The Profitability Case, Carried From Brief to Recommendation

You have a consulting case interview next week and you have read the frameworks, but you have never watched one case driven cleanly from the first sentence to a defended recommendation. This guide does exactly that: it carries a single declining-profits case through every fork, with the real segmentation, the arithmetic worked without a calculator, and two documented wrong turns and how the candidate backs out of each. It is for anyone preparing for an MBB or strategy interview who wants to see the shape of a solved case before running one themselves.

Profitability cases are worth learning this way because they are the workhorse of the round. On published prep data they make up roughly 30 to 40 percent of first-round MBB interviews, so one archetype driven end to end returns more than skimming ten unsolved prompts.

What sequence does the interviewer expect?

Every case moves through four phases in order: opening, structure, analysis, closing. Interviewers are grading you against that sequence whether or not they announce it, so knowing where you are in the flow keeps you from skipping a phase or lingering in one.

The phases are not equal in weight, and the last one carries more than candidates expect. The opening is where you clarify the brief. The structure phase is where you build and present a tree. The analysis phase is where you segment and compute. The closing is the final 60 seconds where you deliver a recommendation. A common and expensive belief is that the case ends when the math ends. It does not. The closing is where the interviewer decides whether they would trust you in front of a real client.

Format changes who drives. McKinsey-style cases differ from Bain and BCG in three ways: the interviewer leads the case, you have less freedom to explore, and you are assessed on each individual question rather than on overall performance. Bain and BCG are candidate-led, so you push the case forward yourself. The four-phase spine holds either way; only the amount of steering you supply changes.

Within that flow, four habits separate strong candidates. Ask at least one clarifying question before building your framework. Ask for 60 to 90 seconds of silence to structure before presenting. Talk through every calculation out loud. End your recommendation with two to three supporting reasons and at least one next step. The exact per-minute timing is not published by the firms; the 60-to-90-second figure is a coaching-page norm, so treat it as a target rather than a rule the interviewer is timing with a stopwatch.

The brief: what the client actually asked

Here is the case this guide carries all the way through. The client is CookieCo, a United States maker of packaged cookies. Its profit has been declining for two to three years while competitors are flat or growing. The interviewer asks: why are profits falling, and what should CookieCo do?

The first move is not a framework. It is to clarify the brief and pin down scope. A strong candidate repeats the client, the objective, and the scope back, then asks two or three clarifiers. In this case the candidate confirms two things that reshape the whole case: that scope is just the United States, and that focus is only the cookie segment even though the client sells other products. Both matter. If the candidate had assumed the whole company was in play, the arithmetic later would have pulled in irrelevant product lines.

Done looks like this: the candidate can state the objective in one sentence and name the metric. "We want to find why CookieCo's US packaged-cookie profit has fallen over two to three years and recommend how to reverse it." That sentence is the anchor for everything that follows.

Building the tree: profit, decomposed

The expected structure for any profitability case starts from one equation: Profit equals Revenue minus Costs. Revenue breaks into Price times Volume. Costs break into Fixed Cost plus Variable Cost. That is the spine, and reaching for it first is a measurable edge, not a stylistic preference.

One coaching provider's drill data found that candidates who open with an explicit Revenue-versus-Cost split and then segment score about 28 points higher on the structured-thinking rubric. The difference is not intelligence; it is the habit of reaching for the equation before exploring. The equation forces MECE coverage - mutually exclusive, collectively exhaustive - so you cannot leave a branch uncovered.

But the equation alone is not a structure yet. You have to segment it, in three levels. The first level is the Revenue-versus-Cost split. The second level is Price times Quantity times Mix on the revenue side and Fixed plus Variable on the cost side. The third level is segment splits by geography, customer type, product line, or channel. For CookieCo, geography is fixed at US and product is fixed at cookies, so the useful third-level splits are customer type and channel - for instance retail versus wholesale, or grocery versus club stores.

The tree you present must be customized. A textbook Profit tree with no customization reads as memorization, and McKinsey interviewers spot a generic framework immediately. Rename your buckets to CookieCo's business and add industry-specific sub-drivers: ingredient cost for the variable branch, plant overhead for fixed, promotional discounting under price. The test is simple. Could this exact tree apply to any company? If yes, it is memorized.

Take structuring time to build this. Ask for 60 to 90 seconds of silence, draw the tree on paper, then present it top-down. Some coaches say about 30 seconds is fine for a simple prompt, so sources disagree on the exact window - but the point holds: build silently, present cleanly.

Reaching for the equation first is a measurable edge, not a style choice, and interviewers can tell.

The analysis: isolate, segment, compute

Once the interviewer accepts the tree and points you to a branch, the analysis phase begins. The first fork is diagnostic: is the decline on the revenue side or the cost side, and are competitors facing the same pressure? Answering the competitor question early distinguishes an internal problem from market-wide pressure, and that distinction changes the entire recommendation.

For CookieCo, the interviewer confirms that revenue is roughly flat and competitors are stable or growing. That points the case at costs, or at something hiding inside flat revenue. This is the first fork where a candidate can go wrong, and we will return to it.

The candidate requests per-segment numbers and computes on paper. The rule of the analysis phase: state your approach out loud before you calculate. This prevents dead-ends and lets the interviewer redirect you before you spend two minutes on the wrong number. Find out whether units sold decreased or price per unit fell; if units dropped, analyze why, staying structured and using segmentation to reach the root cause.

The arithmetic, worked without a calculator

Say the interviewer gives the candidate this: CookieCo sells 100 million units a year. Two years ago the split was 60 million premium units at a 40-cent margin and 40 million value units at a 10-cent margin. Today total volume is unchanged at 100 million, but the split is 40 million premium and 60 million value.

The candidate states the approach first: "I will compute total contribution margin then and now, and compare." Then computes on paper.

Two years ago: 60 million times 0.40 equals 24 million dollars, plus 40 million times 0.10 equals 4 million dollars. Total is 28 million dollars.

Today: 40 million times 0.40 equals 16 million dollars, plus 60 million times 0.10 equals 6 million dollars. Total is 22 million dollars.

The decline is 6 million dollars, or about 21 percent, on flat total revenue. Volume did not fall. Price did not fall. The mix shifted toward the low-margin product. That is the answer the aggregate number was hiding.

Keep the arithmetic disciplined. Most calculations should take 30 to 90 seconds; if one runs past two minutes, you are likely overcomplicating it. The accuracy target is strict - about 95 percent, or 90 to 100 percent per step - and McKinsey expects answers to the ones place, so rounding 4.73 million dollars to "about 5 million" can cost points. When you do round inputs, keep it within plus or minus 10 percent. One arithmetic slip can end the interview, because firms treat quantitative accuracy as a proxy for client-facing reliability.

The procedure you can run yourself

Here is the full sequence, condensed so you can follow along on your own case. Each step names what to do and what done looks like.

The two wrong turns, and how the candidate recovers

Most candidates who fail a profitability case fail in one of a handful of documented ways. Two are worth walking through in full because they are the most cited and the easiest to fall into under pressure. Each has a false positive - the version that looks fine from the outside - and a check that catches it.

Wrong turn one: math before a plan

At the first fork, the interviewer hands the candidate the segment numbers. The candidate, eager to show speed, immediately starts multiplying. The arithmetic is fast and correct, but it computes total revenue, which the interviewer already said is flat. Two minutes gone on a number that was never in question. This is fast, correct arithmetic on the wrong number - the false positive that makes math-before-plan hard to catch. The candidate feels productive.

The recovery: the candidate notices the number confirms nothing new, stops, and says, "Let me back up and state my approach. Total revenue is flat, so the interesting question is what is happening inside it. I want to compare contribution margin by product segment across the two periods." Naming the approach out loud is the fix. Always state your approach before calculating; it prevents exactly this dead-end. The check is a single question: did you say what you were computing and why before you computed it?

Wrong turn two: staying at the aggregate level

The second wrong turn is subtler and more common. The candidate correctly finds that total revenue is flat and total volume is unchanged, and concludes there is no revenue problem. That conclusion is a confident, well-delivered mistake - the false positive here is a clean aggregate finding that sounds like an answer. Many candidates compute total revenue and cost but never segment by product, customer, or geography, and this candidate is about to recommend cost cuts for a problem that lives in the revenue mix.

The recovery comes from the interviewer's nudge, or from the candidate's own discipline: "Before I move to costs, let me segment revenue by product line, because flat totals can hide a mix shift." That is the move that surfaces the 6-million-dollar swing computed earlier. The check is to name the segment carrying the decline. If you cannot name it, you have not segmented.

The other documented traps are worth a table, because you should be able to scan for the one you are about to make.

Two of these deserve a note. The industry check is what separates an internal problem from market-wide pressure; an industry-wide decline calls for a very different answer than a company-specific one, so always ask whether competitors are declining too. And the summary trap is the one that undoes candidates at the finish line, which is why the close gets its own section.

Closing: the 60 seconds that get scored

The recommendation is the only part of the case that mirrors presenting to a real client, which is why interviewers weight it so heavily. A lot of candidates think the case ends when the math ends, but the final 60 seconds is where interviewers decide whether they would trust you in front of a client.

The dividing skill is synthesis, not summary. Summary is "what we talked about." Synthesis is "what it all means" - what, so what, now what. That shift, from replaying details to pulling meaning forward, is what interviewers are listening for. Most candidates freeze here: they ramble, hedge with "it depends," or repeat everything they just said.

A strong recommendation follows a clear rhythm. Start with your answer, back it up with two to three reasons, then point forward with a next step and a risk. Specificity beats clichés. Interviewers want practical recommendations, not "cut costs." Say "optimize procurement through supplier consolidation to reduce unit costs by 8 to 10 percent." For CookieCo, whose problem is a mix shift, the close is not about costs at all.

Recommendation: CookieCo should rebuild premium-product volume rather than cut costs.

Why: First, total revenue is flat but contribution margin fell 6 million dollars, about 21 percent, driven entirely by a shift from premium to value units. Second, competitors are flat or growing, so this is an internal mix problem, not market pressure. Third, the premium line still carries a 40-cent margin against 10 cents on value, so the lever is volume, not price.

Next step: Test why premium volume fell - is it distribution loss, a promotional own-goal, or a competitor's premium launch - starting with the top three retail accounts.

Risk: If the shift reflects a permanent consumer move to value, re-premiumizing will fail, so I would validate demand before investing in premium marketing.

Notice the first sentence contains the actual decision. That is the test for whether you are synthesizing or summarizing.

How much to practice, and where the competition is

Published prep pages converge on 30 to 50 full verbal cases for most candidates, though the ranges vary and one coach argues quality beats volume. A common breakdown is 5 to 10 solo cases, 20 to 30 with a partner, and 3 to 5 coached with a former consultant. Coached cases are weighted heavily because structuring and communication faults do not self-correct through repetition without feedback.

The dissent is worth taking seriously: 25 to 30 cases plus targeted math drills, error logging, and partner feedback typically produces better results than 80 cases done mechanically. Returns diminish past roughly 30. Active out-loud practice improves performance about three times faster than silent reading, and 15-minute debriefs improve it about twice as fast as skipping feedback. Volume without a debrief is the trap.

Where you interview shapes how thick the competition is. Consulting talent clusters in a few markets, and the concentration is visible in the profile data.

In Refolk's index of professional profiles, the US "Management Consultant" pool is about 3.9 times the size of Germany's, and the broad "Consultant" pool in the UK is only about 26 percent of the US pool. The US pool skews toward boutique and independent advisories, while the top employer surfacing in the German pool was a large firm. If you are interviewing where supply is thick, the marginal case you run better and the tighter close you deliver matter more, not less.

One efficient way to close the feedback gap the volume data points to is to practice with someone who has sat on the other side of the table. Finding a former consultant to run coached cases is exactly the kind of targeted search that Refolk handles from a plain description of who you want.

Before you walk in: a final check

Run this list against your own practice case the night before. Each item is something to verify, not a topic to think about.

To keep this sharp as the interview approaches, run one fresh profitability case out loud every day this week, record the close, and log every error you make into a short list you reread before the next attempt. The mechanism that improves you is not the count of cases; it is the debrief after each one. Treat the CookieCo walkthrough above as the reference you compare your own case against - where you segmented late, where the arithmetic drifted, where the close became a summary - and fix one fault per rep.