ROAS calculator

How to calculate break-even ROAS

Break-even ROAS is one divided by your contribution margin: a 25% margin needs 4.00x, a 55% margin needs 1.82x, an 80% margin needs 1.25x. It falls out of the profit an ad-driven order leaves, set to zero and rearranged.

Contribution margin is what one order leaves after the costs that move with it: the goods, the shipping, the payment processing and the platform fees. A $40 product with $12 of goods, $4 of shipping and $2 of fees leaves $22, a 55% contribution margin. Shopify's break-even analysis defines the same quantity per unit, average selling price less variable cost per unit. One divided by 0.55 is 1.82 to two places, so every dollar of spend has to bring back a little over $1.82 before that campaign has paid for itself. The rest is subtraction: a 4.00x actual against a 1.82x break-even leaves 2.18x of headroom, worth $12,000 of gross profit on $10,000 of spend.

Fixed costs stay out on purpose. Rent, salaries and software do not change when you sell one more unit, so folding them in gives you a break-even for the whole business rather than for the campaign in front of you. That is a different calculation, and Shopify writes it as fixed costs divided by contribution margin per unit, which gives you a number of units rather than a multiple. One-off costs sit outside both: a trademark filing is the common one, and the trademark cost calculator prices it. Keep the variable lines separate rather than lumping them into one field, because the transaction fee scales with your price and the shipping does not. If you are filling that fee field from memory, price it first: the Stripe fee calculator gives you the figure per order at your average order value. Keep the catalog underneath them separate too: two variants sharing one code blend their costs into one row, which is what the SKU generator is for. The three figures those lines are built from each have a tool of their own: the COGS calculator for what the goods cost, the AOV calculator for what an order is worth, and the CAC calculator for what a customer costs to win.

What is a good ROAS

A good ROAS is one above your own break-even, with enough headroom left to cover the costs that sit outside the margin. That is the only definition that holds across two different stores, so this page does not give you a target number.

The rule you have probably read is 4:1, usually credited to a 2016 Nielsen study. There is no 4:1 in that study, and no universal good-ROAS rule either. Nielsen looked at more than 1,400 ad campaigns from 450 consumer packaged goods brands over a decade, so everything in it is CPG. It reports incremental sales per household exposed of $0.33 for linear TV, $0.26 for magazines and $0.25 for cross-media, which are per-household amounts and not returns on spend. The return-on-spend multiples it does publish are category benchmarks: $3.71 for baby products, $3.06 for pets. Its own conclusions are that size and scale can be stronger predictors of sales performance than even the category, and that advertisers should compare their sales metrics to brands within their product category. Which is the opposite of a single number for everybody.

What 4 actually is: the break-even ROAS of a 25% contribution margin, one divided by 0.25, the red column above. It describes your store only if your margin happens to be 25%. At 55% you clear your costs at 1.82x, and the 4:1 rule is telling you to turn off campaigns that are printing money.

For a sense of where other stores land, the closest thing to public benchmarks comes from Triple Whale, which publishes anonymized cross-brand medians from the ecommerce brands connected to it. On Meta the median ROAS is 1.88 across more than 40,000 brands for the year to July 31, 2026, with Apparel & Accessories at 2.24, Beauty at 1.54 and Health & Wellness at 1.44. On Google Ads the median is 3.68 across more than 18,000 brands for calendar 2025, which the same source reports as a decline of 10.03% year over year.

Handle those two figures with care. They are vendor-published, drawn from a self-selected sample of brands that chose to connect to one analytics product, and they report platform-attributed ROAS, the ad platform's own account of what it caused. Medians rather than averages, because the distribution is skewed. They are useful for one thing: the same campaign quality produces very different ROAS on different channels, so a target set without naming the channel is not saying much.

ROAS is not ROI

Google defines return on investment as the ratio of net profit to costs: revenue less your total cost, divided by that same total cost. The cost side there includes the advertising. Google's own worked example folds the ad spend in and comes out at ($1,200 less $800) divided by $800, a 50% return, or $1.50 back for every $1 spent. ROAS uses revenue where ROI uses profit, and that one substitution is where most ad reporting goes wrong.

Run the arithmetic. A store with a 20% contribution margin at a 5x ROAS turns $10,000 of spend into $50,000 of revenue, which leaves $10,000 of contribution, which pays back exactly the spend. A 5x ROAS, a 0% return. The same 5x at a 55% margin leaves $27,500 of contribution on the same $10,000, a $17,500 profit. Two identical ROAS figures, two different outcomes, and only the margin tells them apart. That is what the headroom row is for. ROAS alone tells you what traffic cost. ROAS against your break-even tells you whether the campaign paid.

When a knockoff takes the click

When a copied listing takes a click your ad paid to create, your spend is unchanged and only the revenue falls, so ROAS scales by one minus the diverted share. Your ads teach people what your product looks like and what it is called. Some share of them go and search that name or that image, and land on a copy of your listing instead of yours. Call that share d.

The break-even line does not move, because your costs per order are the same as they were. So the headroom is what dies first, and it dies faster than the ROAS does. At the numbers in the calculator, a fifth of the demand going elsewhere takes 20% off the ROAS and 37% off the headroom, while the number in the dashboard still reads a respectable 3.20x.

Be careful with d. It is your number, not ours, and nobody has measured it credibly for any store. The published research gives direction, not a rate. Yi Qian's study of a Chinese footwear panel in Management Science found copies cut sales of low-end authentic products through substitution while raising sales of high-end ones through an advertising effect, which puts a direct-to-consumer store well inside the substitution case rather than the advertising one. The OECD runs three substitution scenarios rather than asserting one, and the EUIPO's sector work estimates 5.2% of EU clothing and footwear sales and 8.7% of toy sales lost to counterfeiting, as econometric sector averages rather than predictions about any one store. We traced those figures in the counterfeit statistics that survive checking. The counterfeit loss calculator puts those sector shares against your own revenue, with the data year on each one.

Use the sector figures to calibrate a range for d, not to fill it in. Then read the chart as arithmetic: at 40% diversion a 4.00x campaign runs at 2.40x, most of the headroom is gone, and none of it shows up as a broken campaign. The cheap version of the check is finding out whether anybody is running listings with your product photographs on them. Knockoff watches your catalog's photos across the marketplaces and the open web and prepares the takedown you approve for each confirmed match. Run a free scan to see which listings are using your product photos today.