Delivery complaints per 100 orders, by carrier: 6.7 versus 2.3
Same products, same customers, eight carrier-marketplace pairs. The gap between the best and the worst courier is bigger than the gap between your best and worst product, and it is measurable from your inbox.
A metric your inbox already contains
Carriers report their own performance: scanned on time, delivered on time, first-attempt success. None of those tell you what the customer experienced, and none of them are broken down the way you sell. Your inbox is. Every “where is my order” and “the driver never turned up” is a delivery complaint, and every one of them sits on an order that names the carrier.
Divide one by the other and you have complaints per 100 orders, per carrier, per marketplace, from the customer’s point of view. It costs nothing and it updates every hour.
Eight carrier-marketplace pairs
Two things stand out. The worst pair is a two-man booked-slot courier carrying the biggest items on the marketplace with the strictest delivery promise. And the same carrier, FedEx, scores 3.8, 3.6 and 2.3 on three marketplaces: the carrier is the same, the products and promised dates are not.
Why the same carrier scores differently
- The promised date. Each operator shows its own delivery estimate. The marketplace with the tightest promise gets the most “it is late” messages for the same actual transit time.
- The product mix. A 180 cm TV unit and a floating shelf can share a carrier and behave nothing alike.
- The customer’s route to you. On one operator, order-status questions go to the operator’s support team first; on another, straight to the seller. The reason matrix shows this.
What a delivery complaint costs
Delivery threads are the cheapest to shorten because the answer is almost always in the tracking. An inbox that shows the carrier, the tracking number and the last scan next to the message lets the first reply be the last one. When the carrier has genuinely lost the parcel, that same view tells you on day one, before the customer asks for a refund. Which is the expensive outcome.
What to do with it
- Publish this table monthly, to yourself. Twenty orders per pair is the minimum for the number to mean anything.
- Take the worst pair to the carrier with the order numbers. “6.7 per 100 on 45 orders, here they are” is a conversation; “customers complain” is not.
- Check the promised date on the marketplace with the highest rate before changing carrier. Moving the estimate by a day is free.
- Put tracking beside the thread so the first reply carries it.
- Delivery complaints per 100 orders, per carrier and marketplace, is a customer-eye view of courier performance you already have.
- The same carrier scored 3.8, 3.6 and 2.3 on three marketplaces: promise and product matter as much as the courier.
- Two-man booked-slot deliveries on big items are the high end; fix the slot communication first.
- Tracking beside the thread turns a 5-message thread into a 1-message one.
Questions people ask
How do I measure carrier performance from customer messages?
Count the conversations whose reason is delivery delay or item not received, match each to the carrier recorded on the order, and divide by that carrier's orders on the same marketplace in the same period. Leave out carriers with fewer than about 20 orders, or one bad week dominates. The result is complaints per 100 orders, comparable across carriers and months.
What is a normal delivery complaint rate for furniture?
In 90 days of data from one UK furniture seller, carrier-marketplace pairs ranged from 2.3 to 6.7 delivery complaints per 100 orders, with parcel carriers on small items at the low end and two-man booked-slot deliveries at the high end. Above 5 per 100 on a single carrier is worth a conversation with the carrier.