The hidden cost of cancelled orders

Cancelled orders were a visible, growing problem in operations — but no one had translated them into money. The team could see that orders were being cancelled. They could not see what those cancellations cost, whether customers came back afterwards, or whether the problem was getting worse. Without a number, cancellations stayed an operations metric and never reached the commercial agenda.

The client is a high-volume on-demand delivery platform handling well over half a million orders a month. Cancellation volumes were climbing and starting to strain operations, but the downstream commercial impact was invisible.

The core issue was framing: a cancellation looks like an operational event, so it was managed as one. The question of what happened to the customer afterwards — and what the lost orders were worth across a year — had never been answered.

We took a full year of order data and focused on the moment after each cancellation, not the cancellation itself. Every cancelled order was classified by whether the same customer placed a successful order again within a short window. Cancellations with no follow-up we defined as genuinely lost.

We then applied the platform's average order value to those lost orders to size the revenue at risk, and tracked the re-order rate month by month — the test for whether this was a temporary spike or a permanent leak.

Monthly cancellations climbed from roughly 57,000 to 82,000 across the year. Two in five cancelled orders were never recovered — the customer simply did not order again. Only around 56% of cancellations were rescued by a re-order within two hours, and that recovery rate stayed flat all year. A stable recovery rate against a rising cancellation volume meant this was structural loss, not a seasonal dip.

For the first time, the business had a defensible, month-by-month number for what cancellations were costing it. That reframed cancellations from a back-office stat into a board-level priority, and gave the team a clear basis for deciding where to intervene first — the segments and periods where the lost value concentrated.

The work didn't recover the revenue on its own. It did something more useful as a starting point: it sized the problem honestly, so the business could decide what it was worth fixing.

Revenue Recovery is one of five Hira Deep-Dive Sprints — a focused, four-week engagement that finds where revenue is quietly leaking and sizes it in terms the business can act on. It's the right fit when you can feel a problem in operations but have never seen it costed.