CaseClimbCase interview maths is seldom hard. It's arithmetic, done out loud, in front of someone, while you also hold the whole case in your head. The shortcuts matter because they free up your attention for the thinking.
These are the kinds of calculations an interviewer hands you during a case, with the answers:
- Market share: "The market is $4.5 billion and the client has a 12% share. What is its revenue?" 4.5 billion × 12% = $540 million.
- Price and volume: "The price rises 8% and volume falls 5%. What happens to revenue?" 1.08 × 0.95 = 1.026, so revenue rises about 2.6%.
- Cost savings: "Revenue is $80 million and costs are $68 million. If we cut costs by 5%, what happens to profit?" Costs fall by $3.4 million to $64.6 million, so profit rises from $12 million to $15.4 million (about 28% higher) and the margin from 15% to about 19%.
- Break-even: "Fixed costs are $2 million and each unit contributes $40. How many units to break even?" 2,000,000 ÷ 40 = 50,000 units.
None of these needs more than a few steps. The challenge is doing them without slips, out loud, and knowing which number to use next.
Say your approach before you calculate
Before you touch a number, say what you're about to compute and why: "I'll take the market size, multiply by our share, then by the average price to get revenue." It gives the interviewer a chance to correct the direction before you've spent a minute on it, and it stops you starting a calculation you can't finish.
The habit also catches the wrong base. If costs fell from 40% of revenue to 36% and you announce "I'll take 4% of revenue", the interviewer can stop you if you had meant 4% of costs.
Round, then adjust
Work with round numbers and adjust at the end. 4.9 million × 21% becomes 5 million × 20% = 1 million, and since 4.9 is a little under 5 and 21 a little over 20, the true answer is very close to 1 million (it's 1.03 million). Tell the interviewer you're rounding and by about how much. The main thing they want to see is that you can judge what precision matters, as it does in market sizing.
Percentages
- 10% means moving the decimal one place; 5% is half of that; 1% is two places. So 15% of 240 is 24 + 12 = 36.
- Know the fractions: 50% is 1/2, 25% is 1/4, 20% is 1/5, 12.5% is 1/8, 33% is 1/3, 16.7% is 1/6.
- To increase by 20%, multiply by 1.2. To decrease by 20%, multiply by 0.8.
- Chain changes by multiplying. +10% then −10% is 1.1 × 0.9 = 0.99, a 1% fall, not zero.
Two traps sit beside these rules. A margin that moves from 10% to 12% has risen two percentage points, and 20% in relative terms (2 ÷ 10); say which one you mean. And to undo a change, divide. If revenue fell 20% to $96 million, the original was 96 ÷ 0.8 = $120 million. Multiplying by 1.2 gives $115.2 million, which is wrong.
Big numbers: work in units
Convert to thousands, millions or billions, calculate on the small numbers, and track the zeros on their own. 250,000 × 1,200 becomes 250 thousand × 1.2 thousand: 250 × 1.2 = 300, and thousand × thousand is million, so 300 million. Say the unit each time you state a number ("300 million dollars", not "300"). Dropped units are one of the most common sources of slips.
The same discipline handles time. Revenue of $7.3 billion a year is $20 million a day (7,300 ÷ 365 = 20). A business with 1.2 million subscribers paying $25 a month earns $30 million a month, so $360 million a year. Write "per month" or "per year" beside each intermediate result.
Growth and doubling
The rule of 72 gives a quick estimate of doubling time: divide 72 by the annual growth rate in percent. At 8% growth, something doubles in about 9 years (72 ÷ 8 = 9). For a few years of modest growth you can add the rates as a rough guide, but remember the true figure is a little higher: 5% a year for three years is about 15.8% (1.05 × 1.05 × 1.05), not exactly 15%.
A few values are worth memorising: 3% doubles in about 24 years, 6% in 12, 9% in 8 and 12% in 6. The rule runs in reverse too. A market that has quadrupled in 12 years has doubled twice, so its doubling time is 6 years and its growth rate is about 12% a year (1.12 to the twelfth power is about 3.9).
Over several years the gap grows. A $200 million market growing at 10% a year for five years: adding five lots of 10% suggests $300 million, while compounding, 200 × 1.1 five times over, runs 220, 242, 266.2, 292.8, 322.1. So about $322 million, 61% above the start against the shortcut's 50%. A $22 million gap is bigger than many of the savings a case turns on.
Break-even and margin
- Break-even volume = fixed costs ÷ (price − variable cost per unit).
- Example: fixed costs of $600,000, a price of $50 and a variable cost of $30 per unit give a contribution of $20 a unit, so break-even is 600,000 ÷ 20 = 30,000 units.
- Margin = profit ÷ revenue. With revenue of $200M and costs of $170M, profit is $30M and the margin is 15%.
- One point of margin on $200M of revenue is $2M of profit, the kind of number a profitability case turns on.
A full calculation done out loud
Put the pieces together on a pricing case. The client sells 4 million units a year at $60, so revenue is $240 million. It is considering a 15% price rise and expects volume to fall 10%. Variable cost is $45 a unit and fixed costs are $50 million. The interviewer asks what happens to revenue and profit, then whether a new line with fixed costs of $12 million, a price of $40 and a variable cost of $25 can break even in a market of 10 million units. A strong answer sounds like this:
"Price up 15% and volume down 10% means 1.15 × 0.9, which is 1.035, so revenue rises about 3.5%, from $240 million to about $248 million. Profit moves more, because the volume we lose carried a $15 contribution and the volume we keep now carries $24. Old contribution is 4 million × $15, so $60 million. New contribution is 3.6 million × $24, so $86.4 million. Fixed costs don't change, so profit rises about $26 million, from $10 million to about $36 million. That's a large gain for a small revenue change, so the assumption I'd want to test next is the 10% volume loss."
Then the new line. Contribution is $40 minus $25, so $15 a unit. Break-even is 12,000,000 ÷ 15 = 800,000 units, which is an 8% share of a 10 million unit market. Say that share out loud and compare it with what the client holds elsewhere; the arithmetic is finished, and that comparison is the point you carry into the recommendation.
Reading a table or chart under time pressure
Read the title, the units and the axis labels before any value, and say them back: "This is revenue by segment in millions, for 2024 and 2025." Most exhibits exist to make one point, found in the biggest gap or the one line moving the other way. Compute only what the question needs.
The classic trap is averaging percentages that sit on different bases. If one segment has $60 million of revenue growing at 10% and another has $40 million shrinking 5%, the average of the rates, 2.5%, is the wrong answer. Work the money: 60 × 1.1 = 66, 40 × 0.95 = 38, so the total moves from 100 to 104, a 4% rise. If a chart gives percentages with no absolute base, ask for one before you calculate.
Sanity-check every answer
Before you announce a result, ask whether it's the right order of magnitude. A market share above 100%, or revenue per customer larger than the customer's income, points to a misplaced zero in most cases. Saying "that looks plausible because..." shows judgement as well as accuracy, two of the five skills interviewers score.
Common mistakes and the fix
- Dropping the unit halfway through, so 300 becomes dollars when it was millions. Fix: say the unit with the number each time, and write it down.
- Treating percentage points as a percent change, or the reverse. Fix: state both and use the one the question needs.
- Undoing a percentage change by multiplying instead of dividing. Fix: to recover a figure that fell 20%, divide by 0.8.
- Stopping at the number. Fix: say what it means for the client in one sentence before the interviewer has to ask.
How interviewers score the maths
Quantitative analysis is one of the five scored skills, marked on accuracy and pace, but the other four are in play at the same time. Laying out the formula first, as you would lay out a structure, is structuring. Choosing to round, and knowing the point at which rounding stops being safe, is business judgement. Turning $26 million into "so the pricing move is worth more than the cost programme we discussed" is synthesis. Narrating so the interviewer can follow without a pen is communication.
How to practise
A ten-minute routine, done most days alongside the cases in your prep plan, is more useful than anything longer done once:
- Two minutes: percentages to fractions and back, out loud. 12.5% of 640, 37.5% of 80, a sixth of 4,200 (80, 30 and 700).
- Three minutes: five price-and-volume chains, such as +12% and −8% (1.12 × 0.92 = 1.0304, about +3%), spoken aloud.
- Two minutes: two break-even problems, finding contribution per unit first.
- Two minutes: one rule of 72 question and one three-year compound growth check, done by adding and then by multiplying.
- One minute: write down the cause of each slip (unit, base or arithmetic) and watch for it the next day.
Ten minutes a day of timed mental maths beats an occasional long session, and it doubles as preparation for a numerical reasoning test. Then practise the maths inside real cases, where you have to do it while talking. CaseClimb's numerical drills give you fresh timed sets with an explanation for each answer, and your case interviews include live calculations with an interviewer who checks your working.
Common questions
Do I need to memorise formulas for case interview maths? A handful cover most of it: revenue is price times volume, margin is profit over revenue, break-even is fixed costs over contribution per unit, plus the rule of 72. The work is in applying them at pace, which comes from repetition.
Is it acceptable to round in a case interview? Yes, and interviewers expect it, provided you say that you are rounding, give the direction of the error, and tighten the numbers if two options come out close. Rounding without saying so looks like a mistake even when the answer is sensible.
Does a slow but correct calculation lose marks? Pace is scored, so a very slow calculation does cost something, but a transparent one at moderate speed scores better than a fast silent one. Be steady and audible first, then build speed with daily drills.
Is case maths the same as the numerical reasoning test? The arithmetic overlaps, so drilling one helps the other, but the format differs. The test is silent, multiple choice and timed per question; the case is spoken and judged on how you set up and interpret the number as much as on the number itself.
