Every SKU with what it costs us and what it sells for. Loaded from the cost sheet, not typed again.
Source:
Chairs read sound: 3 of 399 show a cost above the price. Every other prefix inherited a chair's First Cost through a fill-down, so a $29.99 blanket carries a $75.00 cost and a $27.99 cover carries $70.00. That is 95 of 305 accessories showing as sold below cost, and it is a spreadsheet artifact rather than a business fact. Anything marked inherited should not be priced off until the real cost is filled in.
The sheet charges every SKU 10% of its price for refunds. Real refunds ran about a quarter of that.
Measured from Shopify orders and refunds, . Amazon is not in this data.
It is the same thing the sheet's line item is, so the two compare directly. It does not include the outbound shipping already spent, the freight to bring something back, or product that comes back unsellable. The all-in cost of an actual return is much higher and lives on the Tickets page. Use this one for pricing, not for deciding whether to take something back.
Where the real spread is. Covers return at roughly double the rate of filled chairs, and Bushon at several times Juno. That is the signal a flat 10% throws away. Families marked thin are too small to trust on their own and fall back to the catalog rate.
| Family | SKUs | Units | Revenue | Refunded | Rate |
|---|
Measured on catalog rows. Averaging the per-SKU rates straight gives , which runs lower because the long tail of never-refunded SKUs each count once. The revenue-weighted number above is the one to quote.
| Prefix | SKUs | Median cost | Median price | Below cost |
|---|
Refund reserve columns: sheet is the flat 10% it charges today, real is what that SKU measures at. A single SKU sells too few units a year to trust on its own, so the rate is pulled toward its family until the SKU has earned its own history. The last column says which one you are looking at.
| SKU | Kind | Cost | Old price | New price | Margin | Refund: sheet | Refund: real | Rate | Based on |
|---|