strategies in as little as 7 days.

 

Where Do ChatGPT Case Interview Prompts Fall Short?

 

ChatGPT case interview prompts handle practice volume, but they cannot fully replicate interviewer pressure, track your weak spots across sessions, or render real visual exhibits. Knowing these three gaps tells you exactly what to supplement.

 

The first gap is flattery that survives even strict prompts. ChatGPT softens criticism more than any real interviewer would, which is why ChatGPT for case interview prep works best as a volume tool rather than a judge. Asking "name three specific things I did wrong" gets far more honest output than asking "how did I do."

 

The second gap is math reliability on long chains. Many candidates switch to Claude for case interview prep when a calculation runs past 4 or 5 steps, since it drops fewer digits on extended math. Whichever model you use, rebuild every important calculation yourself before trusting it.

 

The third gap is memory. ChatGPT will not remember that you missed three breakeven problems last week, so your written log has to do that job. Without it, drilling becomes random instead of targeted.

 

For honest, graded feedback from someone who has sat on the other side of the table, my case interview coaching gives you the 1-on-1 pressure-testing that no prompt can deliver.

 

Frequently Asked Questions

 

What is the best ChatGPT prompt for case interview practice?

 

The best single prompt casts ChatGPT as a strict McKinsey or Bain interviewer running a timed, interviewer-led case with explicit grading rules. Prompt #1 in this article does exactly that: it forces one question at a time, releases data only on request, and grades every answer from 1 to 5. Generic prompts like give me a case produce agreeable filler that builds false confidence.

 

Can ChatGPT run a full case interview?

 

Yes, ChatGPT can run a passable full case if you cast it as a strict interviewer and set grading rules. It handles case generation, structuring questions, math, and synthesis well. It still falls short on realistic pushback and visual exhibits, so pair it with human mock interviews before the real thing.

 

Do I need ChatGPT Plus for case interview prompts?

 

No, the free tier is enough for daily practice volume. Free users get a capable model plus voice mode, which is ideal for practicing answers out loud. Paid tiers mainly add longer sessions and fewer usage limits.

 

How many cases should I practice with ChatGPT before my interview?

 

Most successful candidates complete 20 to 40 full cases before MBB interviews, and ChatGPT can supply unlimited volume toward that total. Aim for 25 minutes of prompted drills daily plus 1 to 2 human mock interviews per week. ChatGPT covers volume while humans cover realism.

 

Is it cheating to use ChatGPT for case interview prep?

 

No, practicing with ChatGPT prompts is no different from using a prep book or a practice partner. The skills you build are genuinely yours: structuring, math, and synthesis. Some firms now even evaluate how well candidates collaborate with AI tools during interviews.

 

Should I use ChatGPT or Claude for case interview practice?

 

Use ChatGPT for daily drill volume and voice practice, and switch to Claude for long calculation chains where ChatGPT sometimes drops digits. The 15 prompts in this article work in both tools with no edits. The model matters far less than the strictness of your prompt.

 

Everything You Need to Land a Consulting Offer

 

Need help passing your interviews?

  • Case Interview Course: Become a top 10% case interview candidate in 7 days while saving yourself 100+ hours

  • Fit Interview Course: Master 98% of consulting fit interview questions in a few hours

  • Interview Coaching: Accelerate your prep with 1-on-1 coaching with Taylor Warfield, former Bain interviewer and best-selling author

 

Need help landing interviews?

 

Need help with everything?

 

Not sure where to start?