Resale Ran Out of Inventory Before It Ran Out of Buyers

Live resale's binding constraint has flipped from demand to supply, and the wholesalers now filling the gap are being formalised by intermediaries that sell brands discretion about where surplus goes. Customer returns are a named input to that supply, and livestream selling is the resale channel with the least verification in it.

Live resale stopped being limited by how many people want to buy. Whatnot’s sellers moved more than $6 billion of goods by October 2025 against $3 billion in all of 2024, a doubling the channel’s own suppliers now quote back at it in their marketing. Growth on that slope converts into a restocking problem, because a format that clears stock in seconds has to be refilled at industrial speed.

On Black Friday, buyers were taking 125 items a second at peak. Closets do not refill at that rate, and the goods moving through livestream resale were never mostly secondhand anyway. What has changed is who supplies them, and what that supplier sells to the brand at the other end.

The intermediaries formalising that supply arrived five weeks apart, and they are pitching brands rather than resellers. The Reseller Source surfaced in December 2025, placing brand surplus with vetted sellers on Whatnot; in January, Overstock Trader absorbed it as a private placement channel inside its own buyer network. The proposition to brands is control: private, vetted placement that avoids “public discounting and uncontrolled resale exposure.” Overstock Trader’s Gregg Schwartz states the premise without embarrassment, saying the question “is no longer whether to participate, but how to do it in a way that is intentional, well-controlled, and aligned with overall brand strategy.” He is right about the question. The inventory on offer is listed as mixed-SKU assortments, aged styles, fragmented quantities, off-cycle stock and customer returns.

The last of those, customer returns, is where counterfeits are already known to enter. American shoppers are expected to send back $849.9 billion of merchandise in 2025 on the National Retail Federation’s count with Happy Returns, and a returns desk is a receiving function under time pressure, staffed to move volume rather than to adjudicate authenticity. Of the retailers in that study who track return fraud, 64 percent reported a rise in decoy returns, the category covering a fake sent back in place of the real item. Goods that clear the desk are not destroyed. They are graded, palletised and sold onward as returns, which is precisely the label under which they are now being placed with livestream sellers.

A counterfeit that clears a returns desk has passed the only inspection it will ever face.

Live selling carries the least verification of any resale channel, and the gap is structural rather than a failure of will. The RealReal filled 937,000 orders in the second quarter of 2026, and its model runs goods through an expert-led authentication process before they reach a buyer. That process is slow and expensive, which is the reason it catches things. Whatnot never takes possession: sellers print a platform-generated label and ship direct, so no intake step exists at which an inspection could happen. The absence is not an oversight. It is what makes the format fast enough to work.

![Vale weighing one authenticated handbag against a hundred sealed boxes](({{generate: A nautilus in a worn naval jacket stands at a warehouse inspection bench. On the bench, a single handbag under a jeweller’s loupe with a small price card beside it hand-lettered $7. To the right, a hundred identical sealed cartons stacked and receding into the dark, none of them opened. Nautilus at left third, clinical and unhurried.}}))

Authentication does exist on Whatnot, priced as a promotion rather than imposed as a condition. Luxury sellers can pre-authenticate through Entrupy for $7 an item under a partner discount, and nothing obliges a seller to opt in. Seven dollars is not the barrier; optionality is, because the seller holding the least verifiable stock has the least reason to pay it. The fallback runs the other way, through the buyer, who has thirty days to claim an item is fake. That places detection with the person least equipped to perform it, someone who bought from a moving video feed in the seconds before a timer ran out.

The strongest objection is that institutionalisation improves on what it replaced. A vetted placement network is more accountable than an anonymous pallet broker, and a brand that previously lost sight of its surplus at the loading dock now gets told where it lands. For that objection to hold, the vetting would have to travel with the goods rather than stopping at the seller. Read what is being vetted — the reseller, not the item. Overstock Trader screens for brand-protective conduct: private placement, no public discounting, controlled channel exposure. Those are price-architecture controls, and they say nothing about what is in the box.

The exposure lands on the brand that supplied the pallet, because buyers attribute provenance backwards. Someone who receives a counterfeit from a livestream did not buy from a wholesaler she has never heard of; she bought a name she recognised, at speed, from a seller the platform let onto the feed. Brands entering these placements are buying discretion about where their surplus travels and declining to buy proof of what travels with it. The correction is neither complicated nor costly: make authentication a condition of the placement and price it into the pallet, instead of leaving it a line item the seller can skip. If the channel keeps doubling while verification stays optional, brands will be answering for goods they never made, in a channel they chose to feed. Seven dollars buys that answer at the pallet; the alternative is hearing the question from a customer who already has the box open.