Airline case interview, the unit metrics and a worked case

Osprey Air, an invented U.S. airline, carried 9.5% more passengers this year, yet operating profit fell from $148.1 million to $41.8 million. The CEO blames fuel. Is she right, and what should the airline do? This case and every number in it are invented by Soreno for practice.

1. Clarify

   One airline, two years, operating profit

   Pin down the goal (restore last year's operating profit), the time frame (next year) and what changed: Osprey flew 10 billion seat miles last year and 11.5 billion this year, 15% more, after opening new routes. Average trip length stayed at 800 miles and fees at $30 a passenger.

2. Size the drop

   $148.1 million to $41.8 million

   Revenue rose from $1.49 billion to $1.59 billion, but cost rose faster, from $1.34 billion to $1.55 billion. Margin fell from 9.9% to 2.6%. Passengers went from 10.5 million to 11.5 million, so volume is not the problem.

3. Go per seat mile

   Unit revenue fell 1.09¢, unit cost moved 0.03¢

   Total revenue per seat mile (TRASM) went from 14.91¢ to 13.82¢. Cost per seat mile (CASM) went from 13.43¢ to 13.46¢, almost flat. Almost all of the lost margin per seat mile is on the revenue side.

4. Test the fuel story

   Fuel bill up 29%, but offset

   At 70 seat miles a gallon, fuel went from $342.9 million to $443.6 million: more flying and a price rise from $2.40 to $2.70 a gallon. Fuel per seat mile rose from 3.43¢ to 3.86¢. Cost before fuel fell from 10¢ to 9.6¢ a seat mile as fixed costs spread over more flying, which cancels almost all of it. The CEO's story doesn't hold.

5. Split unit revenue

   Load factor 84% to 80%, yield 14¢ to 13.53¢

   Both drivers fell. By route group: the old network sold 83% of its seats at 13.8¢ a mile, close to last year. The new routes sold only 60% at 11¢, low introductory fares to fill planes. The new flying dragged both averages down.

6. Break-even load factor

   New routes need 91.2%, they fly at 60%

   Fees add $30 ÷ 800 miles = 3.75¢ per passenger mile. At this year's CASM of 13.46¢, a new route breaks even at 13.46¢ ÷ (11¢ + 3.75¢) = 91.2% load factor, against 60% flown. Charged the network's unit cost, they lose $69.1 million a year, while the old network earns $110.9 million (break-even 76.7%, flown at 83%).

7. Recommend

   Fix or cut the new routes, not the fuel bill

   The fall is a capacity decision, not fuel. Give each new route a deadline to reach its break-even. At 60% full it would need a yield of about 18.68¢, and at 11¢ it would need 91.2% of its seats sold, more than the 83% the old network flies. Neither lever gets there alone, so a route has to show both fuller planes and firmer fares, and one that can't should go. Move aircraft off the weakest routes to old routes that turn passengers away, and check that the added seats sell, since more seats on a route running 83% full lowers its load factor too. Check before cutting: aircraft leases and crews don't vanish with a route, so cutting saves less than the $69.1 million loss. Next step: route-by-route load factor and fares for the last six months.

Definitions follow the MIT Airline Data Project's glossary. The formulas are what you use in the case.

1. ASM

   Available seat miles: capacity

   One seat flown one mile, whether occupied or not. MIT's example: a 100-seat aircraft flown 100 miles generates 10,000 available seat miles.

2. RPM

   Revenue passenger miles: traffic

   One paying passenger flown one mile. MIT's example: 200 passengers flying 500 miles generate 100,000 RPMs.

3. Load factor

   RPM ÷ ASM

   The share of output actually sold. For a single flight, MIT notes you can simply divide passengers by seats.

4. Yield

   Passenger revenue ÷ RPM

   The average fare per passenger mile, in cents. MIT warns it isn't useful for comparing markets or airlines, since it varies with stage length and leaves out load factor.

5. PRASM

   Passenger revenue ÷ ASM = load factor × yield

   MIT calls PRASM equivalent to the product of load factor and yield. In the worked case, 84% × 14¢ = 11.76¢. TRASM adds bags, change fees, cargo and other revenue to the top line.

6. CASM

   Operating expenses ÷ ASM

   Unit cost. MIT notes fuel is sometimes left out (CASM ex fuel) to compare operating costs more directly.

7. Break-even load factor

   CASM ÷ revenue per RPM

   The load factor at which unit revenue equals unit cost. Above it the flight makes money. It is a standard case formula, not an MIT glossary term.

8. Stage length

   Average miles per departure

   MIT's glossary says operating costs and revenues are significantly affected by distance flown, and describes stage length adjustment as a common practice for comparing carriers. As a rule of thumb (ours), longer flights spread take-off, landing and airport costs over more miles, so CASM falls as stage length rises.

Figures you can use to sanity-check a case, from the Bureau of Transportation Statistics' May 12, 2026 release on 2025 results for 25 U.S. scheduled passenger airlines and its passengers data story.

1. Revenue

   $252.6 billion, 74.1% from fares

   Fares were 74.1% of 2025 operating revenue and baggage fees 2.9%. Ancillary revenue matters, but fares still pay most of the bills.

2. Profit

   $11.4 billion operating profit

   That is a 4.5% operating margin, the figure BTS itself reports for 2025, down from 5.5% in 2024. Thin margins mean a point or two of load factor can decide whether a route makes money.

3. Costs

   Labor 37.5%, fuel 15.2% of domestic expenses

   Those are the shares of 2025 domestic operating expenses. In the second quarter of 2026, fuel rose to 21.7% and labor fell to 33.1% of domestic expenses. Labor is the bigger line, so don't assume fuel drives every cost question. Use the split the interviewer gives you.

4. Load factor

   81.3% in 2024, 79.4% in 2025

   BTS's passengers story, built on T-100 segment data for every U.S. and foreign carrier flying within, to and from the United States, measures load factor as passenger miles over available seat miles. It counted 1.39 billion seats in 2025.

The openings are ours. Each names the first ratio to look at, so the interviewer hears a plan in the first minute.

1. Profit decline

   Unit revenue against unit cost

   Split TRASM and CASM by year, then load factor, yield and fees on one side and fuel and cost before fuel on the other. Check capacity growth before blaming any one cost.

2. New route

   Break-even load factor against demand

   Estimate passengers a day from the market, set seats from the aircraft and frequency, then compare expected load factor with break-even. Add the competitor's likely reply and one-time launch cost.

3. Fleet or aircraft

   Cost per seat, not cost per plane

   A bigger aircraft costs more per trip and less per seat. It only wins if the extra seats sell. Compare trip cost, seats and the load factor each option needs.

4. Fees and pricing

   Fee revenue against lost fares

   A higher bag fee raises revenue per bag but some passengers carry on, pack less or book a rival. Size the fee gain, then the passengers and fares at risk.

5. Merger

   Network overlap and synergies

   Where do the two networks overlap, which costs combine (fleet types, airports, overhead) and what revenue grows from connecting traffic? Then the integration cost and regulatory risk.

Two openings to the Osprey prompt, written by us.

1. Weak opening

   A generic profit tree

   "Profit is revenue minus costs. Revenue is price times quantity, costs are fixed and variable. Let me also look at competitors, customers and the market." Nothing here is wrong, but it ignores that passengers rose, never names a seat mile and would take ten minutes to get to the answer.

2. Strong opening

   Per seat mile, with a hypothesis

   "Passengers grew 9.5% and profit fell, so more flying is not paying for itself. I'd like to compare revenue and cost per seat mile across the two years. If unit cost is flat, the problem is load factor or yield, and my first guess is the new routes. If unit cost jumped, I'll split fuel from everything else."

L.E.K.'s interview preparation page lists this as a brainteaser among its case examples: planes leave JFK and LAX every hour on the hour on identical New York to Los Angeles routes, the flight takes five hours, and you leave New York at 12:00 PM EST. How many planes flying the opposite way will you pass? L.E.K. asks what clarifying questions you would ask and which logical traps to avoid. The page we read gives no answer. The working below is ours.

1. The trap

   Time zones don't matter here

   Los Angeles is three hours behind New York, a whole number of hours, so on the hour in Los Angeles is also on the hour in New York. Do the whole problem in Eastern time.

2. The window

   Planes that left between 7 and 17

   You are in the air from 12:00 to 17:00 Eastern. A plane from LAX is in the air for five hours, so any plane that left after 7:00 and before 17:00 is airborne at some point while you are.

3. The count

   9 in the air, 11 counting the ends

   Departures at 8:00 through 16:00 are 9 planes passed in flight. The 7:00 plane lands as you take off and the 17:00 plane leaves as you land. Say which you count and why. The clear assumption is what earns the credit.