There’s a warehouse tour on YouTube — an Australian seller with about 186,000 DVDs and CDs on the shelves — and buried in it is a question every growing store eventually has to answer.
The interviewer asks how he keeps track of what needs repricing when he isn’t the one picking the stock. His answer:
“You’ve got to remember. I generally look after that myself… I’ll notice something sells and I’ll say to one of the guys, make that back in stock. It’s probably about 15 titles I do that with.”
Fifteen. Out of 186,000.
That’s not a criticism — it’s arithmetic, and it’s the most honest answer anyone gives to this question on camera. Everything outside those fifteen sits at whatever price it carried the day it went up.
You don’t need 186,000 listings for that to bite. It bites at a few hundred. Most stores running 300 to 3,000 listings are managing the same problem with the same tool: memory, and whatever they happened to notice this week.
The bit nobody does
Listing is the part that gets attention. There are a hundred videos on photographing stock, writing titles, picking categories.
Almost nothing on what happens to that listing eleven months later.
It’s still there. It was priced correctly the day it went up — against a market that has since moved. Someone else listed the same thing cheaper, or the market cooled, or a re-release gutted it. Nobody knows which, because nobody’s looked.
And most eBay stores are one or two people. The hours go into sourcing, photographing, listing, packing, answering questions. Nobody is spending their evenings reading back through eight hundred listings one at a time to check whether the market moved under any of them. It’s not that sellers don’t know it matters. It’s that it’s the one job with no deadline attached, so it loses to every job that has one.
That’s the gap. Not pricing skill — attention. And attention doesn’t scale.
Ask around and you’ll find three things people do with stock that’s gone quiet. A few drop the price. A few more leave it and wait. Most do nothing at all, which is the same as waiting without having decided to.
Waiting is the default, and it isn’t free. Every month an item sits at a price the market walked away from is a month of capital you can’t spend on stock that would have moved. The listing isn’t costing you a fee worth worrying about — it’s costing you the thing you’d have bought instead.
And waiting only works if the price was right to begin with. If it wasn’t, you’re not waiting for a buyer. You’re waiting for the market to come back up to a number you picked eleven months ago, and it has no reason to.
So we ran it on our own store
Still In The Box carries 392 live listings, about $109,000 of ask. On 3 August we ran a price audit over the aged end of it: pull every old listing, re-comp each one against what the market is doing now rather than what it was doing when we listed, and flag the ones where those two numbers have come apart.
Eight came back needing a decision. We repriced all eight. Here’s where they stand a month on, as at 4 September.
| item | had been sitting | change | what happened |
|---|---|---|---|
| Ready Player One 4K Steelbook | 401 days | −30% | sold in 24–36 hours |
| NBA 2K21 Mamba Forever | 429 days | −80% | sold in 48 hours |
| AMD Ryzen 7 8700G | 132 days | −10% | sold in 8 days |
| Freaked Collector’s Edition | 142 days | −8% | sold in 25 days |
| Splatoon 2 Squid Sisters | — | −44% | still sitting |
| Destiny 2 Limited Edition | — | −34% | still sitting |
| Marvel vs Capcom Infinite | — | −37% | still sitting |
| LEGO Movie 2 Ultrakatty | — | −16% | still sitting |
Four of eight sold. Two within 48 hours, after sitting more than a year each.
One of those four we later had to cancel — we’d shipped the item to the wrong buyer, and that was our error in the warehouse, not the market’s verdict on the price. The reprice did its job; we didn’t do ours.
Four didn’t sell at all, and they’re in the table for a reason — a tool that only shows you its wins isn’t showing you anything. Splatoon and Destiny took cuts of 44% and 34% and still haven’t moved, which tells us something useful we’ll come back to.
It isn’t a discount button
Look at the Ryzen and the Freaked: 10% and 8%. Those weren’t fire sales. They were listings sitting a little above where the market had settled, and a small correction was enough to find a buyer in both cases.
That’s the part people expect to be wrong about a re-comp. It isn’t “mark everything down until it moves.” Most of the book comes back saying hold — the price is fine, it’s just slow, and cutting it would be giving away margin to fix a problem that doesn’t exist.
The job is finding the handful where the price and the market have quietly come apart, and saying which ones. What you do about it stays your call.
It isn’t clever. It’s a recheck.
Worth being plain about what happened here, because it would be easy to dress it up.
Those eight listings weren’t mistakes. Every one was priced sensibly against the market on the day it went up — we’d done the work. What we hadn’t done was read it again ninety days later, because the market doesn’t send you a note when it moves.
That’s all this is: a second look at a decision you already made properly, once enough time has passed that it might not be true any more.
And you don’t have to become an expert on any of it. That’s the part that actually costs a seller their evening — not choosing the price, but working out whether a title has gone cold, whether a re-release landed, whether four other people have undercut you since March. The recheck does that reading. You just get told which ones changed.
“Not selling” is two different problems
The four that didn’t move point at the thing that makes this worth automating.
Running the same book against eBay’s traffic data splits slow stock cleanly in two. Across 165 of our listings with traffic recorded over a fortnight:
- 66 listings had plenty of views and no sales. People found them, looked, and walked away. That’s a price problem.
- 14 listings were shown constantly and almost nobody clicked. One sat through 611 impressions and collected 3 views in 140 days — a click-through rate of half a percent. That’s a title and photo problem — the price is irrelevant until someone opens the listing.
Same symptom. Opposite fixes. Discount the second group and you’ve handed over margin without touching the actual fault — and you’d never know, because both just look like “it’s not selling.”
A person can hold that distinction across fifteen items. Not across eight hundred.
What it actually does
- Pull the live book — everything listed, with its age, its views and its price.
- Re-comp every line against the current market, not the market on the day it was listed.
- Split the results — hold, reprice, or that’s-not-a-price-problem.
- You decide. It surfaces the ones worth a decision. It doesn’t touch your prices.
Four out of eight isn’t a miracle. It’s $537 of stock that had been sitting an average of 276 days, gone in a month, from listings we’d otherwise never have looked at again — and it took one pass, not a Sunday.
If you’re running a few hundred listings and you’ve got a quiet suspicion that some of those prices stopped being right a while ago: that’s what this is for. Your stock, your data, your call on every one.