Contribution Margin
Date: 2026-08-16
Revenue minus everything that varies with the sale. It’s the number that decides whether growth helps or hurts — and it’s the one most conversion work never looks at, which is how a winning test loses money.
What it is
Contribution margin is what one additional sale contributes towards fixed costs and profit, after every cost that scales with that sale.
Worked, on the model used throughout this domain — £50 average order:
revenue £50.00
− cost of goods £27.50
───────
gross margin £22.50 (45%)
− payment fees (2%) £1.00
− packaging £1.50
− fulfilment and delivery £3.00
− returns provision £2.00
───────
contribution margin £15.00 (30%)
£15, not £22.50. Gross margin stops at cost of goods; contribution margin continues through every variable cost. The gap between them is where most unit economics errors live.
Why it’s the number that matters
Fixed costs — rent, salaries, software — don’t change when you sell one more unit. So the decision on any incremental sale is made on contribution, not on revenue and not on gross margin.
Three things follow directly:
- The most you can pay to acquire a customer is their contribution margin, over whatever horizon you’re prepared to wait — Customer Acquisition Cost, Payback Period
- Break-even is fixed costs ÷ contribution margin, in units — Break-Even Analysis
- A conversion win is worth contribution, not revenue. 450 extra orders is £22,500 of revenue and £6,750 of contribution
That last one is the correction most CRO business cases need. Quoting revenue overstates the value of a test by a factor of three on this model.
Where CRO and growth go wrong
In plain terms: you can increase revenue and reduce profit, easily, and nothing in an analytics dashboard will tell you.
- Discounting to lift conversion. A 10% discount on a £50 order costs £5 — a third of the contribution — so volume must rise substantially just to stand still — Discount Impact on Margin
- Free shipping thresholds set from conversion rate alone, with the £3 fulfilment cost ignored — Shipping Thresholds
- Basket-size tactics that shift mix towards low-margin lines. Revenue up, contribution flat or down — Basket Composition
- Returns. A category with a 30% return rate has a very different contribution margin from one at 5%, and blended figures hide it — Return Rate and Reverse Logistics
- Paid acquisition judged on revenue rather than contribution, which systematically overstates efficiency
Getting the costs right
Two questions decide whether a cost belongs:
Does it scale with volume? Payment fees, packaging, per-order fulfilment, returns — yes. Rent, salaries, software subscriptions — no. Warehouse staff are the awkward middle: fixed in the short run, variable across a year.
Would it disappear if this sale didn’t happen? The cleaner test, and it settles most arguments.
Marketing is deliberately excluded from contribution margin and handled separately as Customer Acquisition Cost — because acquisition cost applies to a customer, not to each of their orders.
Using it
- Compute it per category, not blended. A 45% blended figure can be 60% in one range and 15% in another, and the second is where discounting does damage
- Convert every CRO result into contribution before presenting it — Performance and Conversion, Guide - Unit Economics and Commercial Decisions
- Watch it alongside revenue in any test with a pricing or merchandising component. Revenue up with contribution down is a real and common outcome — a guardrail worth having — Guardrail Metrics
- Re-derive it when costs move. Delivery rates, payment fees and packaging all change, and a margin model nobody has revisited for two years is a fiction