Case Interview Hypothesis: How to Build, State, and Test Your Approach
A case interview hypothesis is a testable proposition about what drives the client's business challenge and which part of your structure to examine first. On its official case preparation pages, Bain advises taking 1 to 2 minutes to structure your thinking before speaking, while BCG and McKinsey emphasize clarifying assumptions, showing your thinking out loud, and sharing solutions built for the specific problem rather than reciting standard frameworks.
Updated
What a hypothesis does in a case interview
In consulting roles, firms expect you to analyze complex business challenges, develop hypotheses, identify opportunities, and recommend practical solutions, as Bain states in its Retail Business Analyst role description. Inside a live case study interview, a hypothesis is not a blind guess or a final conclusion. It is a directional starting point that tells the interviewer which part of your structure you want to test first and why.
Without a hypothesis, you risk treating a framework like a long checklist and asking disconnected questions across every category. With a hypothesis, you lay out a clear structure first, then pick the highest-priority branch to investigate with data. If the numbers support your idea, you drill deeper into that branch. If the numbers refute it, you cross that branch off cleanly and move to the next one.
You can Practice a case free, no sign-up to test how you state your opening structure and hypothesis out loud before your interview.
Generic framework vs issue tree vs hypothesis tree
Do not blur the distinction between three different tools: a generic framework, an issue tree, and a hypothesis tree. Understanding how these tools differ helps you move from passive memorization to active problem solving.
A generic framework is an off-the-shelf business taxonomy, such as the 4Ps of marketing or the 3Cs. Reciting a generic framework signals to the interviewer that you are working from memory rather than listening to the client's specific business context.
An issue tree is a neutral breakdown of the question. It divides the client's core problem into mutually exclusive and collectively exhaustive components. An issue tree does not assume an answer. It simply maps out all potential drivers of the problem.
A hypothesis tree turns those neutral question branches into falsifiable statements. Instead of asking what is happening to revenue, a hypothesis tree posits that revenue declined because discount pricing eroded gross margin, and then outlines the tests needed to prove or disprove that claim.
Comparison of problem-solving tools in case interviews
| Tool | Core structure | How it sounds in an interview | When to use it |
|---|---|---|---|
| Generic framework | Pre-memorized topic buckets such as marketing mixes or generic profitability splits. | "I want to look at customers, competitors, costs, and capabilities." | Avoid using as a rigid checklist. McKinsey advises sharing solutions built for the scenario. |
| Issue tree | A neutral, MECE map of the problem built specifically for the client's objective. | "Profitability is driven by revenue per store, volume per store, and operating cost per store." | Use immediately after clarifying questions to map all possible drivers of the client's goal. |
| Hypothesis tree | An opinionated tree of testable statements organized by likely business impact. | "Store profit fell because local labor rates rose faster than customer traffic. Let us check hourly wages first." | Use once your issue tree is established to prioritize which branch to analyze first. |
What McKinsey, BCG, and Bain officially look for
All three firms publish explicit guidance on how they evaluate your reasoning during a problem-solving interview. None of them grade you on guessing a hidden answer in the first minute. BCG and Bain both state on their case preparation pages that there is not always a single right answer, and that what matters most is how you approach the problem and the quality of your reasoning.
On its interviewing page, McKinsey advises candidates to listen carefully and share tailored solutions vs standard frameworks. In its Beautify sample case, McKinsey also reminds candidates that calculators are not allowed, so you should write out your calculations on paper and talk your interviewer through your steps to demonstrate an organized approach.
Read our McKinsey case interview, BCG case interview, and Bain case interview guides to see how each firm runs its problem-solving rounds.
Official case interview expectations at McKinsey, BCG, and Bain
| Firm | What the case evaluates | Official guidance on reasoning and communication | Official practice cases and tools |
|---|---|---|---|
| McKinsey | How you structure tough, ambiguous challenges, identify important issues, deal with facts and data, and formulate conclusions. | Share tailored solutions vs standard frameworks. Talk your interviewer through your steps on pen and paper without calculators. | 5 sample cases: Beautify, Diconsa, Electro-Light, Transforming a national education system, and Talbot Trucks. |
| BCG | How you structure your approach, ask thoughtful questions, analyze data, perform quick calculations, and identify the most important factors. | Listen actively, think structurally, communicate clearly, clarify assumptions, and show your thinking. Do not rush, stay silent, or overcomplicate. | OneDay@BCG free job simulation and case preparation tips. |
| Bain | How you think in a logical and structured way, balancing precision and creativity on a real-world business scenario. | Start by understanding the question, ask questions if you need information, take 1 to 2 minutes to think about your approach, and structure your thinking. | Coffee Shop Co. sample case and case interview prep hub. |
Case structure vs case hypothesis
It is easy to confuse your structure with your hypothesis, or to try to replace one with the other. You need both, and they do different jobs in the first few minutes of the case.
Your structure is the complete map of the problem. It breaks the client's objective into three or four distinct drivers so you do not miss a major factor. Your structure stays objective and neutral even if your first theory turns out to be wrong.
Your hypothesis is your first route on that map. Once you lay out the three or four branches of your structure, you state a hypothesis to explain which single branch you want to examine first. If data shows that branch is not the driver, you do not have to invent a brand-new structure. You return to your original map and test the second branch.
- Step 1: Clarify the objective and confirm assumptions with two or three focused questions.
- Step 2: Take 1 to 2 minutes to build a structure built for this client.
- Step 3: Present your structure out loud, naming each branch and what sits inside it.
- Step 4: State your initial hypothesis and ask for the specific piece of data that will test that first branch.
How to formulate a hypothesis: quantitative magnitude vs qualitative business intuition
A strong hypothesis is not a random hunch. You form it by examining two inputs: quantitative magnitude from the initial prompt and qualitative business intuition.
When the prompt provides preliminary financial numbers, prioritize your hypothesis based on quantitative magnitude. Look for the largest cost bucket, the fastest-changing revenue stream, or the segment that represents the bulk of operating expenses. For example, if raw materials represent 70% of cost of goods sold, your initial cost hypothesis should focus on procurement and supplier pricing rather than administrative overhead.
When the initial prompt gives zero numbers, rely on qualitative business intuition. Consider the industry context, customer behavior, and competitive dynamics. If an established retail brand sees traffic drop after a direct-to-consumer competitor launches a mobile platform, common sense suggests exploring customer channel preferences and digital convenience first.
State your hypothesis with a clear rationale. Rather than saying "I hypothesize costs went up," say "Because raw materials make up the majority of direct expenses, my hypothesis is that supplier price increases are driving the margin decline, so I would like to examine the unit cost breakdown first."
How hypothesis specificity evolves from opening to final synthesis
Your hypothesis is not static. It must evolve across the case interview as you uncover new exhibits, finish calculations, and receive hints from the interviewer.
At the opening of the case, your hypothesis is broad and directional. You do not yet have enough data to name exact percentages or specific product stock-keeping units. An opening hypothesis establishes an initial vector, such as predicting that margin contraction is driven by variable manufacturing costs rather than sales volume.
In the middle of the case, your hypothesis becomes diagnostic and segment-specific. As you receive exhibits, you test sub-hypotheses. For instance, after confirming that variable costs rose, your mid-case hypothesis tests whether labor overtime or commodity steel prices drove the increase.
By the final synthesis, your hypothesis transforms into a concrete recommendation supported by data. You present an answer-first conclusion that names the exact operational levers, the financial impact, the key implementation risks, and the next steps.
How to use a hypothesis in candidate-led vs interviewer-led cases
How you vocalize and test your hypothesis depends on whether you are driving the sequence of questions or the interviewer is guiding the prompts.
In a candidate-led case, common at BCG and Bain, you choose which bucket to explore next. After presenting your structure, you state a working hypothesis and ask for the numbers to test it. If you want to check pricing, you propose your reasoning and explicitly request the pricing schedule.
In an interviewer-led format, common at McKinsey, the interviewer asks a set series of questions. You still use hypothesis-driven thinking inside each question: you organize your factors logically, state what must be true for the investment to work, and test that threshold with the case math.
Worked example: Testing a break-even hypothesis with Bain Coffee Shop Co. case math
In Bain's Coffee Shop Co. sample case, a friend asks whether opening a coffee shop in Cambridge, England is a good idea. To test the hypothesis that local resident demand is large enough for the shop to break even in year 1, you can run a break-even analysis using the exact assumptions published in the case.
The case provides the following inputs: Cambridge has a population of 100,000 residents who drink an average of 1 cup of coffee per day. The price per coffee is £3, the variable cost per cup is £1, the upfront cost to open the shop is £245,610, and the annual running cost is £163,740. By our calculation from Bain's Coffee Shop Co. case data, here is our worked example testing the break-even hypothesis step by step.
- Market size calculation: 100,000 residents multiplied by 1 cup per day equals 100,000 cups per day. Multiplying 100,000 cups by 365 days gives a total Cambridge resident market of 36,500,000 cups of coffee per year.
- Contribution margin per unit: The selling price of £3 minus the variable cost of £1 equals a contribution margin of £2 per cup.
- Total Year 1 fixed outlays: Adding the upfront opening cost of £245,610 to the annual running cost of £163,740 gives £409,350 in total first-year fixed cash outlays.
- Break-even volume in cups: Dividing the total fixed outlays of £409,350 by the £2 contribution margin per cup equals 204,675 cups required to break even in Year 1.
- Daily required volume: Dividing 204,675 cups by 365 days equals 560.75, or approximately 561 cups per day.
- Market share required: Dividing 204,675 cups by the annual market size of 36,500,000 cups equals 0.0056075, or approximately 0.56% of the resident market.
- Evaluating the hypothesis: Capturing 0.56% of local resident volume is a plausible share for one shop, but selling 561 cups every day requires high foot traffic and store throughput, so your next step is to test prime location availability and peak hourly store capacity.
Worked example 2: Testing a payback hypothesis with McKinsey Beautify case math
In Question 3 of McKinsey's Beautify sample case, the team evaluates whether training department store beauty consultants to become virtual social media advisors will turn profitable, defined as when incremental revenues exceed the cost of the project.
The case provides the following inputs: Beautify's baseline revenues are €1.3 billion. Virtual advisors are expected to bring a 10% overall increase in incremental revenue in the first year. In that first year, upfront investments are €50 million in IT, €25 million in training, €50 million in remodeling department store counters, and €25 million in inventory. In addition, all-in yearly ongoing costs associated with the shift are €10 million and start in the first year. By our calculation from McKinsey's Beautify case data, here is our worked example testing the payback hypothesis step by step on paper.
- Incremental revenue: Multiplying baseline revenue of €1.3 billion by the 10% expected increase yields €130 million in incremental revenue per year.
- Upfront investment total: Adding €50 million for IT, €25 million for training, €50 million for remodeling counters, and €25 million for inventory yields €150 million in one-time upfront costs.
- Year 1 net cash flow: Adding the €10 million ongoing cost to the €150 million upfront investment gives €160 million in total Year 1 costs. Subtracting €160 million in costs from €1-30 million in incremental revenue leaves a net deficit of -€30 million at the end of Year 1.
- Year 2 net cash flow: In Year 2, the upfront investment of €150 million does not repeat. Incremental revenue is €130 million and ongoing costs are €10 million, generating a positive net cash flow of €120 million in Year 2.
- Cumulative payback: Combining the Year 1 deficit of -€30 million with the Year 2 gain of €120 million results in a cumulative net profit of €90 million by the end of Year 2. The project therefore recovers its investment and turns profitable during Year 2, taking approximately 1.25 years to pay back.
How to pivot when the data disproves your hypothesis
A disproven hypothesis is completely normal in a case interview and does not hurt your evaluation if you respond logically. Interviewers test whether you listen to the facts or cling to a theory that the data has eliminated.
When an exhibit or calculation contradicts your opening hypothesis, follow a clean three-part pivot out loud: state what the data shows, explicitly rule out the current branch, and name the next branch in your structure that you will test.
- Weak response (ignoring the data): "Even though unit costs were flat, I still think procurement might be an issue, so I want to ask more questions about supplier contracts."
- Strong response (clean pivot): "Since unit costs and overhead held steady over the past two years, cost is not driving the profit drop. I am going to rule out the cost branch of my structure and move to the revenue branch, specifically testing whether average selling price or customer volume fell."
When not to force an upfront hypothesis
Not every question inside a case interview calls for an upfront hypothesis. Forcing one where it does not belong makes your answer sound mechanical.
Do not state a directional hypothesis during pure market-sizing estimates before you have cost or revenue figures. For example, when estimating the volume of coffee consumed in Cambridge, an upfront hypothesis about profitability is premature because you are simply sizing a demographic baseline.
Do not force an upfront hypothesis during open-ended creative brainstorming prompts, such as Question 2 of McKinsey's Beautify case, which asks what website features would make a department-store shopper switch to virtual consultations. In that question, the goal is to generate a structured range of consumer value propositions rather than proving a single theory.
Finally, avoid stating an upfront hypothesis during initial clarifying questions when you do not yet know the client's concrete objective or operational constraints. In those moments, focus on clarifying assumptions and building a clean structure first.
Checklist: 4-step routine to rehearse hypotheses out loud
McKinsey notes on its interviewing page that many candidates find it helpful to practice aloud with family or friends, and that responsible use of AI to practice interview questions and explain concepts is encouraged during preparation. Use this four-step drill written by our team to make hypothesis generation automatic before your interview day.
To put these steps into practice with live pushback on your reasoning, continue with the AI interviewer.
- 1. Structure within 1 to 2 minutes: Practice reading a business prompt, taking 1 to 2 minutes to plan your approach, and sketching a three-branch structure built for that specific client.
- 2. Say the bridge sentence out loud: Practice finishing your structure with one clear transition sentence that states your initial hypothesis and requests the first data point.
- 3. Practice the forced pivot: Have a practice partner or AI tool tell you that your first branch is flat or unchanged, and practice pivoting out loud to your second branch in under 15 seconds without losing your composure.
- 4. Talk through paper math step by step: Work calculations on paper without a calculator, stating the formula and units out loud before you compute the digits so the interviewer can follow your logic.
From McKinsey & Company: Getting Ready for Your Interviews, McKinsey & Company: Beautify Practice Case, McKinsey & Company: Interview Tips and Problem Solving, BCG Careers: Consulting Case Study Interview Preparation, Bain & Company: Case Interview Preparation, Bain & Company: Coffee Shop Co. Sample Case and Bain & Company: Retail Business Analyst Job Description (Job ID 111871).
1 practice question and sample answer
Say your answer out loud first, then read the sample answer.
Practice question
Using the setup from Bain's Coffee Shop Co. sample case, your friend wants to know whether opening a coffee shop in Cambridge, England is a good idea. After taking 1 to 2 minutes to organize your thoughts, how do you present your structure and initial hypothesis out loud?
Sample answer. Example answer written by our team. The applicant is invented. 'To help our friend decide whether to open the coffee shop in Cambridge, I would structure this around three areas. First, local market demand and competition: how many cups Cambridge residents buy each day, and what share a single independent shop can realistically win against existing cafes. Second, unit economics and first-year break-even: the price per cup minus the variable cost per cup, and how many cups she must sell in year one to cover both the upfront build-out cost and the annual operating expenses. Third, operational execution and location: whether she can secure a high-foot-traffic storefront in Cambridge that can physically serve that daily break-even volume. My initial hypothesis is that opening the coffee shop is financially viable if she can reach first-year break-even by capturing less than 1% of the local resident coffee market. To test that hypothesis first, I would like to estimate total annual coffee demand among Cambridge residents and compare it against her first-year break-even volume. Should we start by sizing the Cambridge resident market?'
It lays out a three-part structure built specifically for a single coffee shop decision, states a clear and falsifiable hypothesis tied to break-even market share, and asks to start with the exact analysis that tests that hypothesis.
Practice the case interview, out loud
Soreno's AI interviewer runs a full case one question at a time, checks your math and scores your structure, any time you have 30 minutes.
Questions people ask
Does a case interview require arriving at a single right hypothesis?
No. Both BCG and Bain state on their official case preparation pages that there is not always a single right answer in a case interview. Interviewers evaluate how you structure the problem, clarify assumptions, and adjust your reasoning as new data appears.
Should I memorize standard frameworks to build my hypothesis?
No. McKinsey explicitly advises candidates to listen carefully and share solutions built for the specific business scenario rather than relying on standard frameworks.
Why should I talk through my hypothesis and math steps out loud?
In its Beautify sample case, McKinsey explains that talking your interviewer through your steps demonstrates an organized approach, and the more you talk, the easier it is for your interviewer to help you. BCG likewise lists 'Show Your Thinking' and 'Clarify Assumptions' among its core case interview rules.
Can I use a calculator or AI when testing my hypothesis in a live interview?
No. McKinsey's Beautify case and assessment integrity rules state that calculators and real-time AI tools are not allowed during interviews, so you should write out your calculations using pen and paper.
How long should I take to structure my approach before stating a hypothesis?
Bain's Coffee Shop Co. sample case recommends taking 1 to 2 minutes to think about how you will approach the problem and structure your thinking before communicating your ideas to the interviewer.
More guides
- Case Interview Coaching: Prepare and Practice Out Loud
- Clarkston Consulting Interview: Questions & 6 Steps
- Market Questions in Case Interviews: Examples & Math
- Mini Case Interview: Format, Steps, and Practice Cases
- PwC Management Consulting: Services, Strategy& and Careers
- Slalom Interview Questions: Core Values, Cases and AI Rules
- Consulting Analyst Interview Questions and Answers (2026)
- Consulting Client Interview Questions, With Answers
- Environmental Consulting Interview Questions and Answers
- Life Sciences Management Consulting Firms (2026 Guide)
- McKinsey Acceptance Rate: Official Hiring Process & Stages
- Risk Consulting Interview Questions and Answers (2026)