LoveNspire
↳ Click-only attribution, validated against LoveNspire's Shopify order data over the measurement window. Items-per-order measured against a held-out control group.
LoveNspire sells Indian ethnic décor and gifts to a US diaspora audience: a culturally specific catalog where product names mix Hindi and English (Hinglish) and meaning shifts with the festival calendar.
The challenge
Generic search couldn't tell one culturally-specific product from another, and shoppers searching in Hinglish hit dead ends. Merchandising the right festival items at the right time was a constant manual effort.
What we deployed
Our embeddings learned the catalog's real relationships (festival context, product pairings, and mixed-language titles) without a hand-built synonym list. Search and recommendations began surfacing culturally coherent sets on their own.
“PersonalizerAI completely unblocked our discovery problem during peak festival season, effortlessly surfacing relevant bundles that generic apps missed.”
Measured in revenue.
AOV lift
rolling 30-dayitems per order
vs. controlunderstood
no synonym list↳ Click-only attribution, validated against LoveNspire's Shopify order data over the measurement window. Items-per-order measured against a held-out control group.
