Harpy Glossary

Pre-fulfilment Cancellation Rate

Amazon & D2C glossary · Harpy Media

Pre-Fulfilment Cancellation Rate measures how often a seller cancels orders before shipping them — a performance metric for merchant-fulfilled orders, evaluated on a rolling basis against a hard ceiling of 2.5%.

What is Pre-fulfilment Cancellation Rate?

A cancellation before dispatch is almost never a customer problem. It is the visible end of an inventory or process failure: stock that was promised but not there, quantities that were wrong in the system, or orders taken that could not be served. The customer is told no after being told yes, and the platform records it against the seller.

Why this metric has no tolerance

Breaching the threshold can suspend seller-fulfilled privileges outright — which strands inventory, stops revenue from that channel, and requires an appeal to restore. That is the escalation, but the everyday damage is quieter: cancellations erode the customer’s trust, generate negative feedback and claims, and signal to the platform that the seller cannot be relied on to serve demand.

Because cancellations usually indicate an inventory failure rather than a fulfilment one, the diagnostic path runs backwards from the cancel: what did the system think was in stock, what was actually in stock, and what was the mechanism by which those two numbers diverged? The answer is nearly always process — updates lagging, reserves missing, or manual spreadsheets.

Building the accuracy that prevents cancellations

The core fix is real-time inventory truth. Stock reserved the moment an order is placed across every channel; safety buffer held back deliberately so a spike cannot oversell a listing; and quantities updated by system rather than by end-of-day spreadsheet entry. Where multiple channels sell the same physical stock, the reservation has to be shared, not sequential.

Then the checks that catch the rest: confirm stock levels before running promotions, keep unlisted buffer units during high-traffic events, and audit the discrepancy between physical counts and system counts regularly enough that drift is corrected while it is small. Cancellation rate is best understood as a report on how well you know your own inventory.

Cancellation Rate (%) = (Seller-Cancelled Orders ÷ Total Merchant-Fulfilled Orders) × 100Evaluated over a rolling window, and only for orders the seller cancels. Customer-initiated cancellations within the standard grace period are excluded.

In practice

A seller running spatula sets through merchant fulfilment reserves stock the instant an order lands on any channel, and deliberately holds a small unlisted buffer. Across a promotional weekend they take 200 orders, serve every one from accurate available inventory, and finish with a cancellation rate of zero — retaining Buy Box eligibility on listings that a single oversell could have weakened.

⚠️ Watch out. Updating quantities by hand, once a day. During a busy event a merchant takes fifty orders against a warehouse holding forty units, because the listing still showed stock that had sold earlier. The oversold orders have to be cancelled, the metric takes the hit, feedback follows, and the root cause — a manual process with a day-long latency — was the same one that had caused the last three incidents.

💡 Harpy tip. Make availability real-time and reservations shared: system-updated quantities, stock reserved at the point of sale across all channels, and a deliberate buffer kept out of the listing during peaks. Audit physical against system counts regularly — cancellations are almost always the visible symptom of invisible drift.

How Harpy Media helps

Inventory accuracy is part of our operations work: real-time reservation across channels, buffer stock held deliberately, and the counting discipline that keeps system quantities matching what is actually on the shelf.