eBay Product Research: How to Tell If an Item Is Worth Listing

Research is a go/no-go, not a report

You are standing in a thrift store holding a thing. You have about 45 seconds before you are just a person loitering. Here is the order that matters.

1. Does it sell at all?

Search the item, filter to Sold Items. If there are zero solds in roughly 90 days, you have no evidence. That does not always mean no — genuinely rare items sell slowly — but it means you are buying on a theory, and theories belong at low prices.

Put a hard rule on it: if there are no comps, the most you pay is what you would happily lose.

2. Does it sell fast enough?

This is the test people skip, and it is the one that fills garages.

Run the same search without the sold filter. Compare:

• 12 sold / 90 active — one in eight listings sells per quarter. Slow and crowded. Buy only at a price where sitting on it for six months is fine.
• 12 sold / 6 active — demand outpaces supply. This is the buy.
• 40 sold / 400 active — high volume, brutal competition. Fine if you are cheap and fast, painful otherwise.

Sell-through is the difference between a business and a storage unit. An item that nets $30 but takes nine months is worse than one that nets $12 in a week, because the $12 recycles four times.

3. What does it actually net?

Take the middle of the sold comp spread — not the top result — and subtract:

• 13.6% final value fee in most categories (13.25% for trading cards and coins)
• $0.40 fixed order fee on orders over $10 ($0.30 at $10 and under)
• Shipping, if you are covering it
• What you are about to pay for it

A $25 comp on a $6 item that ships for $9 in a box: $25 − $3.40 − $0.40 − $9.00 − $6.00 = $6.20. That is a real number, and it is a lot lower than "it sells for $25."

Weight is where sourcing decisions die. The same $25 comp on something that ships in a $5 bubble mailer nets $10.20 instead of $6.20 — a 65% better return on the identical comp. Two identical comps, two different businesses.

4. Can you describe it accurately?

If you cannot identify the model, read the size, or tell whether a piece is missing, the listing will be vague, and vague listings sell low and come back. Unmarked, incomplete, and unidentifiable items are not bargains — they are listings you will write badly.

The honest version of this test: would you be comfortable writing three specific sentences about this item tonight? If no, put it back.

The three walk-away signals

Counterfeit-heavy category. If the sold page is full of the same brand at prices that make no sense, you are looking at a category with an authentication problem. Someone else's problem becomes yours at case time.

The comp spread doesn't close. If genuinely similar items sold between $30 and $190 and you cannot work out why, you do not understand the item well enough to price it. Variants, completeness, and condition are hiding in there.

A single very high comp. One $200 sale and nothing else is a rumor, not a market. Widen the search. If the number does not survive a broader search, it was an outlier and you are not the one who finds the second buyer.

Categories worth understanding before you source them

Fees are not uniform. Trading cards, collectible card games, and coins run at 13.25% rather than 13.6%, and a few categories have their own structures. On a $400 sale the difference between rates is real money, and the category also controls which item specifics you will be required to fill.

If you are about to specialize in a category, spend twenty minutes reading its fee rate and its required item specifics before you buy a pallet of it.

Terapeak, and what it adds

Terapeak is eBay's own research tool, included with a Store subscription. It reaches further back than the 90-day sold filter and shows aggregate sell-through, which is exactly the metric from step 2 without you counting by hand.

It is worth the Store subscription if you are sourcing at volume. It is not worth it if you list ten things a month — the free sold filter plus the ratio check gets you most of the way.

What I actually do

I stopped doing steps 1 through 3 by hand somewhere around item eight of a cart. Not because the method is wrong — because I get sloppy, and sloppy at item fourteen is how a $6 item becomes a $40 mistake.

FlowLister does the comp pull from a photo and shows the band and the payout after fees. I still make the call. The software's job is to keep me as careful at item twenty as I was at item one.

If you source ten items a week, do it by hand. The method above is the whole thing, and it is free. Volume is what breaks human diligence, not difficulty.