The Return Rate Estimator measures the share of sold units or orders that customers send back during a selected period. It also estimates retained units, refund value, and net revenue after returns when sales and average order values are supplied.
Retailers can use the result to compare products, channels, campaigns, or time periods and to identify where return behavior is eroding revenue. For a useful comparison, count sales and returns on the same basis and use a period long enough for returns to be reported.
1. Choose a consistent count
Use either units or orders for both sales and returns.
2. Enter sold volume
Count completed sales from the selected period.
3. Enter returns
Use returns associated with those sales where possible.
4. Add sales and refund values
These optional amounts reveal the revenue impact in addition to the count-based rate.
5. Review both rates
The unit return rate and refund share can differ when returned products have different values.
The numerator and denominator must refer to the same measurement basis and a compatible cohort or reporting window.
The primary percentage shows how much of sold volume was returned, while the revenue metrics show the financial effect.
Late returns and exchanges can cause temporary differences between operational and financial reports.
Given: 2,500 orders sold, 180 returned, $125,000 gross sales, and $9,000 refunded.
Calculation: Return rate = 180 ÷ 2,500 × 100 = 7.20%. Refund share = $9,000 ÷ $125,000 × 100 = 7.20%. Net revenue = $116,000.
Result: The period has a 7.20% return rate and $116,000 in sales remaining after refunds.
Count them if your operational definition treats the original item as returned. Keep the same policy across periods.
Yes, but do not mix order counts with unit return counts.
Returned items may have higher or lower prices than the average sold item, and partial refunds may also occur.
A cohort view that tracks returns back to the original sale is more accurate than a simple same-month comparison.
Compare similar products, channels, customer segments, and seasons rather than relying on an unsupported universal benchmark.