Refund Rate
How many of the units you sold came back.
What it means
Refund Rate is the units Amazon refunded in the period divided by the units ordered in it, shown as a percentage. It counts units rather than euros, so a returned €5 add-on weighs the same as a returned €200 item. Lower is better: every extra point is stock coming back and a charge you don't get back.
Show the math
Formula and a worked example
Units refunded is every unit Amazon processed a refund for during the period, whenever it was ordered.
Units ordered is the units customers ordered in that same period — a different set of orders entirely.
Worked example. In March customers order 4,000 units and 260 are refunded. Refund Rate = 260 ÷ 4,000 = 6.5%. At €35 a unit that's €9,100 of sales reversed, plus the refund administration Amazon keeps on each one.
Because the two halves cover different orders, the rate wobbles whenever volume does. Double your orders and most of this month's refunds are still for last month's smaller batch of orders, so the rate reads better than the product deserves. After a peak it reads worse. Compare like months.
It answers the question
Of everything shoppers ordered, how much did they send back? A rising rate means the listing is promising something the product doesn't deliver.
Why it matters
A refund costs more than the sale it cancels. The unit comes back needing inspection or disposal, and you've paid to move it in both directions — so Net SalesNet SalesSee the full entry. drops while the refund administration charge lands in FeesFeesSum of commission, fulfilment, closing, storage, service, refund admin, shipping label, statutory deduction, and related fee fields for the row..
It's also the number Amazon itself watches. Return rates that run high for a category put the listing and the account at risk, and no amount of ad spend fixes a product people send back.
What good looks like
There's no universal target — apparel returns at rates that would be alarming for consumables, and an expensive item tends to come back more often than a cheap one. Judge each SKU against its own last quarter, then against the rest of your catalogue by sorting the column; the outliers are where the listing over-promises. A rate that climbed after a listing edit or a supplier change is the one to chase today.
How to improve it
| Lever | What you do | Expect | How long | Watch out for |
|---|---|---|---|---|
| Fast Fix the size and fit information | Add real measurements, a chart and true-to-life photos on the SKUs with the most refunded units | Refund Rate down 1–3 points on those listings | 2–4 weeks | Fewer shoppers talk themselves into it. Orders and page conversion fall before the rate improves. |
| Fast Stop advertising the worst offenders | Pause campaigns on SKUs whose returns run far above the rest of the catalogue | Ad budget stops funding returns and the account rate falls | 1–2 weeks | Rank decays while those listings are dark, and buying the position back costs more than holding it did. |
| Slow Tighten inbound quality checks | Inspect at the supplier for the defect behind your most common return reason | Refund Rate down and fewer damaged units coming back | 1 quarter | Inspection adds cost per unit and delays shipments, so stock cover gets thinner on your fastest lines. |
| Slow Retire the products that always come back | Discontinue SKUs whose returns eat their margin, however well they sell | Account-wide Refund Rate down | 1–2 quarters | You lose their revenue and their traffic. Related listings that fed off their visitors usually dip too. |
Every lever costs something somewhere. The last column is the one to read twice.
Read it with
Refund Rate tells you what came back. Next to what went out, it tells you whether that's a growing problem or a steady cost of selling.
Growing cleanly
More units going out and a smaller share coming back. This is the shape that compounds — every point off the rate is margin you keep on a bigger base.
Scaling the problem
You're selling more of something people return. Volume multiplies the cost, and a rate that stays high for the category eventually threatens the listing itself.
Quieter and cleaner
Fewer units and fewer returns. If you paused a problem SKU on purpose this is working; if you didn't, the improvement may just be a slow month.
Fewer units, worse returns
The rate is climbing on a shrinking base, which usually points at one cause rather than a trend — a bad batch, a changed photo, a new claim in the bullets.
Net Sales already has refunds inside it, so a soft month looks the same whether nobody bought or everybody returned. Put the rate beside it and the two separate: a flat rate means demand, a rising one means the product. That decides whether you touch campaigns or listings.
Views tell you how many shoppers the page convinced; the rate tells you how many it convinced wrongly. Rising views alongside a rising rate is the signature of a listing that sells better than the product delivers — usually a new main image or a bullet the item can't live up to.
Common misreads
It's a unit count. 3% on your most expensive SKU can be a bigger hole than 15% on a cheap one. Check the euros behind it in Net SalesNet SalesSee the full entry. before you move on.
Refunds land when they're processed, not when the order was placed. A big sales month followed by a quiet one pushes the rate up with nothing at all wrong at the supplier.
Not on its own — a listing nobody buys from has a wonderful refund rate. Read it beside the units shipped, and treat a near-zero rate on a low-volume SKU as too little data rather than a win.
Also called
Return Rate · refund percentage · unit return rate
See yoursYour Refund Rate by date, next to the units that went out.
Open Amazon Summary →