Free advertising measurement tool
Refund-Adjusted ROAS Calculator
Ad platforms report revenue the moment an order is placed — before refunds and returns. Enter your spend, reported revenue, and return rate to see ROAS on the revenue you actually keep.
Refund-adjusted ROAS
Enter campaign totals and a return rate — or derive the rate from order counts.
Approximation: this path assumes returned orders have average value. See Limits.
Net revenue = reported revenue × (1 − return rate); adjusted ROAS = net revenue ÷ ad spend
Why it matters
The problem with reported ROAS
When someone buys, the ad platform logs the full order value as attributed revenue within hours. The refund — if it comes — arrives days or weeks later, and the platform rarely revises the original number. Your ROAS report therefore describes a world in which every order stuck.
Reported ROAS is correct arithmetic on the wrong revenue. The revenue you keep is reported revenue minus refunds, so the ratio that matters for decisions is computed on net revenue. A 20% refund rate does not shave a little off the top — it multiplies the whole ratio by 0.8. A campaign reporting 5× is really 4×. If your break-even ROAS is 4.5×, the platform says scale and your bank account says stop.
Refund-adjusted ROAS = net revenue ÷ ad spend
Revenue lost to refunds = reported revenue × refund rate
Worked example
Illustrative: an apparel-style store with a 25% return rate (illustrative — apparel-like — not a benchmark). Ad spend $4,000, reported attributed revenue $20,000.
Net revenue = $20,000 × (1 − 0.25) = $15,000
Refund-adjusted ROAS = $15,000 ÷ $4,000 = 3.75×
Revenue lost = $20,000 × 0.25 = $5,000
The platform reports 5×; you keep 3.75×. If this store's break-even ROAS is 4×, the reported number clears the bar while the adjusted number misses it. That gap is the entire decision — and it is invisible in the platform report.
Limits
Timing lag. Returns arrive weeks after the click. A return rate computed on this month's orders understates reality because most of those orders cannot have been returned yet. Measure the rate on a settled cohort — orders old enough that their return window has closed — and apply it to current spend.
Order rate vs revenue rate. Deriving the rate from order counts assumes returned orders are worth the average. If bargain-hunters return more, or high-ticket items return more, the revenue refund rate differs from the order return rate. The correct input is the share of revenue refunded; the order-count path is an approximation — label it as one.
Partial refunds. A 50% appeasement refund is not a full return. Fold partial refunds into the revenue refund rate: a $20 concession on a $100 order contributes 0.2 of a full return. Or compute the rate directly as refunded dollars ÷ gross dollars.
Costs are not modeled here. Restocking fees, return shipping, and non-refundable payment fees change your costs, not your revenue. They belong in your break-even inputs — see break-even ROAS — not in this calculator.
Your rate, your responsibility. This calculator does not estimate your return rate; you supply it. A stale or guessed rate produces a precise-looking wrong answer. Re-measure it when season, product mix, or sizing changes.
Key takeaways
- Gather: ad spend, reported attributed revenue, and a return rate from a settled cohort — not this week's orders.
- Produce: net revenue and refund-adjusted ROAS from the formulas above; check the platform's number against yours.
- Decide: compare the adjusted ROAS to your break-even ROAS. The reported number is not the decision input.
- Refresh: return rates drift with season, product mix, and sizing changes — re-measure the rate, don't set and forget.
Before you act
Common calculator questions
Why is my adjusted ROAS lower than the platform's number?
The platform reports gross attributed revenue and rarely revises it when refunds land. The gap between the two numbers is exactly your refund rate's share of revenue: a 20% refund rate turns 5× into 4×. Neither number is "wrong" — they measure different things. Use the adjusted one for decisions.
Should I use the order return rate or the revenue refund rate?
The revenue refund rate — refunded dollars divided by gross dollars — is the correct input. The order-count path on this page is an approximation that assumes returned orders have average value. If your returns skew cheap or expensive, compute the rate from dollars instead.
Does this handle partial refunds?
Only inside the rate you enter. Convert each partial refund into its revenue share — a $20 concession on a $100 order is 0.2 of a full return — and include it in the refund rate. Exchange-only returns that keep the revenue should not count at all.
Does MetricClue receive my campaign numbers?
Calculator inputs and results are processed locally by the page. No account, upload or server-side calculation is used. Hosting requests and any email you send are separate; see our privacy page.
Next reading
Examples are illustrative, not client results or benchmarks. The calculations use the inputs shown; real-world decisions depend on your records and assumptions. Report a correction.