AI-powered personalisation that turned first-time buyers into loyal beauty enthusiasts

Tira Beauty's launch in April 2023 was Reliance Retail's most ambitious bet in the beauty category. The market was not short of competition — Nykaa had built a decade of brand equity and customer data, while Sephora's luxury positioning attracted premium shoppers. Tira's thesis was that neither incumbent had truly solved personalisation at scale: Nykaa's recommendations were category-level at best, and Sephora's in-store experience did not translate online.

The foundation of Tira's differentiation was a unified customer profile that persisted across every touchpoint. When a customer tried a Charlotte Tilbury foundation in Tira's Bandra store, that trial was logged against their profile. When they opened the app three days later, the recommendation engine already knew their shade, their skin type, and their price sensitivity — and could suggest complementary products with 94% accuracy. This level of personalisation required JCP's CDP to stitch together POS transactions, app sessions, web browsing, and loyalty activity in real time.

The AI recommendation engine was the second major pillar. With 2,000+ brands and hundreds of thousands of SKUs, the challenge was not having enough products — it was surfacing the right ones at the right moment. JCP's ML models were trained on purchase history, skin-type profiles, wishlist behaviour, and seasonal signals. Early A/B tests showed a 38% lift in add-to-cart rates when AI recommendations replaced manually curated carousels. By Q2 2024, the engine was driving 40% of all revenue across digital channels.

Inventory synchronisation was a less visible but equally critical capability. Beauty customers are notoriously sensitive to stockouts — a customer who cannot find their shade online after trying it in store is unlikely to return. JCP's real-time inventory mesh connected all 50+ Tira stores and fulfilment centres into a single stock pool, enabling ship-from-store fulfilment that turned physical stores into mini-warehouses. The result was a near-zero stockout rate for top-selling SKUs across all channels.

The loyalty programme, powered by JCP's loyalty module, was the final piece. Rather than a simple points system, Tira's loyalty engine tracked customer lifetime value, predicted churn risk 30 days in advance, and triggered personalised re-engagement campaigns — a birthday gift recommendation, a replenishment reminder, an early-access invite to a new brand launch. Within 12 months of launch, Tira's repeat purchase rate reached 40%, a benchmark that rivals established global beauty retailers and significantly exceeds the Indian beauty e-commerce average of 22%.

By 2025, Tira had confirmed a 2.5× average order value lift compared to its first-month baseline — the combined result of AI cross-sell, upsell recommendations, and a loyalty programme that rewarded customers for exploring new categories. In a market where most beauty retailers compete on catalogue breadth, Tira chose to compete on intelligence — and the numbers validated that bet.