M&S wasn’t short on products; we were short on relevance. Customers were sifting through broad lists and missing the pieces that would complete their baskets. That’s why we built

Phoenix Recommendations: a real-time engine that uses intent signals like search, clicks, dwell, size, and stock, wrapped in brand guardrails, so every visit feels personally merchandised. It surfaces the next best item, assembles full outfits, offers similar styles when something’s out of stock, and brings in the right accessories at the right moment.