Contribution Margin, and the Break-Even That Actually Matters

Two subtractions that are not the same subtraction

Gross profit is revenue minus cost of goods sold. Contribution is revenue minus variable cost. Those look like the same sentence with a different noun, and if every cost inside COGS moved with volume they would be the same number. They almost never are. Cost of goods sold in a manufacturing business normally carries an allocation of fixed production overhead — factory rent, supervision, machine depreciation, quality inspection — spread across the units the plant expects to make. That allocation does not disappear when you make one unit fewer. Meanwhile some genuinely variable costs sit below the gross-profit line: sales commission, outbound freight, payment fees, packaging charged to distribution. Gross margin therefore contains fixed cost it should not, and misses variable cost it should.

Here is the whole thing on one product. The price is $100. Direct materials and direct labour come to $52 a unit. The plant absorbs $150,000 of fixed factory overhead across a plan volume of 10,000 units, which is $15 a unit, so reported COGS is $67 and gross profit is $33 — a gross margin of 33%. Below that line, commission and outbound freight take another $8 a unit, and fixed selling and administrative costs are $150,000. The variable cost of one more unit is therefore $52 + $8 = $60, and the contribution margin is $40, or 40%. The contribution margin is higher than the gross margin, which surprises people who assume more subtraction always means a smaller number.

Both routes have to reach the same operating profit, and they do. Gross profit of $33 on 10,000 units is $330,000; subtract $150,000 of fixed selling and administrative cost and $80,000 of variable selling cost and you have $100,000. Contribution of $40 on 10,000 units is $400,000; subtract the full $300,000 of fixed cost and you have $100,000. If your two routes disagree, a cost has been classified twice or not at all.

Break-even is fixed cost divided by contribution margin — and nothing else

Every unit you sell hands you its contribution margin, and the fixed cost pool sits there waiting to be filled. Break-even in units is therefore fixed cost ÷ contribution margin per unit: $300,000 ÷ $40 = 7,500 units. In revenue it is fixed cost ÷ contribution margin ratio: $300,000 ÷ 0.40 = $750,000, which is 7,500 units at $100, as it must be. Above 7,500 units every further unit adds $40 straight to operating profit. At the plan of 10,000 units, the 2,500 units above break-even deliver exactly the $100,000 of profit we already computed. Margin of safety — how far sales can fall before the profit is gone — is (10,000 − 7,500) ÷ 10,000 = 25%.

Now watch what happens if you use gross profit instead. Divide the full $300,000 of fixed cost by the $33 of gross profit per unit and you get 9,091 units — 21% higher than the truth, because the $150,000 of factory overhead is already inside the $67 of COGS and you have just charged it a second time. Try to correct for that by dividing only the $150,000 of non-manufacturing fixed cost by $33 and you get 4,545 units, which is far too low: at 4,545 units the real result is a loss of $118,200. Both errors come from the same place. Gross profit per unit contains an overhead rate that was computed at plan volume and is simply not valid at any other volume. Contribution margin per unit contains no such assumption, which is why it is the only per-unit number that survives a change in volume.

Operating leverage: the same profit, twice the sensitivity

Take two businesses that both sell 10,000 units at $100 and both make $100,000 of operating profit. Business A makes the product itself: contribution margin $40 a unit, fixed cost $300,000. Business B outsources manufacturing and pays a converter per unit: contribution margin $20 a unit, fixed cost $100,000. Same revenue, same profit, completely different machines underneath. A breaks even at 7,500 units; B breaks even at 5,000. A's margin of safety is 25%; B's is 50%.

The degree of operating leverage is total contribution ÷ operating profit. For A that is $400,000 ÷ $100,000 = 4.0; for B it is $200,000 ÷ $100,000 = 2.0. Read it as a multiplier on volume: a 1% change in units produces a 4% change in A's profit and a 2% change in B's. Push volume up 10% and A's profit goes from $100,000 to $140,000 while B's goes to $120,000. Pull volume down 10% and A falls to $60,000 while B falls to $80,000. A 10% swing in demand — one lost distributor, one wet summer — moves A's profit by 40% in either direction. That is the whole content of the phrase high operating leverage, and it is why the same profit can mean very different things about how much risk a company is carrying.

Two things follow. First, leverage is not a constant: it falls as you move away from break-even. A's degree of operating leverage is 4.0 at 10,000 units, 2.67 at 12,000 and 2.0 at 15,000 — and 76 at 7,600 units, just above break-even, where a 1% move in volume swings profit by three quarters. Second, high leverage is not a defect. It is a bet that volume will hold, paid for with a lower variable cost. A earns $40 per unit against B's $20; every unit above 10,000 rewards A twice as much. The question is never whether leverage is good, but whether you can see far enough into next year's volume to be comfortable with the one you have chosen.

The price cut that quietly needs a third more volume

A price cut lands entirely on contribution margin, because none of your costs move when you change the number on the invoice. Cut the $100 price by 10% and you lose $10 of revenue, but that $10 comes out of a $40 contribution margin, not out of the $100. Contribution margin falls to $30 — a cut of 25%, two and a half times the headline. The multiplier is exactly price ÷ contribution margin, which is 1 ÷ the contribution margin ratio: at a 40% ratio it is 2.5, at a 20% ratio it is 5.0.

The volume you need to stand still follows from that. To hold total contribution flat you need old margin ÷ new margin, minus one. After a 10% price cut that is $40 ÷ $30 − 1 = +33.3%: you must sell 13,333 units instead of 10,000 simply to end the year where you started, and you will have carried a third more freight, a third more warranty exposure and a third more working capital to do it. A 20% cut needs double the volume. A 25% cut needs 2.67 times the volume. The table above runs the ladder. Nobody who has seen those numbers proposes a 20% discount casually.

The same arithmetic runs pleasantly in the other direction. Raise the $100 price by 10% and contribution margin goes from $40 to $50; you can lose 1 − 40/50 = 20% of your volume and still be no worse off. And it explains why discounting is far more dangerous for a low-margin business: at business B's $20 contribution margin, a 10% price cut halves the margin outright and requires the volume to double. The lower your contribution margin ratio, the less room you have to buy share with price.

Building the number without lying to yourself

Classify each cost by asking one question: if I made and sold one more unit tonight, would this invoice be bigger? Direct materials, piece-rate labour, freight out, payment fees, sales commission and per-unit royalties all pass. Rent, salaried supervision, machine depreciation, software subscriptions, the accountant and the insurance all fail, however carefully they were allocated to a product in the costing system. Overtime is the awkward middle: it is variable at the margin but only after a threshold, so treat it as a step.

Two limits are worth stating out loud. Contribution margin is only valid inside a relevant range — a volume band in which the fixed costs really are fixed. Add a second shift, a second warehouse or a second machine and the fixed pool jumps, break-even jumps with it, and the old margin of safety evaporates overnight. And contribution margin is a decision tool, not a reporting basis: under IFRS and most national frameworks, fixed production overhead must be absorbed into inventory at normal capacity, so the accounts you file will still show gross margin, not contribution. Run both. Report the first, manage on the second.

What to do with the number on Monday

Rank products by contribution margin per unit of the thing you are short of, not by gross margin. If the constraint is machine hours, rank by contribution per machine hour; if it is shelf space, by contribution per facing; if it is your own selling time, by contribution per call. A product with a 60% gross margin that occupies the bottleneck for an hour is worse than a 30% product that clears it in ten minutes, and only the constrained ratio shows it.

Use it to price marginal business honestly. A one-off export order at $70 looks like a loss against a $67 COGS and a disaster against a $100 list price, but it contributes $70 − $60 = $10 a unit, which is $10 more than declining it — provided it genuinely does not displace a full-price sale, does not require new fixed cost and does not become the price your existing customers discover. Those three conditions are where marginal pricing usually goes wrong; the arithmetic almost never does. For the mechanics of setting a price from a target margin, and the margin-versus-markup trap that sits next door, see the companion article in this series.

Frequently asked questions

Can contribution margin be higher than gross margin?

Yes, and it usually is in manufacturing. Gross margin has already absorbed fixed factory overhead into cost of goods sold, while contribution margin excludes it. In the example here, gross margin is 33% and contribution margin is 40%, because $15 a unit of fixed overhead sits inside COGS and only $8 a unit of variable selling cost sits outside it. The reverse happens in businesses with almost no fixed production cost but heavy per-order costs — most ecommerce — where contribution margin lands well below gross margin.

What is a good contribution margin?

There is no universal figure, and any number quoted without a named source and a defined industry should be ignored. The only meaningful test is internal: total contribution has to exceed total fixed cost by enough to fund the return you need on the capital employed, at a volume you can actually sell. A 15% contribution margin is comfortable in a business with tiny fixed costs and enormous volume; a 60% contribution margin is thin in one with a large plant and a short season. Compute your own break-even and margin of safety instead of hunting for a benchmark.

Should I use contribution margin per unit or the contribution margin ratio?

Use the per-unit figure when you want an answer in units and you sell one thing: fixed cost ÷ contribution per unit gives break-even units directly. Use the ratio when you sell a mix, because units are not comparable across a catalogue: fixed cost ÷ weighted contribution margin ratio gives break-even revenue. The weighting must be by revenue share, not by unit share, and it is only valid while the mix holds — a shift towards low-margin lines raises break-even revenue without a single price changing.

Does break-even include loan repayments and tax?

Not as normally computed. The classic break-even is an operating break-even: it covers fixed operating cost and stops there. Interest, principal repayments and tax sit below it. A cash break-even adds the cash items you actually must pay — interest and loan principal — and removes the non-cash ones, chiefly depreciation, which is why a company can be below its accounting break-even and still cover its bank standing order. A post-tax target is different again: to earn a target profit after tax, divide (fixed cost + target ÷ (1 − tax rate)) by contribution margin per unit.

My degree of operating leverage looks enormous. Is something wrong?

Probably not — it means you are close to break-even. The degree of operating leverage is total contribution ÷ operating profit, so as profit approaches zero the ratio approaches infinity. In the example here it is 4.0 at 10,000 units but 76 at 7,600 units, just above the 7,500-unit break-even, where a 1% move in volume swings profit by 76%. Read a very high figure as a warning about fragility, not as an error, and recompute it after any change to the fixed cost base.