I stop listening when the first sentence of a structure is a brand name. "I'd use the 4Cs" tells me you optimized for sounding prepared, not for answering the client's question. The score sheet rewards a structure that fits this prompt; it does not reward fluent recall of a template.
Bain's university hiring team says the common mistake out loud: force-fitting a pre-learned framework instead of engaging the problem in front of you. McKinsey's interviewing page frames the case as how you structure ambiguous challenges and identify important issues. BCG says there is not always a single right answer; what matters is the quality of your reasoning. None of those pages ask you to name Porter.
After this piece you should be able to hear your own openings the way an interviewer hears them, catch the six failure modes before they leave your mouth, and rebuild one weak structure into a prioritized tree you would actually use on a live case.
What the sheet rewards (and what it ignores)
Firms publish evaluation language, not confidential rubrics. Treat the table below as what they say they listen for, not as an insider scorecard.
| Firm (public) | Language that maps to structure | What that is not |
|---|---|---|
| McKinsey interviewing | Structure tough, ambiguous challenges; identify important issues; deal with facts and data; formulate recommendations | Reciting a named framework |
| BCG case prep | Structure your approach; ask thoughtful questions; identify the most important factors; quality of reasoning | A single "right" answer or laundry list |
| Bain recruiting tip | Pause, structure, talk through logic; tailor to context; hypothesize what matters most | Force-fitting a textbook structure |
Ned's estimate: in a first-round slate of eight cases, two or three openings are already dead on arrival because the structure never touched the decision. The rest usually fail later on math or synthesis. Structure is the earliest filter.
Mistake 1: Name the template before you name the decision
Before (what I hear):
"I'd approach this with Porter's Five Forces. First, threat of new entrants…"
After (what scores):
"The CEO needs a go / no-go on launching a mid-price meal kit in the Nordics. I'd split that into demand proof, unit economics at their cost structure, competitive response, and the capabilities they still need to build. I'd start with demand because if households will not reorder, the rest is irrelevant."
McKinsey's problem-solving interview is built around a client scenario. Naming Five Forces signals you reached for industry-attractiveness scaffolding when the decision was a product launch. The after version states the decision, names branches that could kill the recommendation, and picks a first test. That is what "identify important issues" looks like in the room.
Mistake 2: Branches that cannot change the answer
A structure is only useful if a finding on a branch would change what you tell the client. Decorative buckets fail that test.
Before:
"I'd look at company, customers, competitors, and context."
After:
"I'd test three killers: (1) will Nordic households pay enough after shipping and spoilage to clear a 12% contribution margin; (2) can they hit 40% reorder in month three with their current recipe pipeline; (3) does a local incumbent already own the cold-chain slots they need. If any one fails, I recommend delay."
The 3Cs can be a private thinking prompt. Spoken aloud as four equal nouns, they are not a plan. Bain's FashionCo sample case asks you to start with the question you are solving, then propose where you would start. That is the same standard: branches must be testable against the objective.
Mistake 3: Overlap that burns the clock
Overlapping branches make the interviewer do the MECE work for you. They will not.
Before:
"I'd look at demand, customer volume, sales, and market size."
After:
"I'd separate volume from price and mix. Volume: households ordering at least once. Price: average order value net of discounts. Mix: share of premium recipes. Market size sits above that as a check, not a fourth peer branch."
Demand / volume / sales / market size are four labels for one idea. In a twenty-five-minute case you just spent the clarifying pause inventing duplicate work. Clean branches let you point at one number and move.
Mistake 4: Equal-weight laundry list
BCG's public prep language stresses identifying the most important factors. Equal weight is the opposite of that.
Before:
"I'd go through market, customers, operations, finance, competitors, product, regulation, and execution one by one."
After:
"Eight areas matter in theory. For this Nordic launch I'd prioritize demand and unit economics first, competitive response second, and capability gaps third. Regulation only if cold-chain food rules create a hard gate; I would not spend the opening minute there."
Worked priority check (reproducible): suppose you have twelve minutes of analysis after the structure. Eight equal branches get ninety seconds each: nothing deepens. Three prioritized branches get four minutes each: enough for a real exhibit or a math path. Same clock, different score.
Mistake 5: Wrong problem type, correct-sounding labels
Profitability math on a pure strategy question, or Five Forces on a one-site cost spike, is the same error: you matched a keyword, not the decision.
Prompt for the transcripts below: HarborLine operates six regional ferry routes. Operating profit fell 18% last year. Management wants to know why and what to do in the next twelve months.
Before:
"I'd use a market-entry framework: market attractiveness, competitive intensity, and our ability to win."
After:
"This is a diagnostic profit problem with a near-term fix lens. Profit is revenue minus cost. On revenue I'd split traffic (passengers and vehicles) from yield (ticket mix and ancillaries). On cost I'd split fuel and crewing from port fees and maintenance. I'd start with traffic by route because an 18% profit drop with six routes usually concentrates in one or two lanes."
Market entry is the wrong problem type. You can still use revenue-minus-cost as the root and speak in ferry language. That is the difference between knowing a label and fitting the question.
Mistake 6: Treating the opening structure as a contract
BCG asks you to listen actively and think structurally as the case moves. Bain's tip is to listen and adapt. Freezing on the first tree is how you fail both.
Before (after new data: "Route 4 volume is flat; fuel cost per trip is up 22%"):
"Okay, so continuing with my market and customer analysis…"
After:
"That finding collapses my traffic-first hypothesis for Route 4. I'd retire volume as the lead branch for that lane and open a fuel-and-routing branch: contract vs spot fuel, speed profile, and whether schedule density forces inefficient trips. The other five routes still get a traffic check."
Recovery is visible. Stubbornness is also visible. Say which branch died, which one you open, and why the objective still holds.
Ned's rule. If your opening structure would still be correct after any exhibit the interviewer might show, it is too generic to score. A useful tree has at least one branch that new data can kill.
Score a weak vs strong opening (same prompt, same clock)
Use the HarborLine prompt. Give yourself ninety seconds, then score yourself the way I would.
| Signal | Weak opening (0–1) | Strong opening (2) |
|---|---|---|
| Decision stated | "Look at profits" | "Explain the 18% drop and pick a twelve-month fix" |
| Fit to case type | Market-entry or 4Cs | Revenue / cost with ferry drivers |
| Branch overlap | Traffic and "demand" and "sales" | Traffic vs yield; fuel/crew vs fees/maintenance |
| Priority | "Go through each equally" | Lead with route-level traffic, then fuel |
| Adaptation ready | Tree is a finished product | Named what would kill the lead branch |
Ned's scoring heuristic: if three of the five signals are at 0, I am already writing a soft no before the first chart. You do not need a firm’s confidential rubric to practice this. You need honest self-grading against fit, not against how many frameworks you remembered.
Fix one opening out loud today
Take any case prompt you already know. Record a sixty-second opening twice: once as a named framework, once as decision-plus-prioritized-branches. Play them back. The second should sound like advice to a CEO, not like a chapter title.
Then run a scored structure rep on CoachNed so someone other than you grades the fit:
Structure drill — write three to four branches for a messy prompt and get instant feedback on whether they are tailored or recited. Or take the five-minute first rep (three typed turns, no account) and listen for whether your first structure names a decision. Everything is open for seven days, no card; then $120 for a recruiting season or $49 a month.
If you want the full live pressure version after that, the voice case with Ned (~18 minutes, seven-score debrief) is where freezing after new data shows up fastest.
CoachNed is independent and not affiliated with McKinsey, BCG, Bain, Deloitte, or any firm named above.
Frequently asked questions
Is it bad to use a profitability framework at all?
No. Using revenue minus cost as the root of a profit case is fine. The mistake is stopping at "price, volume, fixed, variable" without translating those into the client's business language, or using that root when the decision is not about explaining a profit drop.
Should I ever name Porter's Five Forces or the 4Ps in the interview?
I would not. Describe the branches in the client's words. If industry structure truly matters, say "I'd look at how suppliers and substitutes constrain pricing" without the trademarked label. Naming the textbook rarely helps and often signals force-fit.
How many frameworks should I memorize before first rounds?
Fewer than prep folklore suggests. Build fluency with a small set of building blocks (profit trees, demand vs supply, make-vs-buy, capability gaps) so you can recombine them. Depth on six reusable pieces beats shallow recall of twenty named frameworks. Firms' own pages emphasize tailored thinking, not inventory size.
What if I realize mid-case that my structure was wrong?
Say so in one sentence, rename the lead branch, and move. Interviewers listen for adaptation; Bain and BCG both say so publicly. Pretending the first tree still fits after contradictory data is worse than a clean correction.
Do BCG chatbot cases punish framework mistakes the same way?
The format differs, but the standard does not: structure still needs to fit the question and prioritize what matters. If you are prepping that channel specifically, use a dedicated Casey walkthrough rather than treating chatbot clicks as a substitute for spoken openings.
Sources
- Interviewing at McKinsey — problem-solving interview criteria: structure ambiguous challenges, identify important issues, use facts and data, recommend. Checked 2026-09-24.
- BCG Case Interview Preparation — public expectation to structure an approach, identify important factors, and be judged on reasoning quality. Checked 2026-09-24.
- How to stand out in Bain's consulting recruiting process — recruiter language against force-fitting frameworks; evaluate how you think, not the framework memorized. Checked 2026-09-24.
- Bain FashionCo sample case — start from the question, structure, then propose where to start. Checked 2026-09-24.
- Preparing for the Case Interview | Bain — clarify objective, structure thinking, listen and adapt. Checked 2026-09-24.