Every business has a number of sales below which it is losing money and above which it is not, and a striking number of owners cannot say what that number is this quarter. It is not that the calculation is hard — it is three figures and a division. It is that the calculation decays. Costs creep, prices shift, a product mix changes, and the break-even point computed in January is quietly wrong by June, while everyone continues working to the old target with total sincerity.
Break-even analysis deserves better than its reputation as a business-plan ritual performed once for a bank. Used properly, it is the fastest diagnostic in small business: it prices decisions (what does a discount do to the volume we need?), exposes drift (why must we sell more than last year to stand still?), and converts vague anxiety about a quiet month into a specific distance from safety.
This article rebuilds the calculation from its logic, deals honestly with the costs that refuse to classify neatly — because that classification, not the arithmetic, is where break-even points go wrong — and shows how to keep the figure alive rather than laminated.
The logic: one pile, one contribution, one count
Split every cost by its behaviour. Fixed costs are owed in a period whatever you sell: rent, core salaries, insurance, software, the loan payment. Variable costs arrive with each unit: materials, packaging, per-transaction fees, piece-rate labour. What remains is the contribution of a single sale — price minus variable cost — which is the amount each sale throws at the fixed pile. When the contributions sum to the pile, the business has broken even. Units needed = fixed costs ÷ contribution per unit.
With fixed costs of 6,000, a price of 45 and a variable cost of 27, the contribution is 18 and the count is 334 units (333.3 rounds up — the pile is covered by whole sales). Revenue at break-even is 15,030. Everything the business does commercially lives relative to those figures: at 500 units the period yields 334 units of coverage plus 166 × 18 = 2,988 of profit; at 250 units it loses 84 × 18 = 1,512. The model is crude and the model is powerful, in the way a map with three roads is powerful when the alternative is no map.
Classifying the costs that resist
Most costs classify themselves; the trouble lives in the hybrids. Delivery is fixed if you run your own van on salary, variable if you pay per parcel, both if you do either — so split it by its actual behaviour, not its accounting label. Labour is fixed for core staff and variable for casual hours called in per job. Energy has a standing charge and a usage curve. The rule that keeps the model honest: classify by behaviour over the range you actually operate in, and where a cost is genuinely mixed, split it into its two parts rather than voting.
The recurring fraud — usually committed against oneself — is flattering the variable bucket to lower the apparent break-even. Your own time priced at zero, “one-off” costs that recur quarterly, fees booked as overheads because they feel administrative: each moves break-even down on paper and leaves it exactly where it was in the bank. A rough break-even built on true buckets beats a precise one built on flattering ones, every single quarter.
Using the number as a decision instrument
Watch what the model does to a proposed discount. Cutting the 45 price by 10% to 40.50 drops contribution from 18 to 13.50 — and the required volume climbs from 334 to 445 units, a third more sales needed to stand in the same place. That is the promotion's real price tag, stated in the currency the business actually pays: volume. A price rise runs the film backwards — at 50, contribution is 23 and break-even falls to 261 units, buying the business a safety margin even if some volume is lost. Every commercial decision — a new channel with its commission, a cheaper supplier with a quality risk, a hire who converts variable labour into fixed salary — can be run through the same three figures and read as a movement in required volume.
The second use is diagnostic. Recompute break-even quarterly and trend it. A rising break-even with flat prices means costs are eating the business's slack; a rising break-even after a discount-led quarter means the volume bought was rented, not owned. The figure is a fever thermometer: it does not name the illness, but a business that takes its temperature quarterly rarely discovers a crisis that a monthly glance would not have seen forming.
The limits to keep in view
The model assumes one product, one price, and costs that behave linearly — a business of many products breaks even on a mix, and the mix moves. For multi-line businesses the honest approach is a break-even per line where costs can be attributed, plus a blended figure for the whole, treated as a compass heading rather than a coordinate. The model also assumes the fixed pile is truly fixed across the range in question: sell enough and you will need a second premises, a second van, a step up in staff — fixed costs climb stairs, they do not draw a horizon line forever.
None of that invalidates the tool; it defines its maintenance. Recalculate when prices move, when a cost crosses a step, when the mix shifts materially, and at least quarterly regardless. The number nobody recalculates is worse than no number, because it is believed. A break-even you have just computed, from this quarter's true buckets, is the cheapest piece of management information a small business can buy.
Key takeaways
- Break-even units = fixed costs ÷ (price − variable cost). Round up: the pile is covered by whole sales.
- Classify costs by behaviour, not label — split hybrids into fixed and variable parts rather than voting.
- Pricing your own time at zero is the classic flattering fraud; it lowers the paper break-even and nothing else.
- Read every commercial decision through the model: a 10% discount on the example raises required volume by a third.
- Recompute quarterly and trend the result — a rising break-even is the earliest reliable symptom of margin drift.
Mistakes to avoid
- Computing break-even once for a loan application and managing to it for years while costs and mix move.
- Booking delivery, energy and casual labour as fixed because the invoice arrives monthly.
- Judging a promotion by the revenue it produced rather than the volume it now requires of every future month.
- Building the model on the pre-discount, pre-fee price instead of the amount actually banked per sale.
Questions readers ask
My business sells thirty products. Is break-even meaningless?
No — but it becomes per-line plus blended. Compute contribution per line, attribute the fixed costs you honestly can, and treat the company figure as the sum of line coverages. The blended number is a compass; the line numbers are the map.
Should the owner's salary be in fixed costs?
If the business must pay it for you to live — which is the normal case — yes, at the level you actually draw. Excluding it produces a break-even at which the business survives and you do not, which is a definition of profit few owners would sign.
How precise should the variable cost be?
Precise enough to argue with. Per-unit figures gathered from real invoices beat elegant estimates; where a cost genuinely varies unit to unit, use the average of recent months and note it as an average, then review it when volumes change.
Put it into numbers
Rebuild yours from this quarter's figures — fixed pile, true variable cost, the price you actually bank — and compare the count with the one you have been managing to.
Open the Break-Even Calculator