All work
Placeholder marketplace
Fixing supply liquidity
Cut time-to-first-match from nine days to under two.
Context
Placeholder case study. A two-sided marketplace with healthy demand and a thin, slow-moving supply side.
The problem
New buyers waited over a week for a first match, and most churned before it arrived. Supply was there but poorly matched and poorly surfaced.
What I did
- Reframed the north-star metric around time-to-first-match rather than listings created.
- Introduced a lightweight matching heuristic ahead of any machine learning work.
- Rebuilt supplier onboarding to capture the two fields matching actually depended on.
Outcome
Median time-to-first-match fell from nine days to under two, and 30-day buyer retention improved materially.
What I'd do differently
I would have killed the personalisation workstream sooner — it absorbed attention that the basics needed.