Every fee comparison ranks a sale. A seller with stock is ranking a listing — and a listing only pays out if it sells. Once two channels stop selling the same item equally often, the best-paying channel and the best channel come apart.
A $14 item, $1 to pack, $5 outbound postage, listed at $48 plus $6 shipping on both eBay and Etsy. Illustrative US rates: eBay 13.6% + $0.40 with processing bundled in; Etsy $0.20 a listing, 6.5% transaction, and 3% + $0.25 Etsy Payments. Returns as the returns page models them — 8% on eBay, 4% on Etsy.
| Line | eBay | Etsy |
|---|---|---|
| Order total | $54.00 | $54.00 |
| Fees | −$7.74 | −$5.58 |
| Item, postage, packaging | −$20.00 | −$20.00 |
| Keep on a completed sale | $26.26 | $28.42 |
| Expected keep after returns | $23.08 | $26.71 |
Etsy’s lower percentage leaves $2.16 more on a completed sale, and its lower return rate widens that to $3.63 once each channel’s returns are counted. Every fee calculator stops here, and on this evidence you list on Etsy.
Say you would leave the item up for 90 days before giving up, your cash is worth 12% a year, and — from your own sold-listing history — listings like this sell inside 90 days about 65% of the time on eBay, taking three weeks, and about 35% of the time on Etsy, taking six and a half weeks. Those are the two numbers to replace with your own; everything below follows from them.
A listing that does not sell is not free. It costs:
| Line | eBay | Etsy |
|---|---|---|
| Sells within 90 days | 65% | 35% |
| Days to sell when it does | 21 | 45 |
| Worth if it sells | $22.99 | $26.50 |
| Worth if it does not | −$0.41 | −$0.61 |
| One listing is worth | $14.80 | $8.88 |
| Average days on the shelf | 45 | 74 |
| Per 30 days of shelf | $9.83 | $3.59 |
Etsy keeps $3.63 more every time the item sells through and is worth about a third as much per month of shelf. The reason is not the fees at all: two Etsy listings in three are still sitting at day 90, tying up cash and earning nothing, while eBay has already turned two in three and freed the money to buy the next item.
This is the useful test, because sell-through is the number you are least sure of. Holding everything else fixed, Etsy would need to sell 71.9% of the time inside 90 days — better than double the 35% assumed, and better than eBay’s own rate — to earn as much per month of shelf. If your Etsy history is anywhere near 35%, the ranking is not close, and no plausible error in the fee rates changes it.
Neither one always. They answer different questions:
Sell-through and time to sale are numbers you supply. Nothing here measures a marketplace, and no default in the calculator is a published or observed rate — they vary more by category than by channel, and yours are in your own sold-listing history. The arithmetic is transparent so you can check it; the inputs are assumptions until you replace them.
Run these numbers on your own item →
They are different questions. What you keep per sale only pays out if the item sells, so a channel that keeps more per sale but clears one listing in three inside your horizon can be worth less per month of shelf than a cheaper channel that clears two in three quickly. Compare the two directly: weight what you keep by how often listings like yours actually sell, subtract the listing fees charged again on renewal and the cost of the cash tied up while it waits, then divide by the days the listing occupies. If you are stock-limited, take the higher rate per month of shelf. If you are demand-limited and have shelf to spare, take the higher margin.
Yes, in two ways that a fee calculator does not show. Per-listing fees are charged again each time the listing renews, so an Etsy listing that sits for a year is charged four times rather than once. And the money you paid for the item is unavailable for anything else the whole time it sits, which is a real cost whether you would otherwise buy more stock or pay down a card. Postage and packaging are not among the costs — those are only spent when something actually ships.
Your own, from your own sold-listing history, for items like the one you are pricing. Sell-through varies far more by category than by channel: a graded card and a dining table do not share a rate on any marketplace. Any published or default figure is a starting point to replace, not a measurement of your listings. If you have no history, run the comparison at the rate you believe and then at half of it, and see whether the ranking survives being wrong.