A failed delivery is not one cost, it is five: the wasted drop, the return leg, the re-plan, the customer contact and the delayed invoice. This is the operational playbook we see work across European multi-drop fleets.
8 min read · Last reviewed 20 August 2026
| Cause | Typical share | Fix |
|---|---|---|
| Recipient absent | High | Pre-arrival notification with a realistic window |
| Wrong or incomplete address | Medium | Validate and geocode at intake, not at the kerb |
| No access: gate code, dock slot, lift, permit | Medium | Persistent access notes on the stop record |
| Vehicle arrives outside the site's receiving hours | Medium | Time windows enforced in the planner |
| Goods refused or damaged | Low | Photo condition check at load, not at delivery |
Note what is missing from this list: driver performance. Fleets that respond to failed deliveries with driver pressure tend to see the number stay flat and turnover rise. The leverage is upstream, in the data attached to the stop.
An ETA is a promise. It becomes reliable when the planning engine models service time per stop type, mandatory breaks, and traffic patterns for the actual time of day — and when it is recalculated as the route runs, not fixed at 06:00.
When something goes wrong, the driver's options should be visible in the app: report a delay, report a failed attempt with a reason code and photo, or request a re-plan. Each action should reach the dispatcher as a structured item in an exception inbox with an SLA clock, not as a message in a group chat.
Track these for four weeks before changing anything. Most operations discover the problem is concentrated in a handful of customers and one part of the day, which makes the fix far cheaper than a fleet-wide programme.
Depot2Door combines constraint-aware planning, stop-level access data, structured exception handling and photo proof of delivery in one system your drivers actually use.
It depends heavily on stop type. Scheduled B2B dock deliveries should be well above 95%, while unattended residential deliveries of bulky goods run considerably lower. Compare yourself against your own trend per stop type rather than an industry headline number.
Only partly. Optimisation reduces distance and improves punctuality, but it does not know that the gate code changed. Failure rates drop when planning, stop-level data and driver feedback share one record.
Add the vehicle and driver time for the wasted drop, the re-delivery, the administrative handling and the delayed invoice. Most European operators who measure it land on a figure well above the profit margin of the original delivery, which is why prevention pays quickly.
Only against an explicit, recorded instruction from the recipient, captured in the app with proof of delivery. Verbal permission without a record shifts liability onto the carrier.
What an eCMR is, the legal basis under the CMR Additional Protocol and the eFTI Regulation, which countries accept it, and how to move a fleet from paper consignment notes to digital proof.
A practical 2026 guide to EU Regulation 561/2006: daily and weekly driving limits, the 45-minute break, split breaks, reduced rests, compensation rules and how fleets prove compliance.
