Marginal ROAS Calculator: Does the Next Ad Dollar Pay?
The Peak Presence
Meta ads · Budget decisions
Marginal ROAS Calculator: Does the Next Ad Dollar Pay?
Calculate marginal ROAS between two spend levels, set your breakeven from contribution margin, and see what a before and after comparison cannot prove.
Marginal ROAS is the extra revenue that came with extra ad spend, divided by that extra spend. Say you moved a Meta campaign from $1,000 a day to $1,500 a day and daily revenue went from $3,000 to $3,900. The extra $500 came with $900 more revenue, so the marginal ROAS is 1.8. Average ROAS at the new budget is 2.6, and that is the number most dashboards put in front of you. These figures are illustrative, not a client result.
Both numbers come from the same data and answer different questions. Average ROAS describes the whole budget. Marginal ROAS describes the last increase, which is the part you are deciding whether to keep.
Your breakeven comes from margin, not from 1.0
A marginal ROAS of 1.8 sounds profitable, because each extra dollar of spend came back with $1.80 of revenue. Revenue is not profit. If your contribution margin before ad spend is 45%, each extra dollar of revenue leaves 45 cents to pay for the ad dollar that bought it. To break even on the increase, you need a marginal ROAS of 1 divided by 0.45, which is about 2.22.
At 1.8, the extra $900 of revenue contributes $405, and the extra spend was $500. The increase lost about $95 a day while the average ROAS of 2.6 still sat above the 2.22 line. Daily contribution after ad spend fell from $350 at the old budget to $255 at the new one. An account can look healthy on average while its newest dollars lose money. This page is built to catch that case.
Contribution margin here means revenue minus product cost, shipping, payment fees, discounts and returns, before any advertising. Use your own figure, because the breakeven moves with it: at a 30% margin the breakeven marginal ROAS is about 3.33, and at 60% it is about 1.67. The same arithmetic sets a store-wide floor for MER in the ROAS vs MER guide; this page applies it to the increase alone.
Marginal ROAS calculator
Enter two spend levels over the same length of time, the revenue at each, and your margin. Days, weeks or months all work, as long as A and B cover the same number of them. Add any fixed cost that exists only because of the increase, such as extra creative you had to produce for the bigger budget. The fields start with the illustrative numbers above.
Enter your numbers
- Marginal ROAS
- 1.80
- Average ROAS
- 3.00 at A, 2.60 at B
- Breakeven marginal ROAS
- 2.22
- Marginal contribution after ad spend
- -$95
- Contribution per extra ad dollar
- -$0.19
If the whole revenue difference came from the extra spend, the extra $500 lost $95 of contribution, about -$0.19 per extra dollar, while average ROAS at B (2.60) is still above 2.22, the breakeven for the budget as a whole. The average is hiding a losing increase. Consider stepping back toward A or testing a smaller step.
Arithmetic only. It assumes A and B cover the same length of time and that the whole revenue difference came from the spend difference. Nothing you type leaves this page.
The assumption printed under the result is the part to check before you act on it. The next two sections show why.
The worked example, line by line
Illustrative numbers, the same ones the calculator opens with.
| Line | Level A | Level B | Change |
|---|---|---|---|
| Daily Meta spend | $1,000 | $1,500 | +$500 |
| Daily revenue | $3,000 | $3,900 | +$900 |
| Average ROAS | 3.0 | 2.6 | |
| Marginal ROAS: $900 / $500 | 1.8 | ||
| Contribution at a 45% margin | $1,350 | $1,755 | +$405 |
| Contribution after ad spend | $350 | $255 | -$95 |
| Breakeven marginal ROAS: 1 / 0.45 | 2.22 |
What two spend levels cannot prove
Comparing last week at one budget with this week at a higher budget is a before and after comparison, not an experiment. The revenue difference contains the effect of the extra spend plus everything else that changed between the two periods. A sale, an email send, a product drop, a stockout or a holiday can move revenue more than the budget did. Sales can also arrive days after the ad interaction, so a window that closes early makes the higher-spend period look worse than it was.
Treat the output as a planning estimate under a stated assumption, not as the incremental return of the increase. Make the comparison as fair as you can before you use it:
Before you trust the comparison
- A and B cover the same number of days, with the same mix of weekdays.
- No sale, email campaign, launch or restock landed in one period and not the other.
- Revenue means the same thing in both: net of refunds and discounts, with tax and shipping handled the same way.
- You used store revenue or new-customer revenue rather than Meta-attributed revenue, because attributed revenue can include sales that would have happened without the ads.
- You waited long enough after period B to collect its late conversions.
- Nothing else changed alongside the budget: the offer, the landing page or the ads themselves.
If one of these fails, the number still describes the gap between two periods. It is not the return on the money you added. If ROAS fell after an increase and you want to find which part of the funnel moved, the budget up, ROAS down checklist walks through it row by row.
Marginal, incremental and average ROAS
| Metric | What it divides | Question it answers |
|---|---|---|
| Average ROAS | Revenue by spend, for the whole budget | How did the budget do overall? |
| Incremental ROAS (iROAS) | Revenue the ads caused, measured against a holdout that saw no ads, by spend | How much of this revenue would not exist without the ads? |
| Marginal ROAS | Change in revenue by change in spend | What did the last increase return? |
People use incremental and marginal as synonyms, but they compare different things. Incremental ROAS compares the ads with no ads. Marginal ROAS compares one budget with a bigger one. Both need a causal estimate of revenue to be exact. The calculator substitutes the observed difference, which is why its assumption matters. Google's Meridian documentation separates average and marginal return in the same way, using a modeled response curve.
Getting closer to the real number
These designs are roughly in order of effort. Each one removes some of the noise a plain before and after comparison carries.
Four ways to measure the increase more cleanly
- Step up, then step back. Raise the budget for a period, return to the old level for a period of the same length, and compare all three. If revenue follows spend up and back down, the relationship is more believable than a single step. It is still not randomized.
- Split by region. Raise spend in some regions and hold it in comparable ones over the same dates. A sale or a holiday then lands on both groups, so the difference between them is cleaner.
- Run a Conversion Lift test. Meta compares conversions between a test group that sees your ads and a control group that does not. That estimates the incremental return of the campaign tested. It is an average for that campaign, not the next-dollar number, unless the test compares two budget levels.
- Fit a response curve. Media mix models estimate return at different spend levels from a longer history across channels. They need months of consistent data and carry assumptions of their own.
Where to read it in a CBO
With Advantage+ campaign budget, formerly called CBO, you set the budget at the campaign and Meta distributes it across ad sets in real time. The extra $500 a day goes wherever Meta sees the best opportunity. The budget decision is made at the campaign, so read marginal return there: campaign spend at A and B against store revenue over the same dates. An ad-level marginal ROAS inside a CBO mixes your budget change with Meta's allocation. The same allocation shapes creative tests: the creative testing framework shows one opening in an ErgoSleeve test taking nearly all of the delivery.
Filling it in by hand
Marginal ROAS worksheet
- Spend. Spend at A and at B over the same number of days. Extra spend is B minus A.
- Revenue. Revenue at A and at B, same definition. Extra revenue is B minus A.
- Marginal ROAS. Extra revenue divided by extra spend.
- Breakeven. 1 divided by your contribution margin as a decimal. With extra fixed costs, divide extra spend plus those costs by extra spend times the margin.
- Marginal contribution. Extra revenue times the margin, minus extra spend, minus extra fixed costs.
- Per extra dollar. Marginal contribution divided by extra spend.
- What else changed. List everything that moved between A and B besides the budget. That list decides how much weight steps 3 to 6 can carry.
Reading the result
What to do with the number
- Above breakeven, clean comparison
- The increase paid for itself in this comparison. Hold at B and measure again before the next step.
- Above breakeven, but something else changed
- Repeat the step in a quieter period, or run the step-back check, before scaling further.
- Below breakeven while average ROAS is above it
- The increase lost contribution and the account average hid it. Step back toward A, or test a smaller step.
- Revenue fell while spend rose
- Something besides the budget moved. Find it before you read any ratio.
- You want to count repeat purchases in the margin
- A higher margin input lowers the breakeven, so base it on your own customers' repeat data rather than a hoped-for figure.
Sources: Google Ads Help, Optimize for marginal ROI instead of average ROI. Google Meridian, ROI, mROI and response curves. Meta, Conversion Lift and About Advantage+ campaign budget.