How to reduce failed deliveries in Saudi Arabia: 4 steps
A share of KSA deliveries fail before the driver is ever assigned, because standard geocoders can't place the address at all. The fix is a data-layer change, with no checkout flow to rebuild.
Failed deliveries in Saudi Arabia frequently trace back to a structural problem in how addresses are recorded. In 2022, Jeddah Municipality classified 64 of the city's 210 districts as informal settlements. An Amnesty International review of a Jeddah Municipality document from the same year identified more than 558,000 residents across 60 of those neighborhoods, and that figure covers only the districts targeted for demolition, not the full informal footprint of the city. Residents of these areas navigate by landmarks, mosque names, and alley descriptions rather than street numbers, and any geocoding tool built on Saudi Arabia's National Address system (Wasel) returns nothing for those addresses.
The fix is to pass checkout addresses through a geocoding API built for KSA before route assignment, with no change to your checkout flow required. A MEA-specific database covers informal settlements and accepts free-form text, converting landmark-based instructions into delivery coordinates before a driver is dispatched.
Jeddah classified 64 of its 210 districts as informal settlements. Standard geocoders have no data for any of them.
The National Address system assigns formal street names, building numbers, districts, and 4-digit zip codes. In established urban areas such as central Riyadh, commercial Jeddah, and downtown Dammam, coverage is high and standard geocoders perform reliably. The gaps are structural, not accidental:
These residents give delivery drivers directions the way anyone does locally: landmarks, mosque names, alley descriptions. For a geocoder built on National Address records, those directions are unresolvable. In Jeddah, that gap covers nearly a third of the city's districts. The result is a failed delivery, not because the address doesn't exist, but because the geocoding layer can't place it.
Why free-form address text breaks standard geocoding tools
A standard geocoder such as Google Maps, HERE, or an OpenStreetMap-based tool converts address text to coordinates by matching against a structured database. That works when addresses are formatted consistently and the database is complete. In KSA, two things break that assumption.
Informal settlements aren't in the National Address system, new developments take months to be indexed, and rural areas are sparsely covered. A geocoder can only match what's in its database.
"Near the blue water tower, left after the petrol station, ring the intercom" is a real instruction. A geocoder built for "14 King Fahd Road, Al Olaya" can't parse it, and returns nothing.
How Blassa geocodes what standard tools miss
Blassa's geocoding API is built for the two specific problems above, coverage gaps and free-form text, and resolves them in a single call before any driver is assigned.
Nothing in the National Address–based database matches. The dispatcher gets a blank and falls back to a district centroid.
A MEA-specific database covers the settlement; free-form parsing reads the landmark cues and returns a structured point with a confidence score.
4 steps to fix the problem without changing your checkout flow
Each step slots into an existing dispatch workflow. Together, they move address-quality problems from the field to the data layer, before a driver is assigned.
Customers write addresses however they do today. Blassa handles the conversion, and informal-settlement addresses have no National Address code to enter anyway.
One geocoding call between checkout and dispatch returns a coordinate and confidence score. High-confidence results route directly.
A low score triggers a short human check before dispatch, replacing the silent fallback to an approximate centroid.
Measure the share returning a usable coordinate per area. It shows where coverage was weak and confirms improvement over time.
Standard geocoding tools weren't built for KSA's address conventions. The markets they were optimized for don't have nearly a third of a major city's districts classified as informal settlements, with residents navigating by mosque names and alley descriptions instead of building numbers. Blassa's MEA-specific database and free-form text geocoding were built for that context specifically. For more on the structural reasons standard tools miss MENA addresses, see Why informal addressing systems in MENA break standard geocoders.
No. Blassa's API accepts free-form text input, including landmark-based and informal descriptions. The geocoding layer handles the conversion, so customers write exactly as they do today.
Jeddah is the best-documented case. In 2022, the city classified 64 of its 210 districts as informal settlements. Amnesty International's review of a Jeddah Municipality document from the same year identified more than 558,000 residents across 60 of those neighborhoods, though that figure covers only districts targeted for demolition under the Jeddah Central Project, not the full informal footprint. Riyadh has documented informal settlements as well; a 2023 study in MDPI Remote Sensing mapped them across multiple districts of the capital.
Most global geocoders are trained on markets where street numbering is complete and consistently formatted. In KSA, informal areas are physically large, densely populated, and give addresses by landmark rather than street number. Global tools weren't built for this convention; Blassa's MEA-specific database and free-form text processing are.
The integration is a single API call inserted into an existing dispatch workflow. For teams with a developer on hand, this is typically a day's work, not a platform migration.