Why Is My Size Always Sold Out? One Retailer Counts What Nobody Took, Rack by Rack

Marks & Spencer’s women’s sale rack listed 4,225 items on the afternoon of September 20. There’s no honest way to count how many garments sit in the stockroom in a 12, because no retailer on the open web publishes unit counts, and anyone who claims to have them for a public shelf is guessing.

One number is published, though. M&S prints a count in brackets beside every value in its size filter, and that count is the number of styles currently available in that size. Not units. Styles. It’s a thin measurement and it’s the only one anybody hands out for free, so the rule here was simple: read that bracket, on every rack where it renders, on one afternoon, and change nothing else.

Five full-price racks came back, plus the shoe rack, plus the whole women’s sale rack. The jumpers rack was read twice, a day apart, to see how much a single afternoon drifts.

The usual explanation for a missing size is competition. Everyone wants the popular size, the popular size goes first, bad luck. A second explanation, popular among people who’ve been burned, is that the shop is faking it, and the shelf isn’t a record but a pressure device. Both are testable against a bracket.

Is a size really selling out, or was it never stocked that deep?

Across all five full-price racks read on September 20, the count peaks in the middle of the run and falls off at both ends, and the fall at the small end is steep. Tops listed 1,418 styles in a UK 12 and 1,078 in a 6. Dresses peaked at 1,163 in both a 10 and a 14, against 794 in a 6. Coats and jackets ran 577 in a 14 and 313 in a 6. Jumpers were the most lopsided rack of the six: 240 in a 12, 87 in a 6.

Rack, September 20	Items	6	8	10	12	14	16	18	20	22	24	26
Tops	3,557	1078	1389	1407	1418	1400	1380	1230	922	612	479	98
Dresses	2,224	794	1088	1163	1132	1163	1156	1072	729	399	333	116
Trousers	1,405	586	666	650	683	692	684	610	443	296	246	26
Coats and jackets	1,085	313	543	565	568	577	568	486	387	188	147	41
Jumpers	672	87	151	216	240	234	233	206	180	104	105	24
Women’s sale, everything	4,225	1187	1275	966	903	918	1011	993	563	401	378	178

Read the bottom row against the five above it. On every full-price rack the 6 sits below the 12, by a ratio between 0.36 and 0.86. On the markdown rack the ratio is 1.31, and the whole curve peaks two sizes lower, at an 8.

The letters do the same thing without any arithmetic. On the tops rack, S outnumbers M by 1,009 to 907, a ratio of 1.11. On the sale rack, S outnumbers M by 452 to 320, a ratio of 1.41.

An interlude on what a bracket is worth

Three limits, before any of this is used to decide anything.

The count is styles, not stock. A rack can print 1,418 in a 12 and hold one unit of each. Nothing here says how deep the pile is, and nothing a shopper can reach says it either. Don’t read these numbers as inventory.

The sale rack spans all of womenswear, including footwear, so the numbers in a 6 and an 8 include shoes. That correction can be bounded rather than waved at. M&S’s neighbouring shoe sizes on the same sale rack print 132 in a 5 and 141 in a 7, and its full-price shoe rack shows a 6 and an 8 to be ordinary sizes there, 1,757 and 1,629 against 1,814 in a 5. Subtract twice the largest neighbour, 282, from both, and the sale rack still holds 905 in a 6 and 993 in an 8, against 903 in a 12. The flip survives a correction more brutal than it deserves.

And this is one retailer on one day. M&S got read because it publishes the bracket. J.Crew, John Lewis and Gap’s US site were all opened first and none of them render a per-size count, which is its own small finding about who’s willing to be counted.

Why does the markdown rack peak lower than the shelf?

Buyers don’t order equal numbers of every size, and the industry doesn’t pretend otherwise. Rasha Mahmoud, writing for the retail training site Retail Dogma, defines a size curve as “the distribution of quantities across different sizes in the size run a retailer is carrying inventory for,” and notes that a shop starting without sales history reaches for a bell: “the middle sizes are sold and carried in higher quantities, and extreme (small or large) sizes are sold and carried in smaller quantities.” The worked ratios are 1:2:2:1 across four sizes, or 1:2:3:2:1 across five.

So the middle of the run starts deeper and still empties first. Matt Putra of the ecommerce finance firm Eightx put the sequence plainly in a June 8, 2026 post: “The core sizes, usually M and L, sell fastest. They hit 95 percent plus sell-through within weeks and effectively stock out.” He calls the curve “one of the largest controllable margin levers you have,” which is an accountant’s way of saying the ends of the run are where the money goes to die.

That’s the shape in the table. A 12 is bought deepest, sells through hardest and is gone, while a 6 is bought shallowest, sells slower, and arrives at the markdown rack in numbers the 12 never reaches.

Which disposes of the two folk explanations. The shelf isn’t running a scarcity act on the middle sizes, because a fake shortage wouldn’t leave a real surplus two sizes down. And the small sizes aren’t disappearing at all, whatever the empty rail suggests on a Saturday. They’re waiting for the discount.

What should I do with my own size?

If you’re in the middle of a size run, buy early and skip the wait, because the discount you’re holding out for is the one part of the cycle your size is least likely to reach. At either end you’ve got the opposite instruction and the better deal: the thing you want will very likely still be there in six weeks, at less money, and the risk you’re carrying isn’t price but the odds that your size was in the buy at all.

Two cautions before anyone takes that to a till.

These are UK sizes, and the shape matters more than the number, so read for where you sit in a run rather than for a 12. A low count isn’t a verdict on anything except a purchase order, which is a document about last season’s till receipts and nothing else. Sizes are inconsistent enough between brands that the label you reach for is a poor guide on its own, and one shop’s bell won’t tell you how a garment fits.

One last check on whether the shelf can be trusted at all. The jumpers rack was read on two consecutive days: 665 items and 89 in a 6 on the 19th, 672 items and 87 in a 6 on the 20th. The counter moves, slightly, in public, which is more than the words “sold out” have ever done. When a shop says an item’s gone, it’s worth remembering how rarely that claim gets checked against the deadline, and how seldom anyone will say whether it’s coming back.

Back to the 4,225, which reads at first like the size of a sale, and is exactly how a banner wants it read. Held against the five racks above it, the same number reads as a receipt: 4,225 decisions a buyer made months ago about who would be shopping, filed where the ones that missed end up.