Resale
Resale programmes are usually scoped as a demand question: is there a market for our pre-owned product, and which platform should carry it. That question was settled some years ago, and answering it correctly does not produce a profitable programme.
The binding constraints are supply and verification. Supply, because the only economic source of inventory is your own returns, and returns arrive without identity. Verification, because the cost of establishing that an item is genuine is set by your category rather than your strategy, and is routinely omitted from the business case.
On 19 July 2026 the cost of ignoring both rose. What follows is the argument, with the evidence and the assumptions it rests on.
1. Demand
The BoF–McKinsey State of Fashion 2026 forecasts secondhand growing two to three times faster than first-hand through 2027. OC&C puts the UK second-hand fashion market above £7 billion, with close to one in four fashion transactions involving resale. Barclays recorded 38% of UK consumers buying from a resale platform in the preceding twelve months.
The behavioural finding matters more than the market sizing. 35% of UK buyers now consider an item's resale value before purchasing it new (Vinted Impact Report, 2025); in the US the figure is 60%, up thirteen points year on year (ThredUp Resale Report, 2026).
Where a material share of customers price expected residual value into the purchase decision, secondary market liquidity becomes an input to primary willingness to pay. Your resale market is therefore already affecting your full-price performance, whether or not you operate in it.
Participation is not the decision in front of you. Visibility is.
2. Hermès
The clearest demonstration of that principle is the brand that has most conspicuously refused to act on it.
Hermès does not discount publicly and does not participate in the markdown calendar. Access to a quota bag is earned through purchase history across scarves, jewellery, homeware and ready-to-wear before a Birkin or Kelly is offered at all. Production is deliberately constrained; each bag is made by a single artisan. Bernstein's Luca Solca estimates Birkin and Kelly together contribute 25–30% of total group sales.
The consequence is a secondary market where the product routinely trades above retail. Rebag's 2025 Clair Report put Hermès average value retention at 138% — the highest of any brand tracked and a 38-point increase on 2024 — with eight models selling above retail across the year, led by the Mini Kelly II at 282% of retail and the Sellier Birkin at 183%. In July 2025 Jane Birkin's original prototype sold at Sotheby's Paris for €8.6 million, roughly seventeen times the previous handbag record of $513,040 set by a Himalaya Kelly 28 in 2021. The buyer was Valuence Japan — a reseller, not a private collector, which is itself informative about where conviction in the category sits.
Treat the widely circulated Baghunter figure of 14.2% average annual appreciation with caution: it was published by a handbag marketplace, covers 1980–2015, and is not independently constructed. The Rebag retention data and observed auction results are the sounder evidence, and they support the same direction of travel without the promotional framing.
What Hermès has built is an asset class it does not own.
On the Q2 2025 earnings call, Axel Dumas described the secondhand market as a real cause for concern, arguing that buyers acquiring bags to flip them dilute the client base the brand exists to serve, and stated plainly that Hermès is not interested in entering resale.
That position is defensible and it is also a structural admission. Once a bag leaves the boutique, Hermès has no view of who holds it, what it sold for, how many times it has changed hands, or whether the item trading under its name is authentic. It cannot see the market that sets the residual value underwriting its retail proposition. Richemont bought Watchfinder in 2018 precisely to avoid that position in watches. Hermès has chosen the opposite, and the choice is only sustainable while scarcity does the work.
The lesson for brands several tiers below Hermès is not that they should launch resale. It is that a secondary market operates on your product with or without your consent, and that the only lever available to a brand that declines to participate is supply restriction — which most businesses cannot exercise.
3. Verification
Two reference points are usually presented as strategic options. They are more accurately the endpoints of a cost function.
Vinted moved €10.8 billion GMV in 2025 at an €8 billion valuation, verifying almost nothing, holding no stock and signing no brand agreements. It invested in logistics and payments instead — over 500,000 drop-off points and its own wallet — and has 17 million UK users, behind only Primark and Next for reach.
Vestiaire Collective moved just under €1 billion GMV at above 50% gross margin and expects its first annual profit in 2026, fifteen years after founding. Of approximately 600 employees, more than 100 authenticate.
Authentication produces measurably higher realised prices on high-value pieces, so the expenditure returns. But it is a labour cost that does not amortise across volume, and Vestiaire's fifteen-year path to profitability is the empirical consequence.
Position on that spectrum is determined by unit value and counterfeit exposure, neither of which you control. High on both and light verification will place counterfeits into circulation under your name within a season. Low on both and heavy verification consumes the margin the programme exists to produce. Most programmes are scoped without pricing the verification tier the category requires, which is why the cost surfaces in year two rather than in the business case.
4. Identity infrastructure
The industry's response to verification cost has been to attach identity to the object itself, and two competing architectures now dominate. They are not equivalent, and the difference is governance rather than technology.
Aura Blockchain Consortium was founded in 2021 by LVMH, Prada Group, OTB and Cartier (Richemont) and constituted as a Swiss non-profit. It has issued digital product passports for more than 40 million items and offers brands a choice of private or public chain. Implementations include Tod's, which embedded NFC into the Di Bag; MCM, using an NFC chip set above the front plaque on its luggage, supplied by Temera; Loro Piana, Jil Sander, Marni and Cartier; and Louis Vuitton's LV Diamond certificates. Aura issues certificates of ownership alongside certificates of authenticity, which is the component built specifically for resale: a documented transfer converts a private sale into a visible event.
Arianee is Paris-based, founded 2018, and takes the opposite approach — an open-source protocol governed by a member association rather than a consortium of owners. It reports approximately 3.4 million passports in production across around fifty brands and 40+ markets, including Breitling, Moncler, Lacoste, Fnac Darty and the Richemont watch maisons (Panerai, IWC, Vacheron Constantin, Jaeger-LeCoultre), alongside Audemars Piguet. It runs a dedicated Polygon CDK Layer-2 chain, holds SOC 2 Type II and is a GS1 partner. Its commercial argument is portability: because the protocol is open, a brand can change operators without losing product history.
Arianee publishes a threefold uplift in secondary market value under continuous authentication and a 38% re-engagement improvement at Breitling. Both are vendor-reported and should be treated as directional rather than benchmark.
The decision axis is straightforward once stated. Aura offers control and peer governance, and suits houses that regard the identity layer as a competitive asset to be held collectively with equals. Arianee offers portability and independence, and suits brands unwilling to have their product data governed by a body their competitors sit on. The material risk in either case is the same: an identity layer that cannot be exported is a dependency, not an asset.
Note also what neither solves. Both authenticate an item at the point someone scans it. Neither tells you which of your returned units is sitting in a tote in your warehouse this morning, unscanned, losing value by the day.
5. Returns supply
A resale programme requires graded, priced, photographed inventory. There are three sources: customer buy-back, consignment, or your own returns and unsold stock. The first two carry acquisition cost or margin share. Returns do not — the unit is already owned and already paid for — which makes them the only source that clears an internal hurdle rate at scale.
They also arrive unusable, for structural rather than physical reasons.
Bracketing is now standard: estimates range from 46% to 63% of UK consumers, with over half of Gen Z and millennials doing it routinely. UK non-food online returns run at approximately 19.5%, clothing at 23–24% (Eightx UK benchmark, 2026). Processing costs £10–£25 per return before item value, reaching 20–40% of item price once lost margin is included. UK reverse logistics is estimated at £27 billion annually. Roughly 15% of returns are fraudulent.
The operative variable is elapsed time. A returned unit enters reverse logistics without identity and is processed on a queue basis rather than a value basis. By the time it is available again the season has moved and the size curve has been rebalanced. IMRG describes the outcome as stock locked in limbo: unavailable to customers, absent from demand forecasting, physically owned by you.
Illustrative, not a benchmark. Assumptions stated so they can be contested.
Take 10,000 returned units at £100 average selling price, processing at £15 per unit throughout. Secondary recovery at 45% of original retail — above the 25–40% range PwC observed across resale channels in 2024, uplifted for condition-graded, channel-matched product.
| Metric | Current state | With unit-level identity |
|---|---|---|
| Units back to full-price sale in window | 6,000 (60%) | 8,000 (80%) |
| Recovered at full price | £600,000 | £800,000 |
| Residual units | 4,000 → markdown at 40% | 2,000 → resale at 45% |
| Recovered on residual units | £160,000 | £90,000 |
| Processing cost | (£150,000) | (£150,000) |
| Net recovery | £610,000 | £740,000 |
Units back to full-price sale in window
Current state
6,000 (60%)
With unit-level identity
8,000 (80%)
Recovered at full price
Current state
£600,000
With unit-level identity
£800,000
Residual units
Current state
4,000 → markdown at 40%
With unit-level identity
2,000 → resale at 45%
Recovered on residual units
Current state
£160,000
With unit-level identity
£90,000
Processing cost
Current state
(£150,000)
With unit-level identity
(£150,000)
Net recovery
Current state
£610,000
With unit-level identity
£740,000
The delta is £130,000 per 10,000 returned units, or £13 per unit.
Its composition is the point. The resale line contributes £90,000; the movement of 2,000 units from markdown back into full-price sale contributes £200,000.
The dominant return on resale infrastructure is not resale revenue. It is the full-price sales that stop being lost.
The sensitive assumption is the 60% → 80% in-window recovery improvement. At a ten-point improvement the delta halves; below roughly five points the case rests on resale margin alone and becomes marginal for most mid-market economics. Your returns dwell-time distribution determines which case applies, and it is the first dataset we request.
6. Regulation
Under Article 25 of the Ecodesign for Sustainable Products Regulation, large companies — over 250 employees and above €50 million turnover — may no longer destroy unsold apparel, clothing accessories or footwear placed on the EU market. Medium-sized companies follow in 2030. Reuse, donation, repair, refurbishment and remanufacture remain permitted; incineration for energy recovery does not.
Derogations exist for safety risk, damage and IP infringement. Relying on one requires documentation retained five years, produced electronically within 30 days on request, with the exemption declared to the waste operator. From February 2027, discarded volumes must be published in standardised format.
Two consequences follow.
First, the counterfactual has changed. Doing nothing about unsaleable stock was previously costless and invisible. It is now a disclosed figure readable against competitors'.
Second, a pricing effect — and here the evidence is inferential rather than measured. An estimated 4–9% of textiles placed on the EU market were being destroyed before use. That volume must now be routed somewhere. Chanel's answer is instructive: days before the ban it confirmed it had ceased shredding and now directs unsold product to L'Atelier des Matières, its recycling business, held within Nevold, a materials platform launched in 2025 with €50–80 million behind it. Not resale, not outlet, not donation — feedstock. Compliant, and scarcity fully preserved.
That option requires capital most businesses lack. For the rest, the compliant routes are resale, off-price and donation, and donation absorption is finite. The reasonable expectation is downward pressure on mid-market secondhand values over two to three years as displaced volume competes with genuine pre-owned at equivalent price points. Revenue models holding current secondary prices flat are likely optimistic.
One provision warrants specific attention: counterfeits may still be destroyed without justification. Authenticity is now the legal determination separating a second life from disposal — which converts the verification question in section 3 from a commercial choice into a regulatory one.
7. Sequencing
The standard implementation path is platform first: select a resale partner, integrate a storefront, then determine where supply comes from.
The order is wrong. Outsourced platforms handle merchandising, pricing and customer experience competently. None of them solve identification, because identification occurs inside your reverse logistics before any platform sees the item. A programme built platform-first inherits whatever supply the returns operation happens to release, which is the late, unidentified, low-value residue.
Identity first, platform second. Returns triage funds the programme; the storefront realises what remains.
8. Limits
It depends on the product retaining secondary value, and many categories do not.
KID ASA, operating Kid Interior in Norway and Hemtex across the Nordics, joined Tekstilpro AS — a shared, competition-neutral textile return scheme aligned to EU producer responsibility rules — rather than building a proprietary programme. That is correct for home textiles: unit values are too low to carry handling cost, consumer resistance to used bedding is well established, and there is no marketplace demand to serve. The requirement is materials recovery, which is economic only when collection and sorting costs are shared across an industry.
Where your product sits there, we will say so before budget is committed. A resale programme built on product with no secondary demand is an expensive piece of greenwishing.
How we work
The diagnostic covers three things: your returns dwell-time distribution and what it implies for in-window recovery; your realised secondary prices and counterfeit exposure, which set the verification tier you must fund; and your Article 25 position ahead of the February 2027 disclosure.
The output is a build, partner or decline recommendation with the arithmetic attached, defensible in a capital allocation meeting rather than a sustainability update.
Where the answer is build, we design the operating model, the grading standard, and the item-level identity layer that makes returns triage possible — including how it interoperates with Aura, Arianee or a proprietary passport without creating a dependency you cannot exit.
We have run this from the operator's side: buying offices, e-commerce P&Ls, and the returns operations behind both.
A structured review of returns, verification exposure and Article 25 position, with a build, partner or decline recommendation and the arithmetic behind it.