Cost per wear, done properly

$285 divided by 260 wears is $1.10. That is the whole method, and everything else in this piece is about the two numbers on either side of the division sign — which is where the argument always actually is.

Cost per wear is the only calculation that makes a $125 cord necklace and a $1,695 occasion dress directly comparable. It is also the calculation most often used to justify a purchase rather than test one, because the wear count is supplied by the person who already wants the thing. Done badly, it is a story with a number attached.

Our calculator at /tools/cost-per-wear runs the arithmetic below exactly as written, server-side, with no affiliate link on it. Every figure in this article was produced by that tool, and every price comes from a brand's own published catalogue as recorded in our brand files on 4–5 September 2026.

The formula

Total wears is wears per year multiplied by years kept, rounded to a whole number. Cost per wear is the purchase price divided by total wears. That is it.

The $285 example above is Finley Shirts' Sloane Shirt in white silky poplin, worn once a week for five years: 52 × 5 = 260 wears, $285 ÷ 260 = $1.10 a wear. Run it yourself.

The verdict band is fixed, and it keys on the net figure after resale rather than the gross one:

• Under $2.00 a wear — Buy at full price. The wear count has done the work and the price does not need a discount.
• Under $10.00 a wear — Buy on sale. The piece is defensible, but a markdown, or more wears, is what closes the gap.
• $10.00 and above — Skip. You are renting it.

The boundaries are strict. A piece landing at exactly $2.00 falls into the second band, not the first.

Two failure modes account for almost every misleading cost-per-wear figure you will read. The first is inflating wears per year: a dress worn to four weddings a year is not a dress worn weekly, and the tool's default of 52 is once a week, not "often". The second is inflating years kept: five years is a long time to keep wearing one shirt, and ten is a claim about the future, not a fact about the garment.

Correction one: resale recovery

The purchase price is what leaves your account. It is not necessarily what the piece costs you, because some of it may come back when you sell.

The tool handles this with one field: resale recovery, entered as a percentage of the purchase price. Recovery is subtracted from the price to give a net cost, and the net cost is what gets divided by total wears. Take the Anna October Clementine Satin Maxi Dress, listed at $850 at Moda Operandi — the one price in our Anna October file that is verifiably quoted in US dollars, because the brand's own storefront prices and charges US shoppers in euros. Worn four times a year for five years, that is 20 wears and $42.50 a wear. Assume you later recover 30% and the recovery is $255, the net cost is $595, and the net figure is $29.75. Both are a Skip. Resale does not rescue a twenty-wear piece; it just makes an expensive mistake slightly cheaper.

Note what that 30% is: an assumption you supplied, not a finding. The field starts at zero on our tool and stays there deliberately, because we have no sourced resale-recovery figure for any of the fifteen brands we cover. Recovery varies enormously by label, category and condition. On lab-grown diamonds it is close to nothing — Fiametta's file records that lab-grown stones carry little resale value and that no secondary-market data exists for the brand at all. Nomasei appears on The RealReal only through resale and consignment, which is not authorised retail and not a price the brand sets.

What we do have are markdowns, and they are instructive even though they are not resale. Heirlome's Elaine Top carries $590 full retail and is listed at $177 by the wholesale partner Shop Zoe — 30% of retail, in an authorised channel, while the piece is still current. Roxanne Assoulin's Camp Bracelets are marked $45 on the brand's own site against a $75 full price, and are final sale at that. When the first-hand market is already discounting a piece that hard, a second-hand recovery assumption above it is optimism with a percentage sign on it.

Use zero unless you have a comparable listing you can point at. A recovery percentage you invented is not a correction; it is a rebate you have written yourself.

Correction two: alteration and resizing

Not one of the fifteen brands in our files publishes what an alteration costs. That absence is the finding, and it is why the tool has no alteration field: there is no honest default to preload. You apply this correction by hand, by typing the all-in figure into the price box rather than the sticker price.

What the brands do publish is enough to know the cost is real. Finley Shirts runs a Perfect Fit Credit worth up to $50 in store credit toward alterations on a full-price website purchase, when the alteration receipt is uploaded within thirty days — a ceiling the brand set itself, on shirting whose core white block runs $190 to $285. Markarian offers no in-house tailoring at all, recommends outside tailors, and states that its made-to-order pieces "often require on body tailoring for desired fit"; on a $1,695 midi dress that is a line item, not a footnote.

On jewelry the correction runs the other way, as a penalty rather than a fee. Liori Diamonds charges a 15% restocking fee on sized rings — against its own sampled catalogue median of $2,480, that is $372 to undo a sizing decision. AMANDA PEARL's repairs take four to six weeks and are chargeable unless the damage is a manufacturing defect. Fiametta's wear-and-tear repairs run four to six weeks. Noémie's lifetime warranty against manufacturing defects is voided if the piece is taken to an outside jeweller, so the convenient local bench is not free either — it costs you the warranty.

The practical rule: if a piece will need a hem, a resize or a strap shortened before you can wear it, add the quote to the price before you divide. A $285 shirt that needs work is not a $285 shirt.

Correction three: the wears you never get

The third correction is the one nobody applies, and it falls straight out of our survey of what fifteen luxury brands actually give you. A short return window and a final-sale rule do not change the price. They change the probability that you are stuck with the price.

Across our fifteen covered brands, thirteen give you fifteen days or fewer to change your mind, and the modal window is fourteen days. Several publish an explicit deduction on top. Nomasei withholds 20% of the price on a return showing slight try-on or wear marks and a further 10% if the item is not repacked in its original box: on the $850 STRADA suede ankle boot, that is $170, or $255 for a boot returned without its packaging. Roxanne Assoulin deducts a $10 handling fee from every refund — 8% of the $125 Cylinder Cord Necklace. Markarian's shipping-and-returns page applies a $25 restocking fee. Anna October charges EUR 40 for a US return against a EUR 725 dress, about 5.5%, and prices in euros throughout.

The arithmetic that follows is straightforward. A piece you cannot return has a floor of one wear, and one wear is the worst cost-per-wear there is. The $1,695 Markarian midi dress at four wears a year for five years is $84.75 a wear; the same dress worn once, because it did not fit and the fourteen-day full-refund window closed, is $1,695 a wear. So the correction is not a number you subtract. It is a question you answer before you order: if this is wrong, how many days do I have, what comes off the refund, and is it final sale?

What the honest numbers look like

Four worked examples, all real catalogue prices, all run through the calculator at the stated assumptions.

The pattern is the point. Price barely predicts the band. The most expensive piece on the list is not the worst buy because of its price; it is the worst buy because occasionwear is a four-wears-a-year category, and four wears a year cannot carry four figures. The $1,065 ring is the best figure on the board because a plain 14k band is worn on ordinary days, and ordinary days are what the denominator is made of.

Push the Nomasei boot to eight years at the same forty wears a year and it moves to $2.66 — still Buy on sale, not Buy at full price. Three extra years of ownership move the figure less than most people expect. Wears per year is the lever; years kept is not.

Where the method stops working

Cost per wear values a garment as a machine for producing wears, and some purchases are not that. A bridal gown has a wear count of one by design, and dividing by one tells you nothing you did not already know. The same goes for a gift, where the wearer's habits are not yours to estimate.

It also says nothing about whether the thing is any good. A $1.10-a-wear shirt made badly is still a badly made shirt; the arithmetic is silent on fabric, construction and fit, which is why every verdict on this site names the material spec next to the figure.

The Pennyweight Verdict

Before the call, the terms: Finley Shirts gives 30 days from receipt, but you must contact the brand first for a prepaid label or the parcel is refused, shipping and handling are non-refundable once shipped, and clearance and promo items are final sale — the 30-day window covers full-price stock only.

Buy at full price — the Finley Sloane Shirt at $285. At once a week for five years it is $1.10 a wear, which is inside the under-$2.00 band without needing a markdown, and the wear count is the ordinary kind rather than the aspirational kind. Two things to know before you divide: the signature silky poplin is 60% cotton, 35% polyester and 5% spandex rather than an all-natural cloth, and several styles including the Crosby are flagged by the brand as running large. Buy the size that fits and the figure holds. Buy the size you hope to be, and the denominator collapses.

Prices are the brands' own published figures as recorded in our brand files on 4–5 September 2026 and checked against the calculator on this site. Wear counts and resale percentages in this article are stated assumptions, not measurements. The Pennyweight buys nothing, accepts no gifted product, and is not affiliated with any brand named here.