Case interview frameworks that still work in 2026

In this article

1. Before you start
2. Why interviewers ask this
3. Case interview frameworks from first principles
4. Worked example
5. Trade-offs and edge cases
6. How to answer it in an interview
7. Common mistakes
8. Follow-up questions
9. Sources

Before you start

Why interviewers ask this

A framework is the first thing a case interviewer sees, and it is what separates a thinker from a reciter. If the structure you state fits the question, the rest of the case runs smoothly. If it is a memorised list that half fits, you spend the case defending it. Interviewers use your first structure to judge how you reason when nobody has told you the answer.

The published guidance points in two directions. Bain's case tips say "Use a clear framework and walk your interviewer through your approach" (Bain careers, read October 2026). A Bain recruiting specialist writes that "a common mistake is trying to force-fit a pre-learned framework" and that "we're evaluating how you think, not the framework you memorized" (Bain careers blog, March 2026). The do list of Boston Consulting Group (BCG) includes "Think Structurally" and its don't list "Overcomplicate Solutions" (BCG careers, read October 2026). Both are consistent: use a structure, built to fit.

Case interview frameworks from first principles

Case framework

A case framework is a reusable starting structure for a business question, such as splitting profit into revenue and cost, that gives you a first set of branches to adapt to the problem in front of you instead of a script to recite.

MECE

MECE, short for mutually exclusive and collectively exhaustive, is a rule for splitting a problem into branches that do not overlap and together leave nothing out, so that every possible cause or option belongs to exactly one branch.

Hypothesis

A hypothesis is a specific, testable answer you state early in a case, such as "the profit drop comes from unit cost", chosen to tell you which branch to examine first and then confirmed or dropped by data.

In plain words

Think of packing for a trip. A packing list from the internet is a framework: it gets you started. But you pack differently for a beach week and for a ski week. A good packer starts from the trip, builds a list that covers everything needed once, and uses the internet list only to check nothing was forgotten.

The analogy breaks in one place. A forgotten sock costs little. In a case, a branch that overlaps another or leaves a gap sends the interviewer's follow-ups straight into it. In one precise sentence: a case framework is a structure you derive from the question and check against a few standard lenses, not a list you recite.

Step 1: Start from the question, not from a framework

Before choosing any structure, restate the objective and clarify. The input is the prompt. The output is one sentence of goal and a few agreed facts: "The goal is to decide whether to enter, measured by profit within three years. May I assume we cannot change the product?"

BCG's do list names "Listen Actively" and "Clarify Assumptions", and its don't list "Rush Your Response" (BCG careers, read October 2026). The reason is practical: the same words can mean different problems. "Profits are falling" is one case if costs rose and another if a competitor arrived. A framework chosen before you know which is a guess. Ask two or three questions, then take a moment in silence to think, as a pause reads as care. What you now know: the goal sentence you say here is what every branch is tested against.

Step 2: Build a tree from the question, not from memory

An issue tree starts with the question at the top and splits it into branches. Each level must be MECE: no overlap, no gaps. Test it by asking two questions of every split: could one fact belong to two branches, and could a real cause fall in none?

This is the payoff of the lead. A tree built from the question has branches in the question's words, so every branch visibly matters to it. A recited framework has branches in the framework's words, and the interviewer sees the mismatch. For "Should the chain enter a new city?", branches such as "the market is big enough", "we can serve it profitably" and "we can win against rivals" come from the question.

Two to four branches per level is enough. Barbara Minto's principle is that thinking is easy to follow when ideas are "organized as a pyramid under a single point" (Minto, read October 2026); a tree is that pyramid before you fill it. What you now know: the tree is the framework, and it has your words on it.

Step 3: Borrow a named lens only where it fits

Named frameworks are lenses for checking and filling a tree. Three recur, and each fits one kind of question. The profit equation, profit equals revenue minus cost, with revenue as price times volume and cost as fixed plus variable, fits any question about profit. Customers, company and competitors fits an entry or launch question. Porter's five forces name five pressures on an industry: established rivals, customers, suppliers, potential entrants and substitute products, and Porter wrote that awareness of them helps a company find a more profitable, less vulnerable position (Porter, Harvard Business Review, January 2008).

Say aloud why the lens fits and when it would not: the five forces describe an industry, so they say little about a single firm's internal cost problem. If a lens does not fit, drop it. What you now know: a lens earns its place by fitting the question, not by being famous.

Step 4: Test one branch with a hypothesis and numbers

A tree is not an answer until one branch is tested. State a hypothesis, name the branch it points to and ask for the data that would confirm or kill it. Then do the arithmetic aloud with round numbers; the case math drills page covers the habit.

Finish answer first: give the recommendation, then the two or three reasons, then the risks, as in the pyramid. If the data kills the hypothesis, say so and move to the next branch; changing your mind on evidence is the skill being tested. A brief summary at the end ties the branches back to the goal sentence from step 1. What you now know: the structure earns its keep by telling you what to compute next.

1 · Start from the question2 · Build a tree3 · Borrow a lens4 · Test one branchReplay

Step-by-step description

1. Restate the goal and clarify before structuring.
2. Branches that do not overlap and cover the problem.
3. Use a named framework only where it fits.
4. A hypothesis, numbers, then the answer first.

• Start from the question, restate the objective and clarify before choosing any structure.
• Build a tree whose branches do not overlap and together cover the problem.
• Borrow a named lens such as the profit equation or the five forces only where it fits the question.
• Test one branch first with a hypothesis and numbers, and give the answer first.

Worked example

The prompt, in our own words: "A specialty coffee roaster is considering entering a new city. Should it?" Every number is invented and in currency units, not a real currency. The tree has three branches borrowed from customers, company and competitors: is the market big enough, can we serve it profitably, and can we win share against rivals?

Take the first two. The market is 2,000,000 bags a year. The roaster sells at a price of 12.00 per bag with a variable cost of 7.00, so each bag contributes 5.00. Fixed cost for the new city is 600,000 a year. Break-even volume is 600,000 divided by 5.00, which is 120,000 bags, or a 6.0 percent share of the market.¹

Now the third branch becomes a number to test. If the roaster expects 3 percent, it sells 60,000 bags, earns 300,000 in contribution and loses 300,000 after fixed cost.¹ At 2 percent the loss is 400,000 and at 4 percent it is 200,000. At a 3 percent share, the price would have to be 17.00 for break-even, which is 41.7 percent above the current price.¹ The recommendation is to enter only if the roaster can credibly win about 6 percent, or cut the fixed cost.

Text description of the diagram

• The break-even volume is 120,000 bags, which is a six percent share of a two-million-bag market.
• At a three percent share the entry sells 60,000 bags and loses 300,000.

Trade-offs and edge cases

No structure fits every question, which is why the published advice is to build and adapt. The table says where each common one fits and where it breaks.

The first edge case is the interviewer-led case, where the interviewer asks a series of questions and so picks the branch for you; a clear opening structure still matters, because it shows how you would have approached the whole problem. The second is a twist: the numbers contradict the tree, such as a mix shift hiding in an average, and the right move is to say what the data changed and redraw a branch.

The third is time. As our own advice, not a published rule, keep the structure short enough to say in under a minute, so the interviewer can react before you lose them; BCG's don't list includes "Overcomplicate Solutions" (BCG careers, read October 2026). The cited Bain and BCG pages give no rule on how much a candidate should lead, so follow the interviewer's cues.

The tree below shows the worked example's structure, with the lens each branch borrows and the number that tests the economics branch.

Text description of the diagram

• The question is whether a specialty coffee roaster should enter a new city, with invented numbers.
• The three branches are market size, economics and rivals, borrowed from customers, company and competitors.
• The economics branch is tested first: break-even is 120,000 bags, a 6 percent share of 2,000,000 bags.

For the wider loop, see the interview loop guide; case-style estimation is in market sizing, and finance rounds use similar logic in the investment banking guide.

How to answer it in an interview

State the goal, build the tree out loud in two or three branches, name the lens that fits and why, then pick one branch to test. The model answer below is illustrative.

Say it in this order: goal, branches, the lens and why, the hypothesis and the first test. It takes about 45 seconds.

Common mistakes

1. Reciting a framework. (wrong) "I will use the four Ps." (right) "The question is about entry, so I will check market, economics and competition."
2. Overlapping branches. (wrong) "Costs, expenses and spending." (right) "Fixed cost and variable cost: each cost sits in exactly one."
3. Forcing a lens. (wrong) "Let me run the five forces on this internal cost problem." (right) "The five forces describe the industry, so I will use a cost tree instead."
4. A tree with no test. (wrong) "Here are six branches." (right) "My hypothesis points to unit cost, so I will test that branch first."
5. Hiding the answer at the end. (wrong) A long walk-through and then a verdict. (right) "My recommendation is X, for these three reasons."

Follow-up questions

What are case interview frameworks?

Case interview frameworks are starting structures for a business question, such as splitting profit into revenue and cost, that you adapt to the case instead of reciting. Bain's recruiting specialist says the firm evaluates "how you think, not the framework you memorized", so treat a framework as a first draft of the tree, as on the case interview hub.

What are the common case interview frameworks?

Three recur: the profit equation for profit questions, customers-company-competitors for entry and launch, and Porter's five forces for how attractive an industry is. Step 3 shows where each fits. A custom tree built from the question covers everything else; product candidates use the same move in product management interviews.

Is there a best case interview framework?

Not one. The best structure is the one whose branches use the question's own words and do not overlap, with a named lens borrowed only where it fits. Bain's tips page asks for "a clear framework", and its recruiting specialist warns against force-fitting a pre-learned one, so the fit matters more than the name.

What if my framework does not fit the case?

Say so, and change it. Name what the first structure missed, redraw the branches in the question's words and carry on. A visible, reasoned change reads as judgment; defending a poor fit does not. For the non-case half of a consulting loop, see tell me about yourself and the behavioral interview hub.

What if the fixed cost is halved?

Break-even moves with it. Fixed cost of 300,000 instead of 600,000 gives a break-even of 60,000 bags, a 3.0 percent share of the 2,000,000-bag market, so the 3 percent case that lost 300,000 now breaks even.¹ Say which branch changed and redo the arithmetic aloud, as in the case math drills.

Sources

1. Case interview preparation (Bain & Company careers) — Bain & Company (no page date shown; date = read date), 24 Oct 2026
2. How to stand out in Bain's consulting recruiting process — Bain & Company (recruiting specialist, University Hiring), 11 Mar 2026
3. Case interview preparation (BCG careers) — Boston Consulting Group (no page date shown; date = read date), 24 Oct 2026
4. The Five Competitive Forces That Shape Strategy (Harvard Business Review) — Harvard Business Review (Michael E. Porter), 1 Jan 2008
5. The Pyramid Principle: concept (Barbara Minto) — Minto International (date = read date), 24 Oct 2026