In a typical first-round slate of eight candidates, I fail two or three because they announce a brand-name framework before they have restated the question. The score sheet does not pay for "I will use the 3Cs." It pays for a tree that can isolate the driver, stay MECE, and die when the data kill the hypothesis.
That is the thesis of this guide: memorized frameworks are raw material, not answers. After you read it, you should be able to build a custom structure in about ninety seconds, know when to steal a branch from Porter or a profitability tree, and know when naming either would cost you the round.
McKinsey's own interviewing page frames the case as a "problem-solving interview" on a client scenario, not a quiz on named models. BCG says what matters most is how you approach the problem and the quality of your reasoning. Neither firm asks you to recite Ansoff.
What the sheet is actually scoring
When I score structure, I am listening for three moves in the first two minutes:
- Restate the decision or diagnosis in the client's words.
- Split the space so branches do not overlap and jointly cover what could move the answer.
- Bet which branch you will test first, and why.
If you do those three, I can follow you even when your tree is ugly. If you skip them and list Company, Customer, Competitor, I already know you are covering syllabus, not solving.
| Signal | Pass | Fail |
|---|---|---|
| Fit to prompt | Branches map to this metric or decision | Same four boxes on every case |
| MECE | No double-counting; no orphan gaps | "Market" and "customers" as two buckets |
| Hypothesis | "I start here because…" | Tour every branch "to be thorough" |
| Language | Client nouns (payers, SKUs, lanes) | Framework brand names |
A structure drill grades that fit under time. A framework encyclopedia does not.
Build the tree in four moves
Skip the mnemonic. Use this sequence every time.
Move 1 — Name the job. Most prompts are diagnose ("why did the number move?") or decide ("what should we do?"). Mix them and you get a laundry list. Design and allocate prompts exist too (offer design; portfolio capital), but if you cannot say which job you are in in ten seconds, stop and clarify.
Move 2 — Pick the driver metric. Profit, ROIC, share, utilization, contribution per order. Write the identity or definition that owns the number. For a P&L drop, that is usually revenue and cost, then price × volume × mix on the revenue side. For a growth choice, that is attractiveness, right to win, and mode. For deeper driver-tree practice, see the driver tree guide.
Move 3 — Split MECE. Mutually exclusive, collectively exhaustive. Two or three primary branches beat seven. If two branches can claim the same dollar, merge or redefine them. If a real industry lever is missing (payer mix in hospitals, load factor in airlines, churn in SaaS), add it.
Move 4 — State a hypothesis and a kill test. "I expect mix, because volume was guided flat; I want the bridge by segment first." Without a bet, you have a checklist. With a bet, the interviewer can hand you the exhibit that matters.
Ned's rule. If you cannot say which branch you would cut if you had only five minutes left, you do not have a structure. You have a syllabus outline.
Classic frameworks as raw material
These models were built for boardrooms and textbooks, not for a twenty-minute interview. Steal branches. Do not announce the brand.
| Tool | Origin (primary) | Steal when… | Failure mode in a case |
|---|---|---|---|
| Profitability tree | Algebraic identity (Profit = Revenue − Cost) | Diagnose a P&L or margin move | Listing every cost line without isolating one driver — see profitability cases |
| 3Cs (company, customer, competitor) | Ohmae's strategic triangle in The Mind of the Strategist (1982) | Competitive strategy or positioning | Opening a cost-out case with three vague buckets — see 3Cs as a tree |
| 4Ps (product, price, place, promotion) | E. Jerome McCarthy, Basic Marketing (1960), refining Borden's marketing-mix idea | Designing or fixing a go-to-market offer | Using it to "explain" last year's EBITDA |
| Porter's five forces | Michael Porter, HBR March–April 1979 | Industry attractiveness or structural profit pool | Touring all five forces on a single-client operational case |
| Ansoff matrix | H. Igor Ansoff, "Strategies for Diversification," HBR 1957 | Ordering growth options (penetrate, develop market/product, diversify) | Forcing four boxes before you know if any option clears a return hurdle |
| BCG growth-share matrix | BCG, late 1960s; Henderson's "Product Portfolio," 1970 | Multi-business capital allocation | Drawing stars and dogs for a one-product pricing prompt |
Porter's 1979 article is about industry structure and profit potential, not about why one hospital's EBITDA fell fourteen million. BCG's own history page is clear that the growth-share matrix helps prioritize businesses in a portfolio. Ansoff's four growth alternatives are a menu of strategic moves, not a diagnostic bridge. Use them where they earn their keep; mute them elsewhere.
Market entry, M&A, and PE diligence have their own trees. Link out rather than stuffing them here: market entry, M&A, PE due diligence.
Worked example: Cascade Pet Foods, profit down $9m
Prompt. Cascade Pet Foods makes private-label dry dog food for three national grocers. Operating profit fell from $28.0m to $19.0m in one year. The CEO wants the driver and one ninety-day fix. You have twenty minutes.
Weak open. "I'll use profitability, then the 4Ps, then Porter to check industry structure."
Strong open. "This is diagnose first, then decide. Operating profit is down $9m. I will split revenue and cost, dig revenue into price, volume, and mix by retailer, and I will start with mix because private-label contracts often reprice on SKU mix even when tons are flat."
Ask for the bridge:
| Line | Prior year | This year | $ impact vs prior |
|---|---|---|---|
| Tons shipped | 84,000 | 86,500 | Volume +$1.8m |
| Average net price / ton | $1,420 | $1,405 | Price −$1.3m |
| Mix: share of premium SKUs | 38% | 29% | Mix −$8.4m |
| COGS + freight | $78.1m | $79.2m | Cost −$1.1m |
| Operating profit | $28.0m | $19.0m | −$9.0m |
Driver isolated: SKU mix, not "the pet food market is more competitive" and not a five-forces tour. Premium kibbles lost nine points of mix to economy formulas at the largest retailer (42% of tons). Price per ton slipped slightly; cost barely moved; volume helped.
Hypothesis update. The retailer pushed a value reset; Cascade accepted SKU slots that dilute margin. Kill test for alternatives: if freight per ton were the story, the cost line would move more than $1.1m.
Decide on the mix branch only. Options that do not overlap:
- Renegotiate the value-reset contract: restore two premium facings or accept a volume floor with a mix floor.
- Cut list price on economy SKUs to defend tons (worsens mix math).
- Exit the economy line at that retailer and free capacity for the other two grocers' premium packs.
Option 2 fights the driver. Option 3 risks 42% of volume. Option 1 attacks mix with a commercial lever the CEO can run in ninety days.
Recommendation. Recover mix: trade a 3% volume concession at the large retailer for restoration of the two highest-margin SKUs to planogram, targeting +6 points of premium mix (about +$5.6m at current margins if volume holds). Risk: the retailer refuses and sources economy from a rival co-packer. Next step: contribution by SKU × facing for that banner, then a one-page walk-away.
No brand name required. A diagnostic identity, one isolated driver, a MECE solution tree on that driver.
When a canned tree is the wrong move
Math-first exhibits. If they hand you kiln energy dollars per tonne, compute. Structure after the so-what needs a decision.
Behavioral / PEI. Stories need an arc and ownership, not an issue tree.
Inventory of tools. Covering every framework you studied to show breadth reads as indecision. One mutated tree beats seven labels.
Announcing the name. "I'll apply Porter" on a pricing prompt is a meme and a low structure score. Say the branches in client language instead.
Practice this today
Close every framework cheat sheet. Set a ten-minute timer. Take three prompts from different industries (a margin drop, a market entry, a build-vs-buy). For each, write only: job type, metric identity, three MECE branches, and the first kill test. If a brand name appears on the page, cross it out and rename the branch.
Then run a timed structure drill so someone else scores whether the tree fits the prompt.
CoachNed's structure drill and the five-minute first rep at /start are open for seven days with no card; then $120 for a recruiting season or $49 a month. The live voice case at /interview is useful once your trees are short enough to say aloud.
Frequently asked questions
Should I memorize case interview frameworks?
Memorize the identities and the jobs (diagnose vs decide), not the brand names. Keep a mental scrapyard of useful branches from profitability, 3Cs, 4Ps, Porter, Ansoff, and BCG. Pull a branch when it fits; never open with the label.
What frameworks do McKinsey and BCG want?
Neither firm publishes a required list. McKinsey describes a problem-solving interview on a client scenario. BCG emphasizes structured approach and quality of reasoning. Both are scoring fit and logic, not recall of Porter or Ansoff.
Is the profitability framework enough for every case?
No. It is the right raw material for P&L and margin diagnoses. Market entry, M&A, pricing design, and portfolio allocation need different trees. Starting every case with revenue and cost is how you fail a "should we enter" prompt.
How long should my initial structure take?
Roughly sixty to ninety seconds of silent work, then a crisp walk-through of two or three branches and your first test. If you are still drawing after three minutes, you are decorating, not structuring.
Can I say "MECE" in the interview?
You can. Better: show it. Overlapping branches and gaps are obvious to the interviewer whether or not you pronounce the acronym.
Sources
- Interviewing at McKinsey — firm description of the problem-solving (case) interview. Checked 2026-09-24.
- BCG Case Interview Preparation — official guidance that approach and reasoning quality matter more than a single right answer. Checked 2026-09-24.
- How Competitive Forces Shape Strategy (Porter, HBR 1979) — primary origin of the five-forces industry-structure model. Checked 2026-09-24.
- What Is the Growth Share Matrix? (BCG) — BCG's account of the matrix's origin and portfolio use. Checked 2026-09-24.
- Basic Marketing (E. Jerome McCarthy, Google Books record) — 1960 managerial text associated with the 4Ps marketing mix. Checked 2026-09-24.
- Ohmae, Kenichi. The Mind of the Strategist (McGraw-Hill, 1982) — strategic triangle (company, customer, competitor). Book citation; no paywalled reprint linked. Checked 2026-09-24.
- Ansoff, H. Igor. "Strategies for Diversification," Harvard Business Review 35, no. 5 (1957) — four growth alternatives later popularized as the Ansoff matrix. Checked 2026-09-24.
CoachNed is not affiliated with McKinsey, BCG, Bain, Deloitte, or any firm named above.