Launching Secondhand
H&M wanted to prove circular commerce could work at scale — not as a side project, but woven into the existing shopping journey. The launch went smoothly. Then a late-breaking logistics problem threatened to unravel it.
A 65-person validation study found that customers weren't angry about receiving two parcels — they were angry about finding out too late. That reframe changed the solution from a logistics fix to a communication design. Six-figure GMV in month one.
H&M wanted circular commerce — but had never shipped it
H&M Group's goal: a climate-positive value chain by 2040. Secondhand produces 70% less CO₂ than traditional retail. The gap was behavioural — consumers said it mattered, but few bought regularly.The brief: weave secondhand into H&M's existing flow, not build a separate destination. Fulfillment would come from Sellpy, a secondhand partner. That detail would matter later.
My role: solo designer, 12 teams — product, engineering, Design System, logistics, legal, data — no direct authority over any of them.
Three problems. One we didn't expect.
Items shown on mannequins, not models. Customers couldn't assess condition or fit.
Unclear whether H&M's standard policy applied to secondhand items.
Items shipped from Sellpy in a different parcel. Customers were angry — or so we thought.
The first two were solvable. The third needed more investigation.
They weren't angry about two parcels — they were angry about finding out too late
Customers who reacted badly had discovered the separate parcel at checkout — or not at all. Customers told early didn't mind.
The logistics weren't the problem. Late discovery was the problem.
We didn't need to fix the fulfillment model — we needed to fix when we communicated it.
Three hypotheses. One winner. Data that closed the debate.
Communicate the environmental benefit of separate transport
Communicate early that items would arrive from Sellpy
SelectedMake existing Sellpy disclosure more prominent
If late discovery was the cause, earlier disclosure was the fix. The Design System team blocked a new component, so we reused a notification banner. It turned out cleaner anyway. 65 participants per variant confirmed it:
- Negative parcel sentiment: 30% → 17%
- Correct parcel count awareness: 53% → 66%
- Positive sentiment toward communication: 71% → 86%
The data removed any ambiguity. We had evidence, not a judgment call.
Six-figure GMV in month one. A new business model proved.
Soft launched February 3, 2021 — Sweden, then Germany. Six-figure GMV in the first month. Proof that circular commerce could work inside a traditional retailer without a separate platform or brand.
What I'd do differently: I took on more validation risk than necessary early on. And the 12-team coordination would have been smoother with shared decision-making principles from the start — there was no internal precedent, and I felt that gap.
