£52,000 recovered every month.
26% of abandoned carts, back.

By the time an email landed, the customer had already gone.

The e-catering platform sold time-sensitive, often last-minute catering to corporate and private clients who expect quick, reliable service. Every abandoned basket was lost revenue, and the longer it sat unaddressed, the less likely it was to come back.

Email was the default recovery channel, and it was too slow. Messages could go unseen for hours or days, by which point intent had faded and the order had likely gone elsewhere. The business needed to catch abandonment the instant it happened, work out who was worth re-engaging, and do it without feeling intrusive.

Three constraints were non-negotiable:

• Abandonment had to be caught the instant checkout was interrupted
• Email recovery arrived hours or days too late
• Re-engagement had to feel helpful, not intrusive

Real-time detection, then a text that remembers what was in the basket.

An AI monitoring layer was integrated directly with the e-catering platform’s checkout process, detecting in real time the moment a customer abandoned a basket without completing payment.

Instead of email, the system used SMS: text messages achieve 98% open rates within three minutes and get checked far more often than email on mobile. The AI analysed each abandoned basket’s contents, value and the customer’s order history, then generated a personalised message referencing the specific items selected. Repeat corporate clients received a concise confirmation with a one-click return link; first-time customers got a friendlier reminder highlighting menu items and delivery windows.

Automated rules throttled messages and excluded anyone who had already completed an order or opted out. A/B testing refined timing, message framing — item reminder versus value-focused versus incentive — and call-to-action wording, with the AI learning from outcomes to keep improving recovery rates.

A quarter of the lost baskets, back.

Real-time detection plus a personalised text turned a leaky checkout into a recovery channel: 26% of abandoned baskets came back, worth £52,000 a month that had previously been lost. Customers responded well to the timely reminder rather than resenting it, and ongoing A/B testing kept improving recovery without raising complaint rates — a scalable model for cutting abandonment across e-commerce catering.