In a typical first-round panel of eight profitability cases, I score two or three candidates who still have four live branches when they recommend. The algebra is not the hard part. Profit equals revenue minus cost; revenue is price, volume, and mix; cost is fixed plus variable. What the scoring sheet rewards is killing every branch but one, then fixing that driver.
If you leave the room with a tidy tree and a menu of initiatives, you drew a picture. If you leave with one driver and a recommendation that could not have been said in minute one, you did the case.
After this piece you should be able to open a profit case with a bridge, segment the tree until one line explains most of the delta, and catch the mix trap that makes the newspaper branch look right.
Why the tree is arithmetic, not a framework
Candidates treat "the profitability framework" as a brand. Interviewers treat it as bookkeeping. Draw Level 1 as revenue versus cost. Split revenue into price (or yield), volume, and mix. Split cost into variable (moves with units) and fixed (does not, in the case horizon). Stop when one child of one parent explains most of the year-over-year move.
| Level | Split | You are finished when… |
|---|---|---|
| Profit | Revenue − cost | One side owns most of Δ |
| Revenue | Price / volume / mix | One of these owns most of the revenue Δ |
| Cost | Fixed / variable | One line owns most of the cost Δ |
| Unit | Segment (channel, SKU, region, customer) | One segment owns most of the residual |
Naming the tree is optional. Isolating the driver is not. A partner will tolerate a messy whiteboard. They will not tolerate a recommendation that still hedges across three branches.
Real margin structures teach where to look first
Public filings show why "average margin" is a trap. The same company often runs two or three profit engines with different economics. Your first move in a case should be to ask which engine moved.
Costco's fiscal 2024 Form 10-K is the cleanest teaching example. Net sales were $249.6 billion against merchandise costs of $222.4 billion — a merchandise gross margin of about 10.9 percent. Membership fees were a separate $4.8 billion line. Costco itself states that it operates profitably at significantly lower merchandise gross margins than most retailers because volume, turnover, and membership fees do the work. In a case that looks like Costco, a 20-basis-point merchandise-margin miss and a flat membership-fee miss are not the same disease.
Starbucks' fiscal 2024 results show the other common pattern: store labor and occupancy sit on top of product costs. Full-year store operating expenses were 51.4 percent of company-operated store revenues, and GAAP operating margin contracted 130 basis points to 15.0 percent even as the company talked about pricing. North America comparable transactions fell while average ticket rose. Ticket up, traffic down is not "pricing worked." It is a mix-and-frequency story wearing a price label.
Southwest Airlines' 2024 Form 10-K is the fixed-versus-variable warning. Fuel and oil was $5.8 billion, about 21.4 percent of operating expenses, at $2.64 per gallon. Fuel fell year over year; CASM excluding fuel (CASM-X) still rose. The headline variable cost moved the wrong way to explain a profitability miss driven by non-fuel unit costs or yield. If the interviewer hands you a fuel spike and a profit hole, check whether fuel can even add up to the hole before you write a hedge memo.
Nike's fiscal 2024 results show a branded goods shape: gross margin of 44.6 percent and selling and administrative expense of $16.6 billion. COGS is mostly variable; a large share of SG&A behaves like fixed or semi-fixed in a one-year case. Volume down with SG&A flat is deleverage, not a "product cost" case.
Use these shapes as priors, not as answers. Ask which P&L the client actually runs, then bridge that P&L.
Bridge first, initiatives never
Ask for this year versus last year in dollars, not percentages. If the interviewer will not hand you a bridge, build one from the exhibits. Do not brainstorm twelve initiatives until the bridge exists. In the room I stop listening when a candidate jumps to "we should renegotiate suppliers" before they know whether cost moved at all.
A usable bridge answers three questions in order:
- Did revenue or cost own the profit delta?
- Inside the owning side, which Level-2 line owns most of it?
- Inside that line, which segment owns most of the residual?
If step 3 still leaves two equal segments, keep cutting. If it leaves one, you are done diagnosing.
Price, volume, and mix are three different diseases
Volume is units. Price is dollars per unit on a constant mix. Mix is a shift between units that earn different dollars per unit (or different contribution). Lump mix into "price" and you will recommend a list-price hike on a customer cohort you already lost, or celebrate load factor while contribution collapses.
| Move | What changed | Bad recommendation it tempts |
|---|---|---|
| Price | $/unit on same mix | Volume push when willingness to pay fell |
| Volume | Units on same mix | Price cut when demand is already soft |
| Mix | Share of high- vs low-margin units | Blanket discount or blanket cost-out |
Starbucks' North America pattern in late fiscal 2024 — fewer transactions, higher ticket — is the classroom version. Average price rose; the profit story may still be traffic. Costco's gasoline note in the same 10-K family is the other classroom version: higher gasoline penetration lowers reported gross-margin percentage even when the warehouses are healthy, because gasoline carries a lower merchandise margin. Mix moves the average without moving the disease.
Fixed versus variable decides which lever can even work
Variable costs scale with units in the case horizon: COGS, fuel, piece-rate labor, merchant fees. Fixed costs do not: rent, salaried overhead, aircraft ownership, much of brand SG&A. Semi-fixed costs step (add a shift, open a box, add a DC).
If volume fell and fixed cost stayed put, profit fell from deleverage. Cutting more variable cost will not refill the contribution you lost; you need volume, price, or mix back — or you need to cut fixed. If volume rose and variable cost rose faster than revenue, you have a unit-economics problem, not a "scale" problem.
Nike-shaped businesses deleverage when Direct or wholesale volume softens under sticky SG&A. Southwest-shaped businesses can show falling fuel and rising non-fuel CASM in the same year. Costco-shaped businesses can look "low margin" on merchandise while membership fees carry disproportionate profit. Classify the cost before you prescribe the cut.
Worked example: the obvious branch is wrong
Prompt. Meridian Home, a specialty kitchenware company: wholesale to department stores, a DTC site, and 18 outlet stores. Operating profit fell from $42 million to $19 million (−$23 million). Steel and ocean freight were in every trade magazine. The CEO wants a China dual-source program and a 5 percent list-price increase across wholesale.
You get this bridge ($ millions):
| Last year | This year | Δ | |
|---|---|---|---|
| Wholesale revenue | 210 | 198 | −12 |
| DTC revenue | 95 | 118 | +23 |
| Outlet revenue | 55 | 72 | +17 |
| Total revenue | 360 | 388 | +28 |
| COGS | 162 | 178 | −16 (cost up) |
| Variable fulfillment / returns | 28 | 41 | −13 |
| Fixed (SG&A, rent, brand) | 128 | 150 | −22 |
| Operating profit | 42 | 19 | −23 |
Revenue is up $28 million. This is not a demand collapse. COGS is up $16 million — painful, and still less than the $23 million profit hole once you notice fulfillment and fixed also moved. The newspaper branch (steel/freight → dual-source) cannot be the whole story. Kill it as the main play.
Contribution by channel (revenue − COGS − variable fulfillment; outlets share warehouse SG&A in fixed):
| Channel | LY contribution | TY contribution | Δ |
|---|---|---|---|
| Wholesale | 84 | 79 | −5 |
| DTC | 48 | 35 | −13 |
| Outlet | 38 | 26 | −12 |
| Total contribution | 170 | 140 | −30 |
| Fixed | 128 | 150 | −22 |
| Operating profit | 42 | 19 | −23 |
Wholesale is almost flat on contribution. DTC and outlet together lost $25 million of contribution while growing revenue. Fixed rose $22 million (new DTC marketing, three outlet openings, and a brand refresh the board already approved). The profit hole is DTC and outlet unit economics plus fixed spend that assumed those channels were healthy, not a company-wide COGS crisis.
Inside DTC. Orders +18 percent; return rate 12 percent → 21 percent; promotional share of sales 8 percent → 19 percent. Gross margin after returns and promos: 42 percent → 28 percent. The volume look healthy; the contribution does not.
Inside outlet. Share of company units that are closeouts rose from 14 percent to 26 percent because DTC overbought and wholesale cancelled. Outlet ASP fell 11 percent; contribution per outlet unit fell from $14 to $8. Outlets became a scrapyard for the other two channels.
Isolated driver: contribution collapse in DTC (returns and promos) and outlet (closeout mix), amplified by fixed cost added for growth that did not earn its keep. COGS inflation is real for about $16 million at the gross-profit line and does not explain the $23 million operating-profit miss once channel contribution and fixed are on the page. A network-wide 5 percent wholesale list-price hike fixes neither DTC returns nor outlet closeouts, and dual-sourcing is a second-year project for a first-year hole.
Recommendation. Freeze new outlet openings and the brand-refresh tranche that has not yet committed. Cap DTC promos at last year's share for 13 weeks and raise the return friction on free-return labels (the return rate is the contribution killer). Push excess inventory through a one-time wholesale off-price deal rather than the outlet fleet. Keep the dual-source work as a cost program with a 9–12 month clock — not as this year's profit bridge. Risk: wholesale retailers notice the off-price dump and demand permanent price concessions; ring-fence the dump SKUs.
That is the case: one driver, local fix, newspaper branch demoted.
Ned's rule. If the headline cost cannot arithmetic to the profit delta, it is not the driver. Say that out loud before you analyze it for twenty minutes.
When this tree is the wrong first move
Industry structure for all players. Use Porter when the question is why margins are thin across the category, not why this client's profit moved $23 million.
Where cost sits in the operating system. After you isolate cost, operations and cost and value chain take over. Do not start there.
Which division to fund. Portfolio tools sit after a clean profit bridge per division, not instead of one.
How to pick any structure without naming it. See the framework selection guide. MECE still applies when you segment; see MECE.
Practice this until one branch dies
Open a profitability prompt. Give yourself four minutes: build the bridge, name the owning Level-2 line, name the segment, then state the recommendation in one sentence that mentions only that segment. If your closing sentence still lists three initiatives, restart.
Structure drills on profit prompts live at /drills/structure. For a full voice case with a scored debrief, use /interview (CopperCup, about 18 minutes). Everything on CoachNed is open for seven days with no card; then $120 for a recruiting season or $49 a month.
Frequently asked questions
What is the profitability framework in a case interview?
It is profit equals revenue minus cost, with revenue split into price, volume, and mix, and cost split into fixed and variable, then segmented until one driver explains the delta. Interviewers care about isolation and the recommendation that follows, not about whether you said the word "framework."
How do I start a profitability case?
Ask for a year-over-year profit bridge in dollars. Compare revenue versus cost first. Only then open price/volume/mix or fixed/variable. Do not list initiatives until one branch owns most of the move.
What is the difference between price and mix?
Price is dollars per unit holding the product or customer mix constant. Mix is a shift toward units (or channels, or customers) with different dollars per unit or different contribution. Treating mix as price is the most common analytics fail on profit cases.
Should I always look at costs first when profit is down?
No. Look at whichever side owns the delta. Revenue can rise while profit falls if mix and variable costs degrade, as in the Meridian Home example. Start with the bridge, not with a habit.
How detailed should my profit tree be?
Deep enough that one leaf explains most of the residual, shallow enough that you can quantify it in the time left. Three to five useful splits beat twelve decorative ones. If two segments still split the residual evenly, cut once more.
Sources
- Costco Wholesale Corporation Form 10-K for the fiscal year ended September 1, 2024 — net sales, merchandise costs, membership-fee income, and low-merchandise-margin operating model. Checked 2026-09-24.
- Starbucks Q4 and Full Fiscal Year 2024 Earnings Release (PDF) — store operating expenses as a percent of company-operated store revenues, operating margin, and North America ticket versus transactions. Checked 2026-09-24.
- Southwest Airlines Co. Form 10-K for the year ended December 31, 2024 — fuel and oil expense, fuel as a share of operating expenses, and fuel cost per gallon. Checked 2026-09-24.
- NIKE, Inc. Fiscal 2024 Fourth Quarter and Full Year Results (PDF) — gross margin and selling and administrative expense levels for a branded goods cost shape. Checked 2026-09-24.
CoachNed is not affiliated with McKinsey, BCG, Bain, Deloitte, Costco, Starbucks, Southwest, Nike, or any firm named above.