eBay Sell-Through Rate: How to Calculate STR (and What a Good One Is)

Sell-through rate answers the question every flipper asks in the aisle: if I buy this, will it actually sell? A high sold price is useless if one sold in three months and forty are sitting unsold.

Most "sell-through rate" guides are written for retail stores. This one is for eBay.

#The formula

STR = sold in the last 90 days ÷ (sold + currently active) × 100

Worked example for a used Super Nintendo console:

Some resellers use sold ÷ active instead (68 ÷ 32 = 213%). It's a different scale, not a different answer. Pick one and stick with it so your numbers are comparable.

#What counts as a good sell-through rate on eBay?

There's no official benchmark. These are the rules of thumb resellers use:

• Under 20%: lots of supply, few buyers. Only buy very cheap, or for a specific buyer.
• 20%–50%: it sells, but expect to wait. Price close to the median sold price.
• 50% and up: a steady seller. This is where most flippers want to be.

A slow item can still be worth it if the margin is big. A $4 buy that sells for $90 once a month is fine. A $40 buy that sells for $55 once a month is not.

#Why your sell-through rate is probably wrong

The math is easy. The counting is the hard part. Resellers complain about this constantly: "eBay does not make it easy to get an accurate count of active and sold items that exactly match your item." Three things distort it:

1. Loose matching. eBay pads results with "results matching fewer words." One parts seller found the "exact match" for a part number was a steering wheel for a completely different car.
2. Multi-quantity listings. One listing that sold 254 drills shows up as a single sold result in search.
3. Different tools, different counts. One reseller measured 16% with one method and 43% with another for the same item.

Fixes: search by part number, model number or exact title phrase in quotes, filter the right condition, and exclude lots. If the result list contains things that aren't your item, your STR is fiction.

#A faster check while you're sourcing: sales per week

When you're standing in a store, skip the active count:

Sales per week = sold in the last 90 days ÷ 13

68 sold in 90 days is about 5 a week. That's a fast seller, no division by active listings needed. Under 1 a week means you're buying something that will sit. Combine it with the median sold price and you have the two numbers that matter: what it sells for and how often.

#Doing it for many items at once

For one item, eBay search is fine. For a pallet, an estate or a whole inventory sheet, you want the counts without the clicking. A sold-listings API gives you the sold side directly:

curl "https://api.soldcomps.io/v1/sold?q=super+nintendo+console&condition=used" \
  -H "Authorization: Bearer $SALEHISTORY_KEY"
# → "count": 68, "stats": { "median": 89.00, "excluded": 6 }

count minus stats.excluded is your clean 90-day sold count. Divide by 13 for sales per week. Pair it with the active count from a normal eBay search (or eBay's Browse API) if you want full STR. Our Google Sheets guide shows how to put the count next to every row of an inventory list.

Want this inside an app? See building a reseller app that scans to a sold price.