Recommendation engine driving 40% revenue lift for a $3 billion e-commerce group
A multi-brand fashion retailer saw only 2.8% conversion from personalized recommendations.
StartDate replaced legacy rules-based engines with a transformer-based sequence model trained on 180 million sessions across web, app, and in-store POS. Real-time recommendations now factor style affinity, weather, local trends, and inventory velocity.
Conversion rate jumped to 10.9%; average order value rose 26%; incremental revenue attributed to AI personalization exceeded $420 million in the first year.