An 8,500-SKU home retailer improves weak discovery
Weak search responses made best-selling pieces harder to find. Continuous re-ranking improved discovery and lifted collection conversion.
modeled revenue recovered per quarter
| Metric | Before | After |
|---|---|---|
| Collection click-through rate | 3.1% | 4.1% |
| Zero-result search rate | 11% | 4% |
| Products rewritten citation-ready | 0 | 6,200 |
| Revenue per 1,000 sessions | $420 | $511 |
Large catalogs and static, hand-curated collections produced weak search and click-through signals for high-margin best-sellers. Merchandising reviews happened weekly, far slower than demand actually shifted.
Search
Regor repaired zero-result and low-CTR searches across a wide long tail of room, material, and style queries, connecting shoppers to inventory they could not previously find.
Ranking
It continuously rebalanced collections using trusted search and click-through signals, promoting products with rising engagement and demoting stale SKUs without waiting for the weekly manual pass. High-impact ranking changes were held for review.
Product data
More than 6,000 products were rewritten into structured, citation-ready copy with materials, dimensions, and use-case attributes filled in, improving both on-site search and visibility with AI assistants.
The modeled result was a 31% lift in collection click-through and an estimated $128,900 in recovered revenue per quarter, with every change logged and supported mutations carrying rollback information.
"Our best products were always there. Shoppers just struggled to find them. Having something rebalance the catalog every hour instead of every week is the difference."
Modeled scenario. The figures above are illustrative projections from Regor's recovery model and published discovery benchmarks, not results from a named customer. Search shoppers convert roughly 2 to 3 times higher than browsers (Forrester), and one day of an Adobe Analytics feed from a major US retailer, representing 15,968 unique visitors, showed that 73.5% of purchasers searched before buying. Actual results vary by catalog, traffic, and configuration.
Discuss your product discovery workflow
A 20-minute conversation about evidence, workflow, and fit.
Book a demo