There are two ways to build a resale marketplace and they are much less similar than they look.
The first is the open version. Anyone can list anything, the platform mediates, and scale comes from breadth. Poshmark, Vinted, Depop, eBay. The second is the brand version, where a company that already makes and sells a product operates the secondary market for its own goods. Patagonia does this. Levi’s does this. Increasingly, so does anyone selling something durable enough to have a second owner.
The open version is a liquidity problem. The brand version is a trust and operations problem, and the brand starts with advantages that make it a different build.
What The Brand Already Has, And What It Costs Them
A brand launching resale has supply it can identify, demand it already owns, and authentication that is close to free. It knows what it made, when, in what colourways, with what construction. Verifying that an item is one of its own is a problem it has already solved by having made the thing.
Compare that to an open marketplace trying to authenticate a luxury handbag from photographs. Entire businesses exist to do only that.
What the brand does not have is a resolved position on the questions resale forces. Three of them, and none is a technology question.
Does resale cannibalise new sales.The honest answer is that it does, some, and it also does other things: it puts a floor under the product’s residual value, which supports the new price; it brings in a buyer at a lower price point who may buy new later; and it keeps the product out of a competitor’s secondary market. Whether the net is positive is a real question with a real answer, and it depends on the category and the price point. It is worth modelling before the build rather than arguing about afterwards.
Who owns the customer relationship on a resale transaction. If a customer buys a used jacket from another customer through the brand’s platform, whose customer are they. This matters because it determines whether resale data flows into the same CRM, whether resale buyers get marketing, and whether the resale programme is measured as a channel or as a sustainability line item. Programmes measured as sustainability line items tend to get defunded.
What happens to the item physically. This is the fork that determines the entire build, and it is the next section.
Peer-To-Peer, Consignment, Or Buy-Back
Three models. They produce three different companies.
Peer-to-peer. The brand hosts the market, seller ships direct to buyer, brand takes a commission. Lightest operationally, fastest to launch, and control over condition and quality extends only as far as the listing flow enforces it. This model is a marketplace.
Consignment. Seller ships the item to the brand, the brand inspects, photographs, grades and lists it, and the buyer buys from the brand. Heaviest operationally, because you now have a warehouse, an inspection process and inventory. Highest control, best buyer experience, and the margin has to cover the handling. This model is a retail operation with sourced inventory.
Buy-back or trade-in credit. The brand purchases the item outright, usually in store credit rather than cash, then refurbishes and resells. Not a marketplace at all. It is a procurement and refurbishment business, and the credit is a retention mechanic. The economics are the most attractive of the three when the credit multiplier is set well, and the operational load is similar to consignment.
| Model | What it actually is | Operational load | Control over condition |
|---|---|---|---|
| Peer-to-peer | A marketplace | Lightest | Only what the listing flow enforces |
| Consignment | A retail operation with sourced inventory | Heaviest | Highest, item is in hand |
| Buy-back | A procurement and refurbishment business | Similar to consignment | Highest, you own the item |
These get conflated easily, because the peer-to-peer economics and the consignment buyer experience are each the obviously desirable one, and nothing in a wireframe shows you that they are incompatible. The gap between them tends to surface in the middle of a build. It is worth forcing the choice at the model stage even though it is an uncomfortable meeting.
Condition Grading Is The Product, Not A Dropdown
On a new-goods marketplace, the product description is largely fixed. On resale, every single unit is different, and the buyer is being asked to commit money to an item they cannot inspect. The entire transaction rests on whether the condition description is accurate.
Which makes condition grading the core product problem, and it is almost always implemented as the weakest part of the listing flow: a select box with four options, “New, Excellent, Good, Fair”, filled in by the person whose interest is in choosing the highest one.
The version that works is structural rather than declarative:
- The taxonomy is category-specific. “Excellent” means different things for a wool coat, a pair of trainers and a bicycle, and a single scale across all three guarantees disputes in at least two of them.
- The seller does not choose a grade. They answer specific questions about specific wear points, and the grade is derived. Pilling on the elbows, sole tread depth, chain stretch. Answering a question about a knee wear point is harder to shade than picking a word.
- Required photographs are prescribed rather than free-form. Not “add up to eight photos” but “photograph the left cuff, the inside collar label and the sole”. The prescribed shots are the ones that carry disputes.
- Publication is gated on completion. A listing that has not been through the grading flow does not go live.
That last one gets pushback, because gating publication reduces listing volume, and listing volume is the metric everyone is watching in month one. It also reduces the dispute rate, and the dispute rate is what determines whether a resale buyer ever comes back for a second purchase. Resale platforms carry a structurally higher dispute rate than new-goods marketplaces, and the second transaction is where the whole model either works or does not.
“On a resale marketplace the condition description is the product. Whoever writes it is setting the buyer’s expectation and your dispute rate at the same time.”Darren Cody, Marketplace Studio
Authentication, And Where It Belongs In The Flow
Authentication is not required in every category. It is required where the counterfeit market is active and the price point justifies the cost: luxury goods, sneakers, watches, high-end electronics. For a used garden tool it is a cost with no corresponding trust benefit.
Where it is required, the design question is not whether to authenticate but where the authentication sits relative to payment. If the item is authenticated before listing, you need the item in hand, which pushes you toward consignment. If it is authenticated after purchase, the transaction has to pause: money taken, item in transit to an inspection point, buyer waiting, and a defined outcome if it fails.
That pause is a state your payments architecture has to support natively. Funds held, not captured. A release trigger tied to an inspection result rather than a delivery confirmation. A refund path that does not require a human to intervene. Retrofitting that onto a payment flow built for immediate capture is one of the more expensive rebuilds in this category, which is why it is worth deciding on the authentication model before the payment integration is written rather than after.
What The Brand Version Gets To Skip
Worth ending on the advantage, because it is substantial.
A brand running resale on its own goods skips the cold-start problem almost entirely. Supply is identifiable: it is in the wardrobes and garages of a customer list you already hold, and you can email them. Demand is on your site already. Authentication is cheap because you made the item. Category taxonomy exists because it is your own catalogue.
What is left is the grading flow, the dispute process, the payment states and the commercial model. That is a real build, and it is a much smaller and much more predictable one than launching an open marketplace from zero.
We have built recommerce platforms where condition grading is enforced through the listing flow rather than displayed next to it, and treat the authentication decision as a payments architecture decision rather than an operations one. Our resale and recommerce page covers the category more broadly, and launching a marketplace inside an existing business covers the channel and margin questions a brand-side programme runs into. If you are working out which of the three models your programme actually is, that is worth an early conversation.
