How do you structure a profitability case?

Profitability is the most common type of case interview question in consulting. A client is making less money than they used to, and you have to work out why.

If you have never done a case before, here is what one actually is. The interviewer describes a business problem. You ask questions, they give you information, and together you work toward an answer over about twenty minutes. It is a conversation, not a test with a right answer at the back of the book.

The underlying structure of a profitability case is almost embarrassingly simple: profit is revenue minus cost. So if profit fell, either revenue fell or costs rose, or both.

Everything after that is disciplined narrowing. The skill is not knowing the framework — everyone knows it. The skill is using it without getting lost.

The worked example throughout this guide is a regional airline losing money.

Clarify before you structure

The most common early mistake is diving straight into a framework. Spend thirty seconds asking questions first.

Two or three, no more. Good ones:

"When you say profitability, do you mean total profit or profit margin?" These can move in opposite directions. A company can grow total profit while margins shrink.

"Over what period has this been happening?" A problem that appeared last quarter is different from a five-year decline.

"Is the rest of the industry facing the same pressure, or is this specific to our client?" This one question can save you ten minutes. If fuel costs rose for every airline, the question is not "why did costs rise" but "why is our client hurt more than everyone else."

Asking these signals that you understand the problem before solving it. It is also just what a consultant does on a real project.

State your structure out loud

Before diving in, say what you are going to do. Something like:

"Profit is revenue minus cost, so I would like to establish which side has moved. Then I will break down whichever one is driving the problem."

This takes five seconds and gives the interviewer something to follow.

What MECE means

You will hear this word constantly in consulting. It stands for mutually exclusive, collectively exhaustive.

Mutually exclusive means your categories do not overlap. Nothing belongs in two buckets.

Collectively exhaustive means your categories cover everything. Nothing is left out.

Splitting customers into "business travellers" and "leisure travellers" is MECE — every passenger is one or the other, and nobody is both on the same trip.

Splitting them into "business travellers" and "people flying to Spain" is not. A business traveller might be flying to Spain, so the buckets overlap, and travellers going elsewhere for other reasons are missing entirely.

Overlapping buckets suggest you have not thought it through. Gaps suggest you are reciting something you memorised. Interviewers listen for both.

Breaking down revenue

For most businesses, revenue is price times volume. Then you split further.

For our airline:

Revenue = number of passengers × average fare

Passengers breaks into: how many routes they fly, how many flights on each route, how many seats per plane, and what percentage of those seats are filled.

That last one has a name in the industry: load factor. It is the percentage of available seats actually sold. Airlines watch it obsessively.

Average fare breaks into: the mix of ticket types, the mix of cabin classes, and extra revenue from bags, seat selection and food.

You can also cut revenue by segment: by route, by customer type, by time of year. This matters, and we will come back to it.

Breaking down cost

Split costs two ways. Both are useful.

By type. For an airline: fuel, staff, aircraft ownership or lease payments, maintenance, airport and landing fees, and distribution costs like booking commissions.

By behaviour. Are they fixed or variable?

Fixed costs stay roughly the same whether you fly or not. Aircraft leases, most staff, headquarters.

Variable costs change with how much you fly. Fuel, catering, landing fees.

This distinction matters enormously for airlines because they are heavily fixed-cost businesses. Once the plane is flying, one more passenger costs almost nothing. That is why an empty seat is so painful — it earns nothing but the flight cost the same. It is also why load factor drives airline economics so hard.

Ask for data and narrow

Now you start working with numbers.

Ask which side moved. Suppose you learn revenue is down 8 percent while costs are flat.

Good. Now costs are not the problem. Say that out loud and move on. Do not spend eight minutes on cost structure to prove it.

Then narrow inside revenue. Passengers or fare? Suppose passengers are flat but average fare is down 9 percent.

Now you are somewhere useful. Fares have fallen. Why?

Segment, always

Aggregate numbers hide the answer. This is the single most useful habit in case interviews.

"Average fare is down 9 percent" is where the analysis starts, not where it ends.

Down on which routes? For which customer type? At what time of year? Since when?

You might discover that fares are flat on most routes but collapsed on three, because a low-cost competitor started flying them last year. That is a specific, actionable finding. "Fares are down" is not.

Look outward, not just inward

Always ask what the rest of the industry is doing.

If every airline's fares fell 9 percent, this is a market problem, and the answer is about how to respond to industry conditions.

If only our client's fares fell, this is a company-specific problem, and the answer is about what they are doing differently.

Same symptom, completely different recommendation. Candidates who only look inside the company miss half the picture.

Form a hypothesis early

Strong candidates do not examine every branch equally. They form an early view and test it.

"Given the timing and the fact that costs are flat, my hypothesis is that a new competitor is putting pressure on fares in specific markets. I would like to look at route-level fare data to test that."

This is faster, and it is how real consultants work. Weaker candidates check everything methodically and run out of time before reaching a conclusion.

Land a recommendation

The last minute matters more than any other. Interviewers remember whether you finished.

A good synthesis has four parts:

Name the driver. "The profit decline is driven by a 9 percent fall in average fare, concentrated on three routes where a low-cost competitor entered last year."

Recommend two or three actions. "I would look at whether we can compete on cost on those routes, whether we should redeploy capacity to less competitive routes, and whether there is a differentiated service angle."

Size the impact roughly. "Restoring fares on those three routes would recover roughly two-thirds of the profit gap."

Name the main risk. "The risk is that the competitor has a structurally lower cost base, in which case competing on price is not sustainable."

A case that ends with a description of the problem is unfinished, however good the analysis was.

Saying it out loud

The whole shape, compressed:

"Profit is revenue minus cost, so first I would establish which has moved.

On revenue, I would break it into passengers times average fare, then segment by route and customer type. On cost, I would split fixed from variable, since airlines are heavily fixed-cost and load factor drives the economics.

Once I know which side is driving it, I would narrow down within that branch, and I would compare against the industry to see whether this is market-wide or specific to us.

Then I would come back with the driver, two or three recommendations, a rough size of the impact, and the main risk."

What the interviewer is really checking

Do you clarify before diving in?

Two or three questions at the start prevent ten minutes of analysing the wrong thing. Interviewers watch for this specifically.

Is your structure MECE?

No overlaps, no gaps. This is the most-used word in consulting recruiting for a reason.

Are you hypothesis-driven?

Do you form a view and test it, or do you check everything equally and run out of time? The first is how consultants work.

Can you do arithmetic out loud?

Cases involve numbers, delivered verbally, under mild pressure. Comfort here shows immediately.

Do you segment?

Averages hide answers. The driver almost always lives in a segment rather than the total.

Do you synthesise?

The final sixty seconds carry disproportionate weight. Naming a driver, recommending actions, and stating a risk is what a finished case sounds like.

Common mistakes

Reciting a memorised framework.

Interviewers can tell within seconds. Describe your structure in your own words instead of naming a framework from a book.

Skipping clarification.

Thirty seconds of questions at the start saves the whole case.

Jumping straight to cutting costs.

It is the reflex, and it is frequently wrong. Find out which side moved first.

A structure that is not MECE.

Overlapping buckets undermine everything built on top of them.

Never segmenting.

"Revenue is down 8 percent" is a starting point. Down where, for whom, since when?

Examining every branch equally.

Without a hypothesis you will run out of time before reaching an answer.

Getting lost in a branch that does not matter.

If costs are flat, say so and move on. Do not spend eight minutes proving it.

Ignoring the industry.

A market-wide problem and a company-specific problem need different answers.

Ending with findings instead of a recommendation.

"So those are the factors" wastes the case. Land it.