Cutting failed deliveries when addresses are unreliable

Software written elsewhere assumes an address is a street number on a named street, and that a map can turn it into a point. In much of the Gulf that assumption does not hold. Addresses arrive as landmarks, as villa numbers on unnamed roads, as a building known locally by the shop on its ground floor. The parcel is not failing because the driver is careless. It is failing because the address was never a coordinate.

Use the systems the region actually built

Dubai solved this with Makani, a ten digit code that points at a specific building entrance. Saudi Arabia built the National Address, operated by SPL, with a short code that resolves to a precise location. Both are more reliable than any text address a customer types, and both are ignored by software that only has a field called street.

Treat them as first class address data rather than notes. A code that survives from the order into the driver app is worth more than three lines of description, and from 1 January 2026 the Saudi National Address is required on parcel deliveries in Saudi Arabia, so this stops being optional there.

Confirm before you dispatch, not after you fail

The cheapest failed delivery is the one that never leaves the hub. A message the day before, on the channel the customer actually reads, converts a doubtful address into a confirmed one while the parcel is still with you. In this region that channel is WhatsApp, and the difference in read rate against SMS is the whole reason to bother.

Give the customer something to do in that message, not just something to read. A link that lets them correct the address, drop a pin or pick a different day removes the failure before it costs a trip.

Make the driver notes outlive the delivery

The most valuable address data in your business is already in your drivers heads. The gate is on the far side. Call from the pharmacy, not the main road. Ask for the security desk. Most operations lose that knowledge the moment the driver closes the stop.

Attach the note to the customer rather than to the shipment, and the second delivery to that address is easier than the first. Over a year this compounds into a quieter operation, and it costs nothing but a field that persists.

Decide reattempts by reason, not by rule of thumb

A customer who was not home and a customer who refused the parcel are different events, and giving both three automatic attempts wastes trips on one and gives up too early on the other. Group the reasons your carriers and drivers actually report into a small set you can act on, then set the retry rule per reason.

Then work the queue by what it costs to ignore. A failed delivery holding a large cash on delivery amount deserves attention before one holding a prepaid item, and a queue sorted by date will never tell you that.