Designing Local Delivery Zones Around Real Fulfilment Capability

A delivery map is a promise to customers, not merely a set of coloured polygons. A postcode may sit inside a service area while roads, traffic, building access, vehicle restrictions, or order preparation make the advertised window unrealistic. Multi-branch retailers should design delivery zones from actual fulfilment capability and update them when capacity changes, rather than expanding coverage simply to make a campaign appear larger.

Build zones from delivery evidence

Build the zone model from completed deliveries and failed attempts. Examine travel time by daypart, distance, order size, building access, loading restrictions, and the branch where the order is prepared. Straight-line radius is a poor proxy for urban travel, especially where rivers, railway crossings, or restricted roads separate nearby neighbourhoods. Record the reason for a failed promise so geography can be distinguished from address quality or customer unavailability.

Separate serviceability from speed

Separate eligibility from speed. A location may be serviceable for scheduled delivery but not for same-day fulfilment, bulky goods, installation, or temperature-sensitive products. Display the available options only after the customer’s basket and address have been evaluated. A broad “delivery available” label should not conceal that the selected item requires a different vehicle or trained crew.

Calculate the complete promise

Set a promise using the full operating chain: order acceptance, picking, packing, dispatch queue, travel, and handover. Use a conservative percentile from reliable history rather than the best observed trip. Add contingency for predictable peaks, but avoid padding every order so heavily that the offer becomes meaningless. When the evidence is sparse in a new area, launch with a narrower schedule and expand after performance is measured.

Test edge addresses

Test the boundary with real addresses and edge cases. A zone can fail at a gated complex, a high-rise without loading access, a rural road with seasonal restrictions, or a business address closed outside office hours. Validate the delivery instructions and customer contact process. Do not promise an exact arrival time if the carrier can only provide a broad window. Explain what happens when the recipient is unavailable.

Coordinate campaign demand

Coordinate marketing with dispatch. A paid campaign can create a surge that overwhelms the very zone advertised as fast. Before launching, reserve capacity, confirm stock, and agree a pause trigger when the backlog grows. If the promise must change, update landing pages, checkout estimates, customer messages, and branch scripts together. Leaving one channel with an old window creates avoidable complaints and refund requests.

Route special cases

Create a controlled exception process for unusual orders. Large installations, fragile items, restricted products, or multi-stop business deliveries may require manual approval. The exception should collect the information needed to assess feasibility without implying confirmation before the team has accepted the job. Preserve the original request and the final agreed promise so customer service can resolve later disputes.

Review outcomes by zone

Review on-time delivery, first-attempt success, distance cost, cancellations, rescheduling, damaged goods, and customer contacts by zone. A high on-time percentage can look impressive if difficult addresses are quietly excluded after checkout, so include all accepted orders and track declines separately. Change boundaries only after identifying the cause and validating the effect on adjacent branches.

Field implementation note

Use a serviceability map that dispatch teams can edit through a controlled workflow. Drivers may identify a new road restriction or a building access problem before the analytics team sees it in the data. Require the report to include date, delivery type, and a reason code, then validate recurring reports before changing the public boundary. A single failed attempt may be caused by a customer not answering the phone, while repeated failures at the same complex may justify a specific access instruction or a revised promise window.

Decision record and follow-through

Make delivery estimates conditional on the actual basket, address, and operating window. Review rejected orders as well as successful deliveries so the map does not appear accurate by excluding difficult cases. Change the public promise only after dispatch evidence supports the new boundary.