Contribution margin is the revenue left from a sale after you subtract every cost that rises with that sale, such as the product itself, shipping, payment fees and returns. It’s usually expressed as a percentage of revenue, and it’s the pot of money that has to pay for your marketing, your fixed costs and your profit. That’s why it sets your breakeven ROAS: if each dollar of revenue leaves you 40 cents of contribution, a dollar of advertising has to bring in $2.50 of revenue before it pays for itself, so breakeven ROAS is simply 1 divided by your contribution margin.
When I ran a growth marketing agency, the first thing I needed from a new client was their margin, before any talk of channels or creative, because without it a ROAS number means almost nothing. A 3x return is excellent for a software company with 80 percent margins and a disaster for a retailer with 25 percent margins, and the dashboards both of them look at every morning will show the same green arrow. In the course I put it as “a strong revenue multiple can still lose money,” and this page is about the arithmetic that tells you when.
What contribution margin is, and how it differs from gross margin
The contribution margin formula is revenue minus variable costs, and the contribution margin ratio is that number divided by revenue. Variable costs are the ones that scale with each extra order: the cost of the goods, picking and packing, shipping, payment processing, the refunds and handling costs of returns, and sometimes a marketplace commission or sales royalty. Fixed costs, such as salaries, rent and software subscriptions, don’t change when you sell one more unit, so they’re left out. The Corporate Finance Institute’s definition is a good plain reference if you want to check your finance team is using the same one.
Gross margin is the number most people already know, and it isn’t the same thing. Gross margin only subtracts the cost of goods sold, so a DTC brand with a 65 percent gross margin might only have a 45 percent contribution margin once shipping, fees and returns come out. Using gross margin in that example credits each revenue dollar with about 40 percent more contribution than it really earns, and drops the breakeven ROAS from 2.22 to 1.54, and I’ve seen that exact mistake justify a lot of spend that was quietly losing money.
The other trap is the opposite one: subtracting advertising twice. Many ecommerce finance teams label their margins in layers (some call them CM1, CM2 and CM3), where the last layer already has marketing costs taken out. That’s a useful view of the business, but it’s the wrong input for breakeven ROAS, because the whole point of the calculation is to compare the margin with the advertising cost. What you want is contribution margin before advertising. Labels vary from company to company, so it’s worth agreeing the definition with finance in writing before anyone builds a target on it.
How to calculate contribution margin for an ecommerce order
Let’s say you sell an $80 order. The product costs you $28, shipping and fulfillment cost $9, your payment processor charges 2.9 percent plus 30 cents, and 8 percent of revenue is refunded through returns. Assuming returned products go back into stock but you don’t get the shipping or the processing fee back, the calculation per order looks like this:
| Line | Per order |
|---|---|
| Gross order value | $80.00 |
| Refunds (8% of revenue) | −$6.40 |
| Product cost on kept orders (92% × $28) | −$25.76 |
| Shipping and fulfillment | −$9.00 |
| Payment fees (2.9% × $80 + $0.30) | −$2.62 |
| Contribution before advertising | $36.22 |
| Contribution margin (÷ $80) | 45.3% |
I’ve measured the margin against gross order value, not net revenue after refunds, on purpose. Ad platforms report ROAS on gross conversion value, and your model is likely to be fitted on gross revenue too, so the margin has to be on the same base as the ROAS you’re going to compare it with. If your revenue numbers are net of returns, calculate the margin on net revenue instead; the rule is just to be consistent.
Why breakeven ROAS is 1 divided by contribution margin
If an ad dollar brings in dollars of revenue and each revenue dollar leaves of contribution, the ad dollar produces of contribution, and it breaks even when that equals the dollar you spent. So breakeven ROAS is , and the equivalent for cost per acquisition is that breakeven CPA equals contribution per order. In the example above, breakeven ROAS is and breakeven CPA is $36.22.
| Contribution margin | Breakeven ROAS |
|---|---|
| 20% | 5.00 |
| 25% | 4.00 |
| 30% | 3.33 |
| 40% | 2.50 |
| 50% | 2.00 |
| 60% | 1.67 |
| 75% | 1.33 |
The course’s invented Northstar case shows how easy it is to be fooled. Northstar estimates £250,000 of incremental revenue from £100,000 of advertising, a 2.5x return. If 30 percent of that revenue remains after product and fulfillment costs, that’s £75,000 of contribution to pay a £100,000 ad bill, so the campaign loses £25,000. At a 40 percent margin the same 2.5x exactly breaks even, and at 50 percent it makes £25,000, which is three different answers from the same campaign and the same ROAS.
Find profit-maximizing spend on a response curve
Incremental revenue = a × spend / (k + spend), in $k per period. Profit after ads = contribution margin × incremental revenue − spend.
Try entering your own numbers. Shipping and returns usually move the result more than people expect, and a product with a low price point often has a much higher breakeven ROAS than the gross margin suggests, because fixed per-order costs like shipping take a bigger bite out of a small order. The repeat orders field shows what happens when you count repeat purchases, which I’ll come back to below.
Target ROAS: building profit into the number
Breakeven is the floor, not the goal. If you want advertising to deliver a profit after its own cost of, say, 10 percent of the revenue it drives, the arithmetic becomes , where is the profit share you want to keep. With a 45.3 percent margin and a 10 percent profit target, the target ROAS is . Notice how quickly that climbs as the profit target approaches the margin: if you asked for 40 percent profit on a 45 percent margin, you’d need a ROAS of 20.
Be careful when you plug these numbers into a platform’s target-ROAS bid strategy. Those strategies optimize against the platform’s own attributed conversion value, which is not the same as incremental revenue, so a target set at your true breakeven will usually let the platform spend on conversions you would have had anyway. In practice, teams that use platform targets tend to set them higher than breakeven by a factor they’ve calibrated from an incrementality test or a model.
Why incremental revenue, not attributed revenue, goes in the calculation
Breakeven ROAS answers a causal question: does this advertising pay for itself? So the revenue in the numerator has to be revenue the advertising caused. Platform-attributed revenue includes customers who would have bought anyway, such as people who searched for your brand name and clicked an ad on the way to a purchase they’d already decided to make, and it double counts customers that several platforms all claim. If a campaign reports 4x in the ad platform but only drives 2x incrementally, comparing the 4x with a breakeven of 2.5 says “profitable” when the truth is “losing money.”
The incremental number has to come from somewhere that can separate cause from coincidence, usually a marketing mix model or an incrementality test. That’s also where MMM adds most value for profit decisions, because it can estimate incremental revenue across every channel on the same basis. The incrementality guide goes into the gap between reported and incremental results, and the marketing efficiency ratio page covers the blended alternative many DTC operators use when they don’t trust platform numbers.
Breakeven for subscriptions and repeat purchases: LTV and payback
For a subscription business, or any brand where customers come back, the first order isn’t the whole story. A new customer who places two more orders over the next year brings three orders’ worth of contribution, so the breakeven CPA on a 12-month view is three times the first-order contribution, and breakeven ROAS measured on first-order revenue drops to a third of its first-order value. That’s the logic behind acquiring customers at a loss on the first order, and it’s legitimate as long as the repeat behavior is real.
The two rules I’d hold to are to use contribution lifetime value, not revenue lifetime value, and to make the evidence for it as strong as the evidence for the ROAS. A revenue LTV of $300 at a 45 percent margin is $135 of contribution, not $300. And as the course puts it, a lifetime-value assumption needs its own definition and evidence; it isn’t a free adjustment that makes every campaign profitable. Customers acquired through a heavy discount or a new channel often repeat less than your existing base, so I’d use cohort data for the channel in question where you have it, and a shorter window (12 months rather than “lifetime”) where you don’t.
Payback period is the cash version of the same idea: how many months of contribution it takes to earn back the cost of acquiring a customer. A business can be profitable on a 24-month LTV and still run out of cash if it scales acquisition faster than payback can fund it, which is why many subscription businesses set a target payback period alongside a target CPA. If your finance team discounts future cash flows, apply the same discounting to the LTV that goes into the breakeven.
Finding the profit-maximizing spend on a response curve
Breakeven ROAS tells you whether advertising pays on average, but it doesn’t tell you how much to spend. For that you need a response curve, because each extra dollar of spend brings in less incremental revenue than the last. With incremental revenue and contribution margin , profit after advertising is , and it peaks where its slope is zero, which is where . In words: keep spending until the marginal ROAS, the return on the next dollar, falls to your breakeven ROAS. For the curve used in the calculator, that spend is .
This is the most important point on the page, and it’s often missed. At the profit-maximizing spend, average ROAS is still far above breakeven. In the calculator’s default example, with a curve of and (in $k) and a 45.3 percent margin, profit peaks at about $113k of spend, where average ROAS is 4.70 and marginal ROAS is exactly the breakeven of 2.21. If instead you kept scaling until average ROAS fell to breakeven, you’d end up at about $353k of spend, where total profit after advertising is zero, because the last $240k or so of spend lost all the money the earlier dollars had made. “Scale until ROAS hits breakeven” is a common rule, and it’s a rule for giving all your profit back to the ad platforms.
The same logic extends to several channels: each channel should get spend until its marginal ROAS reaches breakeven, and if the total budget is capped below that point, you split it so the marginal returns are equal, which is what a budget allocator does. When I review a budget plan, the first thing I check is whether the marginal ROAS at the proposed spend is above or below breakeven for each channel, because that one comparison catches most of the plans that look good on revenue and bad on profit.
What breakeven ROAS leaves out
Contribution after advertising is not accounting profit. It ignores fixed overheads, tax and financing costs, so a positive number means advertising is helping to cover those costs, not that the company is profitable. It also ignores cash timing: an order whose contribution arrives after the inventory has been paid for and the ad bill has cleared can still strain cash flow. And it treats advertising as if it only creates sales within the measurement window, which undervalues activity whose effect builds slowly. If your model captures longer adstock or brand effects, include them in the incremental revenue; if it doesn’t, say so in the decision note rather than quietly assuming them away.
If you want to see the full chain, from response curves to a budget recommendation that finance would sign off, the MMM scenario planning page and the Northstar decision case in the Vexpower MMM course put margin, uncertainty and constraints together in one decision. The old budget optimization brief covers the response-curve side if you want a simpler starting point.
