Cost Per Wear: 30 Staples, and the Number Everyone Guesses

Nobody is arguing about the formula

Price divided by wears. It is unit pricing for clothes, the same idea supermarkets use when they put a price per 100g on the shelf edge, and it has one good piece of research behind it. Lisa Eckmann at the University of Bath and Lucia Reisch at Cambridge Judge ran six preregistered experiments on it, published in Psychology & Marketing in October 2025. Showing people a cost-per-wear figure moved them toward the better, more expensive garment.

Two of their findings are the ones worth carrying around. The effect only appeared when people could compare two garments' figures against each other; shown next to a single item it did nothing. And it disappeared for occasion wear, working only on things bought for everyday use.

The second finding is the uncomfortable one. The metric stops working precisely where people reach for it hardest, which is standing in front of a suit they will wear to one wedding.

Eckmann's own worked example, written up in The Conversation, is a man buying a work shirt he will wear once a week, “so roughly 50 times per year”. A £20 shirt that lasts a year costs 40p a wear. A £50 shirt that lasts four years costs 25p. She also notes that durability could be measured properly, by the abrasion cycles a fabric survives, and that no shop does it.

Seven guides, and not one real price

So the concept is sound and the research is real. Then you read what gets published about it. I read the seven cost-per-wear guides ranking for the term in September 2026, in full, and counted the garments in them you could go and buy.

Zero. Not one names a brand, a product or a price you could check. Every example is a round number holding a slot: a $100 coat, an $80 pair of jeans, a $25 top, a $150 dress. Three of the seven use exactly $100. The two published a day apart, at AOL and Woman's World, run the same five examples, the same two expert quotes and the same industry statistic, and the first of them says at the foot that it was made with AI.

Credit where it is due: one of the seven works from a wardrobe she logs daily in a closet app, which is the only measured denominator on the list, and she is careful to say a good figure is personal rather than universal. She is also the one with the least reach.

This is a metric whose entire value is that it replaces a feeling with a number, and the genre that promotes it has been running on made-up numbers for a decade. That is the gap worth filling, and it is the same job as reading twelve nightlife dress codes end to end rather than repeating what everyone says they contain.

The one number somebody has measured

Half of cost per wear is public: the price is on the tag. The other half needs to know how long a garment survives, and there is exactly one serious dataset on it, which the entire cost-per-wear genre appears never to have opened.

WRAP, the UK climate NGO, surveyed 6,000 people about 44,807 individual items of clothing in late 2021, asking of each how long ago it was acquired and how much longer it would be worn. Coats came out at 6.3 years, jackets at 6.1, jeans at 4.1, T-shirts and polos at 4.0, and clothing overall at 4.3. Every one of those is longer than the 2013 figures. A repair adds 1.3 years on average.

There is one thing WRAP measured that changes how you read the rest: a quarter of what people own, 31 items out of an average wardrobe of 118, had not been worn at all in the past year. Clothes are not failing early. They are sitting still.

How the numbers below were worked out

Wears equals how often you wear a thing multiplied by how long it lasts. WRAP supplies the second term. For the first, the rate is Eckmann's own, once a week or 50 wears a year, applied to every piece.

That is deliberately the best case. A garment in genuine weekly rotation for the whole of its measured life is the ceiling, not the average, and almost nothing you own will hit it. The ceiling is the number worth knowing anyway: if a piece fails the test at its best case, it fails.

Two limits, stated rather than buried. WRAP surveyed clothing, not footwear, so the five shoe rows below borrow the all-clothing average and are the softest figures on the page. And WRAP's respondents are in the UK while the prices are US, which is the kind of seam that exists in any calculation combining two honest sources and is better named than smoothed over.

Thirty staples, priced and divided

Thirty pieces that a man building a wardrobe would recognise, across nineteen houses, every one a real listing in the MVRCK catalog, which is mine, at a real price. No piece appears twice. They are ranked by cost per wear, most expensive first, with their position in the same thirty ranked by price sitting in the last column.

The spread of prices is 89 to 1. The spread of cost per wear is 64 to 1. Dividing by the best lifespan data available compresses an 89-fold difference into a 64-fold one, and that is the whole of what the metric achieves here.

Sorting by cost per wear barely changes the order

Read the last column downward. It runs 1, 2, 3, 4, 6, 5, 8, 11, 9, 7 and carries on like that to the bottom. Reordering thirty garments by cost per wear rather than price moves no piece more than four places.

The reason is arithmetic, not taste. The denominators available to these thirty pieces run from 200 wears to 315, a spread of 1.6 to 1, because that is how little garment lifetimes differ between the longest-lived and shortest-lived categories on this page. The prices run 89 to 1. When one input varies by 89 and the other by 1.6, the first one decides the answer and the second is a rounding adjustment.

The furthest mover in the whole set is A.P.C.'s trench coat, which drops four places down the ranking. On a list running from worst to best that means it comes out four places better than its price alone would put it, and it earns that by being a coat, the category with the longest measured life of any. Three more pieces move three places. Nothing else moves more than two. That is the metric's entire corrective power over a price tag.

Which is worth holding against how cost per wear is usually sold, including by people arguing in good faith: as the calculation that reveals the expensive thing to have been the cheap thing all along. On real prices it mostly reproduces the price list. It is the same pattern as the suede loafers that cost less than the sneakers they dress up, in reverse: the relationship people assume between money and value in a wardrobe is weaker than it feels in both directions.

The denominator is the whole game

So if the category barely matters and the price mostly just reasserts itself, where does cost per wear move? Into the only variable left, which is how often you take the thing off its hanger.

Here is one garment from the table, at one price, under four honest patterns of use.

A 43-fold spread, on one garment, with the price held still. The entire thirty-piece table above spans 64-fold. Almost the whole range of that ranking is reachable inside a single $125 shirt, purely by how often it gets worn.

Put the two ends of the table against each other and it stops being abstract. Loro Piana's cashmere sweater at $2,070 comes to $9.63 a wear if you live in it. The Ralph Lauren shirt at $125, worn once a month for a year and then forgotten, costs $10.42. The sweater is sixteen times the price and cheaper every time you put it on.

Four things the arithmetic does not support

None of that makes cost per wear useless. It makes it a comparison tool rather than a justification tool, which is what the research found and what almost nobody uses it as. Run it across two shirts you are choosing between and it will tell you something. Run it on one thing you have already decided to buy and you will produce whatever answer you went in wanting, because you are the one supplying the denominator.

What decides whether you reach for it

The denominator is not about discipline, and telling people to wear their clothes more has never worked. A garment gets worn when there is an outfit waiting for it. The pieces that rack up wears are the ones that finish four different looks with things already in the wardrobe; the ones that stall are the ones that finish exactly one, which is why WRAP's respondents named “for occasions only” as a leading reason for the 31 items they had not touched all year. Counting those looks before you buy is its own exercise, and on a real twelve-piece wardrobe only a fifth of the combinations hold up.

That is the count worth having before you buy anything, and it is the one nobody can do in their head, because it depends on the other forty things hanging up. It is what I built MVRCK to do. Photograph what you own once and it becomes a wardrobe the stylist can see, so the question stops being “will I wear this a lot” and starts being “here are the six outfits this shirt completes with pieces you already have, and here are the two the other shirt completes”. It remembers your sizes, brands and budget, you can see any look on your own avatar before spending anything, and the catalog behind it is real stock at real prices, which is where the thirty figures above came from.

Cost per wear asks you to predict the future. This is the closest thing to checking.

Common questions

What I would do with all this: stop using cost per wear to decide whether to buy something, and start using it to decide which of two things to buy. The first job is a rationalisation with arithmetic on top. The second is the only one it has ever been shown to do. And if the two figures come out close, which on real prices they usually will, buy the one that goes with more of what you already own.