Multi-brand fashion retailer (anonymized) — Doubling the head of the range without buying more stock | Particular Audience
Computer vision solved the cold-start problem on new arrivals, spreading demand across the catalog instead of concentrating it in a handful of SKUs.
Doubling the head of the range without buying more stock
Computer vision solved the cold-start problem on new arrivals, spreading demand across the catalog instead of concentrating it in a handful of SKUs.
- +21.1% More SKUs in head of range
- +14.7% Average order value
- +11.9% Revenue per visitor
As ranges grow, a static site means fewer items get discovered and revenue concentrates in a shrinking subset of products — the power law working against the buying team.
New arrivals are invisible to legacy recommendation engines because they have no behavioral history. Computer vision surfaced visually similar items to in-market shoppers from day one, paired with bought-together recommendations for bolt-on purchases.
The result was a wider revenue base: more SKUs earning their place in the top 80% of turnover, and higher revenue per visitor without additional traffic.