PartnerRoundBy the end you can
- Draw a profit tree tailored to the client's business in 60 seconds
- Run the diagnosis in order: trend, segment, compare, root cause
- Build a profit bridge that explains every dollar of a change
- Turn the root cause into a recommendation with a number attached
On this page
01The profit tree
Profitability prompts come in a few forms: profit fell, margins trail competitors, a division loses money. They are among the most common cases at every major firm, because they test structure and maths at once.
Volume = market size × market share
Fixed cost does not move with volume in the short run
Two details separate good trees from average ones. Splitting volume into market size and share tells you whether the market shrank or the client lost ground in it. Putting mix under price catches the case where no price changed but customers bought cheaper products.
If the client has several revenue streams, split by stream first, then volume × price within each. An airline sells tickets, bags and cargo; a gym sells memberships and personal training.
For questions about volume, such as breakeven or dropping a product, use the contribution form: profit = (price - variable cost per unit) × volume - fixed cost. The bracket is the unit margin: what each extra sale adds to profit.
02The diagnostic order
The tree tells you where to look. The diagnostic order tells you in what sequence, so each question narrows the problem instead of touring every branch.
- 01TrendWhen did it change, and how fast?
- 02SegmentRevenue or cost? Which product, region, customer?
- 03CompareCompetitors too, or just us?
- 04Root causeWhy, in terms a manager can act on
- Trend. Ask for three to five years, not just this year against last. A sudden drop points to a shock: a new entrant, an input price, a lost contract. A slow slide points to something structural: mix, an ageing product, creeping overhead.
- Segment. Ask first whether revenue or cost moved more: the answer halves the tree. Then split the side that moved by product, customer, region or channel. Problems are rarely everywhere at once, and one segment usually explains most of the change.
- Compare. If competitors' margins fell by as much, the cause is industry-wide and the question becomes how to adapt faster than them. If only the client's fell, the cause is inside the company.
- Root cause. Keep asking why until you reach something a manager can change. "Costs rose" is a symptom. "Overtime doubled in two warehouses after a rival opened nearby paying $2 an hour more" is a cause.
Check yourself
Revenue is flat at $500m. Costs rose from $400m to $420m, so profit fell 20%. Competitors' margins are stable. What should you do next?
03The profit bridge
A profit bridge, also called a margin bridge or profit walk, starts at last year's profit and adds the effect of each driver until it reaches this year's. Every dollar of change is explained exactly once: MECE in numbers.
Example: a manufacturer sold 10m units at $50, with variable cost of $30 a unit and fixed costs of $100m, for a profit of $100m. This year it sold 9.5m units at $51, variable cost rose to $33 and fixed costs to $105m. Profit is now $66m.
The convention: value the volume change at last year's unit margin, then the price and unit cost changes at this year's volume. Other orders work if you apply them consistently, so say which one you are using.
Then read the bridge for the biggest bar. Here variable cost per unit explains $28.5m of a net $34m fall, so the next question is which input got more expensive.
Check yourself
In the bridge above, what would this year's profit have been if price had stayed at $50?
04Common root causes by industry
Knowing the usual suspects lets you form a hypothesis in the first minute. Treat them as hypotheses to test, never as answers to announce.
05Worked case: a regional airline
Worked example
Why did a regional airline's profit fall by two thirds?
Prompt. Our client is a regional airline with 20 aircraft. Operating profit fell from $60m to $20m in a year. The CEO wants to know why, and what to do.
Clarify. The goal is to restore profit, ideally to last year's level. The airline flies 40,000 flights a year with 100 seats each: 4.0m seats. Revenue is almost all ticket sales, so one revenue stream is enough.
Step 1, trend. Profit had been steady at about $60m for three years, then fell in one. That points to a shock, not a slow decline.
Step 2, segment. Revenue explains $32m of the $40m fall and fuel the other $8m. Passengers and load factor are flat, so the revenue fall is all price: the average fare dropped $10.
Split by route and the picture sharpens. A low-cost carrier launched on three of our routes this year. Those routes carry a quarter of our seats.
Step 3, compare. The other regional carrier's margin slipped from about 11% to 9%. That is roughly what fuel alone did to ours: on its own it would have taken our margin from 12.5% to 10.8%. Fuel is an industry problem; the fare collapse on three routes is ours.
Step 4, root cause. We matched the entrant's fares to protect load factor. It worked, load factor held at 80%, but we gave away $40 on every contested ticket. The entrant's cost per seat is about $70 against our $107, so a fare war is one we cannot win.
Check before you cut. Spreading all costs by seat, the contested routes now lose $7m: $100m of revenue against $107m of cost. But only their fuel ($27m) and passenger fees ($20m) disappear quickly if we stop flying them. They still contribute $53m towards aircraft, crew and overhead, so simply exiting would cut profit, not raise it.
Recommendation. Stop matching the low-cost fare on every seat. Protect business-hour flights and corporate contracts, where schedule and loyalty matter, and match only on a capped share of seats. Each $10 of average fare recovered on contested routes is worth $8m a year (0.8m passengers × $10).
Separately, test whether the 5 aircraft on those routes could earn more than $53m a year elsewhere. Handle fuel through hedging and fare surcharges, as competitors will.
06The questions that unlock the case
Good questions are tied to branches of your tree. Each one should close off part of the tree or point to the next split.
Weak
So we need to cut costs. They could reduce headcount, renegotiate with suppliers and close underperforming stores.
Strong
So the problem is on the cost side. Before recommending anything, I'd like to know which line moved. Could I see costs split into fixed and variable, per unit, over three years? If variable cost per unit rose, I'd look at input prices and product mix. If fixed costs rose, I'd look for new capacity or overhead that hasn't paid back yet.
Why: The weak answer jumps to generic cuts before finding the driver. The strong answer narrows to the right branch, asks for exactly the data that splits it, and says in advance what each result would mean. That is hypothesis-driven work, and it is what interviewers score.
Check yourself
A product line shows a $5m loss: revenue $40m, variable costs $30m, allocated overhead $15m. If you drop it and the overhead stays, what happens to company profit?
Cheat sheet
The lines worth rereading the night before.
- 1Profit = revenue - cost. Revenue = volume × price. Cost = fixed + variable cost per unit × volume.
- 2Split revenue by stream first if there are several; split volume into market size × share.
- 3Diagnose in order: trend, segment, compare with competitors, root cause.
- 4Ask "revenue or cost?" early: it halves the tree.
- 5Bridge the change: volume at last year's unit margin, then price and unit cost at this year's volume.
- 6Competitors hit too means an industry cause; only the client hit means an internal one.
- 7An allocated loss is not a reason to exit. Check contribution and which costs really go.
- 8Recommend with a number: "every $10 of fare is worth $8m a year".
Now drill it
Reading builds the model; drills make it automatic. Start with the first topic, easiest drills first.
- Profitability7 drills · Find out why profit moved by splitting it cleanly, segmenting early and asking the question that localises the problem.
- Profit & margins8 drills · Gross, operating and contribution margin, margin moves in points, and what a cost or price change does to profit.
- Cost reduction7 drills · Find where the cost sits, read benchmarks with care, and separate quick wins from structural change.
Read the full guide
The same ground on the blog, with more worked examples and practice on the page.
Questions candidates ask
What is the profitability framework in case interviews?
It is a profit tree: profit = revenue - cost, with revenue split into volume × price and cost into fixed and variable. You then work through it in order: trend, segment, compare with competitors, root cause. The tree is the skeleton; tailor its labels to the client's industry every time.
How do I start a profitability case?
Clarify the goal and the time frame, ask how the client makes money, then present a profit tree in under a minute. Your first analytical question should usually be whether revenue or cost moved more, because the answer halves the problem.
What is the difference between a profit problem and a margin problem?
Profit is a dollar amount; margin is profit as a share of revenue. Profit can fall with a stable margin if the business shrinks, and margin can fall while profit grows if revenue grows faster than profit. Clarify which one the client cares about before you analyse.
What if both revenue and costs changed?
Build a profit bridge. Start from last year's profit and add the effect of volume, price, variable cost per unit and fixed cost in turn until you reach this year's. The biggest bar tells you where to dig first.
Do I need industry knowledge for profitability cases?
No specialist knowledge is expected, but knowing each industry's key metrics helps you ask sharper questions. Load factor for airlines, like-for-like sales for retail and churn for software are enough to sound credible. If you do not know a metric, ask: interviewers will explain.