Resale
Operations Briefing

The leaders have stopped optimising. They've started building things that can't be copied.

Vinted closed 2025 with €10.8bn GMV - up 47% - and net profit of €62m. Down from €77m the year before. Every headline you'll read will frame that as a stumble. It isn't. Vinted made a deliberate choice to trade margin for infrastructure: Germany turnaround, Vinted Go expanding into Spain and Portugal, Vinted Pay wallet launched, 500,000+ pick-up and drop-off points across Europe. CEO Thomas Plantenga's explanation is worth reflecting on: the lower the cost, the more attractive it becomes to trade lower-value goods. That's not efficiency for its own sake. That's a deliberate expansion of the total addressable market - built on rails that took years to lay and will take years to replicate.

ThredUp posted its first ever positive annual cash flow in 2025. Revenue up 20%. Active buyers up 30%. Still loss-making on GAAP - but the losses are investment losses now, not structural ones. The unit economics question has been answered. The 2026 focus is supply acquisition and AI-driven discovery. The RealReal crossed $2bn GMV, posted positive Adjusted EBITDA every quarter for the first time, and is guiding 12-15% GMV growth for 2026. Average order value hit $641 in Q4. eBay paid $1.2bn for Depop. These aren't isolated data points. They're the same thesis - discipline first, then scale, then reinvestment - arriving simultaneously across structurally different models.

We spent months interviewing operators for the Resale Operations Report. The pattern that kept surfacing was this: the operators achieving real numbers weren't the most innovative. They were the most sequential. Unit economics first. Then operational leverage. Then technology on top of a clean foundation. Vinted didn't build Vinted Go because growth gave them spare cash. They built it because years of cost discipline gave them the conviction that logistics infrastructure was the right next bet. ThredUp's discovery investment works because their supply data is clean enough to train on. The RealReal's AOV holds because their authentication reputation was built before they needed it to hold margins.

The sequence isn't a strategic preference. It's a prerequisite. And it doesn't run in reverse.

ThredUp's 2026 Resale Report puts the global market at $393bn by 2030 and identifies supply as the binding constraint on getting there. But those who've run the numbers know the constraint is more specific than that. There's no shortage of secondhand clothing. There's a shortage of secondhand clothing worth processing at a margin that works. 75% of operators are still managing listings manually. The leaders building infrastructure right now aren't solving a volume problem. They're solving a quality and economics problem - and doing it at a scale that makes the distinction invisible to everyone watching the GMV numbers go up.