5 min read
Refund Adjusted ROAS: Separate the Refund Date From the Acquisition Cohort
The date money reverses and the campaign that acquired the buyer answer different questions.
· Grometrics Team
The short version
A campaign looks profitable in its first week, then refunds appear after you have already increased the budget. The problem is not solved by subtracting every new refund from whichever campaign is currently active. Grometrics revenue analysis starts with the original payment and its supported source connection. Keep the refund's real date, and also retain the relationship to the customer or transaction being reversed. Those two facts let you answer different operational questions without rewriting either one.
- Use refund dates to explain current period payment reversals.
- Use acquisition cohorts to judge the later outcome of customers acquired together.
- Compare cohorts at a similar age and keep gross revenue separate from money retained.
State which question the report answers
A calendar view asks what payments and reversals occurred during the selected dates. A cohort view groups customers by when they were acquired and follows their outcomes over an observation period. Neither view is automatically wrong. The confusion starts when the spend comes from one definition and the revenue or refund total comes from another, without telling the reader that the periods refer to different things.
For example, November cash movement can include refunds for September purchases. That does not establish that November advertising acquired those refunded customers. Preserve the refund date for cash reporting and the supported original source for acquisition analysis. If your reporting tool cannot show both views, use a clearly defined export and state the calculation. Do not present a custom cohort calculation as a built in capability without checking it.
Work through an illustrative campaign
Imagine a fictional campaign that spends USD 1,000 and produces USD 1,600 in confirmed payments from a defined group of new customers. Its initial payment revenue divided by spend is 1.6. Over the following month, USD 400 is refunded against those purchases. Retained payment revenue is then USD 1,200, and that amount divided by the original acquisition spend is 1.2. The later result changes the assessment of the original group.
Neither ratio includes every business cost. Payment fees, taxes, service costs, and the treatment of disputes may require separate calculations. Call the measure what it actually is. A payment revenue ratio after refunds is not automatically gross profit or cash payback. When you label a dashboard result simply profitable, the reader may assume costs are included that the report has never measured.
Keep the transaction relationship intact
A refund should remain connected to the original payment where the provider supplies that relationship. Partial refunds need their real amounts rather than a rule that turns every refunded order into a full reversal. Multiple legitimate partial refunds must also remain distinct from repeated delivery of the same notification. The business requirement is to reflect the actual reversal once, with enough reference information to explain the result.
If the original purchase has a supported acquisition source, use that evidence when analyzing which customers later refunded. If it does not, keep the reversal in an unknown source group instead of spreading it proportionally across known channels. Proportional allocation can make totals reconcile, but it introduces an assumption about which channel caused the refund. That assumption should never be mistaken for observed customer history.
Compare customers at a similar age
A cohort acquired yesterday has had very little time to request refunds or renew. An older cohort has had more time for both. Comparing their current totals without an observation window can favor whichever group is newest. Choose a window that fits your purchase cycle and review each group at that same age. Keep still developing groups visible, but avoid ranking them as final winners before the relevant period has passed.
The right window is a business choice, not a universal benchmark. A downloadable template and an annual subscription have different customer experiences and reversal patterns. Use your own observed payment and refund timing to decide when a cohort becomes informative. Then keep that definition stable long enough to compare campaigns meaningfully, recording any changes rather than silently switching the rule when one channel looks weak.
Make the weekly decision explicit
When a source has high initial revenue and a high later refund amount, inspect the offer, audience, and expectations set before purchase. A checkout conversion improvement may be attracting buyers who were promised the wrong thing. The corrective action could be clearer qualification or more accurate product copy, rather than more traffic. Revenue quality becomes useful when it changes a concrete acquisition or product decision.
Keep a short reporting note with the spend period, acquisition definition, payment observation window, refund cutoff, currency treatment, and costs included. Stripe distinguishes cash receipt from accounting revenue recognition, so do not expect every financial report to match a payment based marketing ratio. A trustworthy explanation shows why the numbers differ and which business decision each view is suitable for.
Sources and further reading
READY TO SEE YOUR REAL NUMBERS?
Connect your payment provider and see which channels actually drive revenue.
Start tracking for free →RELATED POSTS