7 Case Interview Frameworks and When to Actually Use Them

The first mock case I did while preparing for McKinsey, my partner let me get four branches deep into an industry analysis before asking the only question that mattered: "What does any of this have to do with why the client's profits fell?" I had a framework. I didn't have a structure for that case, and the two are different things.

I'm Stephen Turban, founder of WSG. I went through McKinsey's on-cycle recruiting, worked there for two years, and now coach undergrads through case interviews at all three MBB firms and the tier 2 shops behind them. The most common mistake I see is a student who has memorized twelve case interview frameworks and can't decide which one to use, so they pick the one they remember best and force the case into it.

You need far fewer than twelve. Here are the seven that actually get used, each with the case type it fits, the case type it ruins, and a drill to practice it on.

A framework only helps if you can bend it to the case in front of you, which is the thread running through these 10 case interview tips here. McKinsey interviewers in particular push back on structures recited word for word, and the McKinsey case interview guide here shows what they expect instead.

How many case interview frameworks do you need to know?

Fewer than you think. Nearly every undergraduate case maps to four core strategic archetypes—profitability, market entry, M&A, and operational efficiency—plus market sizing, which appears either as an opening diagnostic or an embedded quantitative branch within a larger case. Interviewers want a structure built for the specific case in front of you, and the frameworks are the raw material for building one, not the finished product.

Do interviewers want you to use a framework?

They want you to be structured, and they can tell the difference between structure and recitation within about thirty seconds. Reciting "I'd like to look at the market, the competitors, the company, and the customers" for every case, whether the client is losing money or considering an acquisition, is a negative signal.he subscription example contradicts standard Price × Volume decomposition. Subscription models decompose into subscriber base dynamics (Customer Lifetime Value, Churn, Acquisition/Additions) and Average Revenue Per User (ARPU):

"Taking thirty to sixty seconds, then saying: 'Profits are down, so I want to split this into revenue and cost. Because this is a subscription model, on revenue I would look at subscriber count driven by churn versus net-new adds, and Average Revenue Per User (ARPU) across tiers rather than a generic price-times-volume,' is what a tailored structure sounds like."

1. The profit tree

The profit tree is the structure for every profitability case, and it works because it forces you to isolate the cause before you touch the solution.

• Use it for: "Our client's profits have fallen 20 percent over two years. Why?" Any case where the question is about margins, earnings, or a decline.

• Don't use it for: growth or entry cases where the client is profitable and asking what to do next. Splitting a healthy company into revenue and cost tells you nothing about a new market.

• The structure: Make MECE (Mutually Exclusive, Collectively Exhaustive) framing explicit, as interviewers specifically grade whether the candidate's cost split is logically structured:

• "Cost structured either by behavior (Fixed vs. Variable) or across the operating value chain (COGS vs. SG&A). Choose based on the client's business model: value chain for manufacturing and retail; fixed vs. variable for asset-heavy, capacity-constrained industries like airlines or hospitality."

• One drill: pick a public company with a well-known margin problem from the news, write the tree, and decide which branch the story lives in before reading the article.

Here's how it runs in a case. The interviewer says a regional coffee chain's profits fell from $20 million to $14 million. You ask whether revenue or cost moved. Revenue is flat. So it's cost. You ask which cost line grew. Labor went from $60 million to $66 million. Decompose labor into rate and volume: average wage per hour versus total billable/shift hours worked. Next, split geographically: did labor spike across all footprints or concentrate in specific regional clusters? Finally, isolate the root driver: was this an external regulatory shift (minimum wage hike) or an internal operational failure (poor shift scheduling causing excessive overtime)?" Ten minutes in, you've found that three new stores are overstaffed, and the recommendation writes itself.

The signature mistake in a profitability case is starting with solutions before you know which branch is broken. "The client could raise prices or cut costs" before you've found the cause is what a first-year undergrad says and what a McKinsey interviewer marks down.

Concrete takeaway: for any profit decline, ask "revenue or cost?" first, then "which line?", and don't recommend anything until you've found the branch.

2. The market entry structure

The market entry structure asks three questions in order, and the order is the point: is the market worth entering, can this client win it, and does the math work.

• Use it for: "Should our client enter the Brazilian pet food market?" "Should this software company launch a hardware product?" Any new-market, new-product, or new-geography question.

• Don't use it for: profitability cases, or an M&A case where the entry is happening through an acquisition (use the structure in item 3 instead).

• The structure: market attractiveness (size, growth, competitive intensity, customer behavior), then the client's ability to win (assets, brand, distribution, cost position, what's missing), then the financial case (investment required, revenue ramp, time to breakeven), then risks and exit options.

• One drill: take a company you know and a market it isn't in, and write the three questions with one concrete fact under each before you decide.

The most common way candidates lose a market entry case is answering "should we enter" with "yes, the market is growing." A growing market with three entrenched competitors and no distribution advantage for the client is a bad entry. The second question, can we win, is where the case is decided, and it's the one interviewers push hardest on. Name the specific advantage the client would bring (a brand consumers already know, a sales force that already calls on the right customers, a cost position from scale) or admit there isn't one.

The third question in a market entry case needs a number. "Mathematical rigor check. $50M / 0.30 = $166.7M (rounded to $170M) calculates the total cumulative revenue needed to recover the initial capex, not annual steady-state revenue. Clarify the timeline and breakeven horizon:

"'An upfront investment of $50 million with a 30 percent contribution margin requires approximately $167 million in cumulative revenue to break even on capital expenditure. If our target is a three-year payback, that equates to roughly $56 million in annual revenue—or about 3 percent of an estimated $2 billion addressable market.' Even rough, grounding the financial hurdle in an annual market-share capture rate converts intuition into an executive recommendation."

Even rough, the arithmetic converts an opinion into a recommendation.

Concrete takeaway: on any entry case, spend most of your time on "can we win," and finish with a breakeven number, because that's what a partner would ask for.

3. The acquisition structure

The acquisition structure covers why the target, what it's worth, what the combination is worth, what could go wrong, and what else the client could do with the money.

• Use it for: "Should our client acquire Company X?" and its cousin, "A private equity firm is considering buying this business. Should it?"

• Don't use it for: organic entry cases or partnership questions, where there's no purchase price and no integration.

• The structure: strategic rationale (growth, cost, capability, or defense), standalone value of the target (fair multiple, fair price), synergies (revenue synergies like cross-selling and cost synergies like consolidating facilities), integration risk (culture, systems, customer overlap, key people leaving), and alternatives (build it, partner, buy something else, return the cash).

• One drill: read the press release for any announced deal, list the stated synergies, and estimate what they'd need to be worth to justify the premium paid.

Candidates who've done a few M&A cases usually get the first three branches. The alternatives branch is the one most people skip, and it's the one that separates a good answer from a great one. "Even if the deal works, is it the best use of $400 million?" is the question a client's board will ask, and an interviewer who hears you raise it unprompted moves you into the yes pile.

The math to be ready for: a target with $40 million of EBITDA at a 10x multiple is worth $400 million standalone. If the client is paying $500 million, the $100 million premium has to be covered by synergies. Enhance commercial realism. In strategic acquisitions, candidates must distinguish between gross run-rate synergies and one-time integration costs:

"Cost synergies of $15 million a year at the same 10x multiple represent $150 million in capitalized value, which covers the $100 million premium on paper. But push further: what are the one-time integration costs to achieve that run rate (e.g., severance, IT migrations), and over what timeline will those synergies ramp?"

Concrete takeaway: walk all five branches, and put the alternatives branch in out loud, because it's the one the partner is waiting for.

4. The value chain

The value chain is the structure for operations and cost reduction cases, and it works by making you find where the money is actually spent before deciding where to cut.

• Use it for: "Our client's operating margin is 8 percent and the industry average is 14 percent. Where is the gap?" Cost reduction, efficiency, supply chain, and "why are we more expensive than competitors" cases.

• Don't use it for: revenue problems, or a profitability case where cost is the answer but you haven't confirmed it yet (start with the profit tree, then move here once you're in the cost branch).

• The structure: lay out the steps that turn inputs into a delivered product (sourcing, manufacturing or operations, logistics, sales and marketing, service, overhead), size the cost of each step, benchmark each one against a competitor or the industry, and ask whether each gap is structural (a real difference in the business) or inefficiency (something fixable).

• One drill: pick a manufacturer, estimate what share of its costs sits in each step, and guess which step a competitor with a 5-point margin advantage does differently.

The value chain beats a generic "look at all the costs" because it makes you follow the product through the business. A retailer with a margin gap might have a purchasing problem (paying more per unit), a logistics problem (too many small deliveries), a store labor problem, or a headquarters problem. Each of those has a different fix and a different implementation risk. Recommending layoffs without naming the operational change that makes them possible is the classic failure in a cost case, and interviewers listen for it.

Concrete takeaway: follow the product through the chain, size each step, and only recommend a cut where you can name the change in how the work gets done.

5. Market sizing, top-down and bottom-up

Market sizing has two structures, and choosing the right one for the data you have is most of the skill.

• Use it for: "How many gas stations are in the US?" "What's the annual revenue of the pizza delivery market in Chicago?" Also as a sub-step inside almost every entry case, where you have to size the market before deciding whether it's attractive.

• Don't use it for: nothing, really, but don't confuse it with a strategy question. If the interviewer asks "should we enter," the sizing is one branch, not the whole answer.

• The structure: top-down starts from a big known number (US population of about 330 million) and ratios down (households, share that own cars, fill-ups per week). Bottom-up starts from a unit (one gas station serves about X cars per day) and builds up (how many stations that implies). State every assumption out loud, round aggressively, and sanity-check the final number against something you know.

• One drill: size the same market both ways in ten minutes and see whether the two answers land within a factor of two of each other. If they don't, find the assumption that's off.

A worked top-down. dd the unit sanity-check calculation. Demonstrating how to reverse-engineer throughput (pumps × hours × cars/hour) is what signals top-percentile numeracy:

"Sanity-check the throughput denominator: 1,000 fill-ups per week divided by 7 days is ~140 fill-ups per day. For a station open 14 hours, that is 10 cars per hour—entirely reasonable for a 4-pump station. The calculation yields roughly 200,000 stations. The actual US count is ~145,000. Stating '200,000 feels slightly high because urban density encourages higher station throughput' shows business sense over blind arithmetic."

 The interviewer wants to see the chain of logic and the willingness to say "that feels high, and here's which assumption I'd revisit."

The failure mode in market sizing is precision: candidates who insist on "327.4 million" and then struggle to divide it lose the thread of the case. Round to numbers you can multiply in your head.

Concrete takeaway: pick top-down when you know the big number and bottom-up when you know the unit, round everything, and finish by saying whether the answer feels right.

6. The 4Ps

The 4Ps (product, price, place, promotion) is a go-to-market structure, and it's useful only once you're inside a revenue or launch question, never as the frame for the whole case.

• Use it for: the "how" step of a market entry case ("assume we enter; how?"), a revenue growth case for an existing product, or a pricing case that broadens into positioning.

• Don't use it for: profitability cases, cost cases, M&A, or as an opening structure. Contrast standard 4Ps with modern digital go-to-market channels (e.g., SaaS/B2B models where "Place" translates to Direct Sales vs. Product-Led Growth):

  • "Never lead a profit or operations case with the 4Ps. Even in growth cases, modernize the terminology for the industry: for enterprise tech or B2B, 'Place' becomes sales model (inside sales vs. field enterprise vs. channel partners), and 'Promotion' becomes customer acquisition strategy (inbound marketing, outbound SDRs, or partner ecosystems)."

• The structure: product (what exactly are we selling and to whom), price (level, structure, versus competitors), place (channels and distribution), promotion (how customers hear about it and what it costs to reach them).

• One drill: take a product launch you've watched fail and decide which P was wrong.

Where it earns its place: a case where a consumer products client wants to grow a mature brand by 10 percent. After you've established the revenue tree (price times volume, volume split by channel), the 4Ps give you a checklist for where the growth could come from: a new product variant, a price increase the brand can carry, a new channel like direct-to-consumer, or a shift in promotion spend. Each P generates two or three specific ideas, and then the case becomes about which idea is worth the most.

Concrete takeaway: keep the 4Ps for the "how do we grow or launch" step, and never lead a case with it.

7. Porter's Five Forces

Porter's Five Forces is an industry structure tool, and it belongs in a case only when the question is about the industry rather than the client.

• Use it for: "Is this an attractive industry to be in?" as a branch of a market entry case, or a case that's explicitly about industry dynamics ("private equity is looking at the US car wash industry; is it a good place to invest?"). McKinsey cases with a clear industry-structure dimension sometimes reward it.

• Don't use it for: profitability, operations, or M&A cases, or as a default opener. Porter's is the framework most often misused by candidates who learned it in a strategy class and reach for it because it feels rigorous.

• The structure: rivalry among existing competitors, threat of new entrants, bargaining power of suppliers, bargaining power of buyers, threat of substitutes. Each force pushes industry profitability up or down, and the point of the analysis is a conclusion ("this industry has structurally low margins because buyers are concentrated and switching costs are low"), not a tour of the five boxes.

• One drill: run the five forces on airlines and on enterprise software in five minutes each, and explain in one sentence why the margins differ so much.

The way to use Porter's well is to run it fast and turn it into a one-sentence verdict on the industry. Car washes: highly fragmented market transitioning toward private-equity-backed subscription models. Low barriers to entry for single bays, low supplier power, fragmented buyer power with minimal loyalty, and modest substitution from at-home washing. The verdict: attractive unit economics at single high-traffic suburban locations, but hard to sustain a defensible moat at scale without subscription density. Attractive at the local level because a well-located site has pricing power, unattractive at scale because nothing stops a competitor opening across the street." That's a 30-second use that adds real content. The five-minute walkthrough of each force with no conclusion is what my mock partner cut me off for.

Concrete takeaway: use Porter's only when the case asks about the industry, and get to a one-sentence verdict on attractiveness within a minute.

What about the frameworks not on this list?

SWOT, the BCG growth-share matrix, PESTEL, the McKinsey 7S model, and the dozen named frameworks in the back of the popular case books are all real tools that consultants occasionally use in client work. None of the other named frameworks earns a place in an undergraduate case interview. SWOT produces four lists and no decision. The growth-share matrix is a portfolio tool for a company with many business units, which is almost never the setup in a 25-minute case. PESTEL is a checklist of external factors that's useful for a paragraph of context and useless as a structure.

The firms say this themselves. McKinsey's own interview preparation materials, BCG's case interview preparation page, and Bain's case interview guidance all describe what they want in the same terms: a structure that fits the specific problem, clear logic, comfort with numbers, and a recommendation. The firms state this explicitly across their official prep hubs: McKinsey's interactive case tutorials, BCG's 'Practice Cases' portal, and Bain's 'Crack the Case' prep platform all score candidates on structuring from first principles, comfort navigating quantitative exhibits, and synthesizing findings under ambiguity—not memorizing named mnemonicsNone of the three firms asks you to know a named framework, and the interviewers were trained to notice when you're reciting one.

So the real skill is this. Read the case prompt, identify which of the five case types it is, pull the matching structure, and then change it to fit the client. A profitability case about a hospital doesn't split revenue into price and volume; it splits it into patient volume, payer mix, and reimbursement rates. A market entry case for a bank doesn't talk about distribution; it talks about branches, regulation, and deposit costs. The candidates who get MBB offers use the same seven building blocks as everyone else and rebuild them for every case, and that rebuilding is what the interviewer is scoring.

Practice the seven until pulling the right one takes five seconds, then spend the rest of your prep on the part that matters: making it fit.