Back to blog
Technical Deep DiveAddress Data

Why informal addressing systems in MENA break standard geocoders

A real delivery address that no official database has ever seen — and why the failure is structural, not a data-quality bug.

July 9, 20268 min read
Text to address · one informal input

The same address, two very different answers.

INPUT · JEDDAH
FREE-FORM DESCRIPTION
"Behind the water tank, second alley past the fish market, Briman neighborhood."
no street nameno building numberlandmark-based
STANDARD GLOBAL GEOCODER
No match — null result
Address not found in coverage
lat: · lng: · status: unroutable
PURPOSE-BUILT MENA GEOCODER
Resolved via landmark anchor
Briman, Jeddah · Makkah Province
LAT
21.62°
LNG
39.12°
SIGNAL
anchored
The same free-form address, resolved by two different classes of tool.

A driver in Jeddah receives a delivery address: "Behind the water tank, second alley past the fish market, Briman neighborhood." Google Maps returns nothing. Standard geocoding tools fail here because they match against official address databases — and informal settlements were built without formal street registration. There are no database entries to resolve against.

This post explains why the failure is structural, and what purpose-built MENA geocoders do differently.

Key takeaways
Informal settlements in Cairo and Jeddah lack formal street names and building numbers, so geocoders built on official databases return null results or incorrect coordinates.
Standard geocoders match input against municipal registries and postal databases; where no formal entry exists, no resolution is possible.
Geocoding failure surfaces in delivery logs as null results, centroid fallbacks, or incorrect partial matches — each producing the same outcome: a failed delivery.
Purpose-built MENA geocoders use landmark indexing, community-level reference data, and delivery feedback to resolve addresses that formal systems cannot.

Why standard geocoders fail in informal addressing environments

Standard geocoding tools work by matching input text against a reference database of known addresses. The geocoder parses the input, looks for street names, building numbers, postal codes, and administrative identifiers, then maps those tokens to coordinates in its database.

This process fails at two distinct points in informal areas.

1
No street names in the database

Streets in Cairo's ashwiyyat — unplanned settlements that grew outside formal urban development — often have no official names, or names assigned locally that never appeared in any government registry.

2
No building numbers

Formal addressing assigns numbers sequentially along officially surveyed streets. Informal buildings went up as residents could afford them, often room by room, and were registered years later, if at all.

In Jeddah, Saudi Arabia's National Address system has improved formal coverage in KSA's urban centers substantially. The limitation is the National Address system's reach: informal zones and rapidly expanding peri-urban neighborhoods at the city's edge often haven't been surveyed or assigned codes. An address requires official assignment to be in the system; addresses outside that coverage are simply not resolvable through it.

The matching mechanism

A geocoder can only match what a registry already holds.

Input address
Parse & match against reference database
Result
FORMAL · NASR CITY
12 Youssef Abbas St, Nasr City, Cairo
STREET
Youssef Abbas
NUMBER
12
POSTAL
11371
ADMIN
Cairo
Resolved
30.06°, 31.34°
INFORMAL · ARD AL-LIWA
Third house past the mosque, Ard al-Liwa, Giza
STREET
not in DB
NUMBER
none
POSTAL
ADMIN
Giza
Null
no coordinate
1
No street names in the database

Cairo's ashwiyyat grew outside formal development; streets carry local names that never entered any registry.

2
No building numbers

Informal buildings went up room by room and were registered late, if ever — sequential numbering needs a survey that never happened.

The same pipeline resolves a formal address and breaks on an informal one at two structural points.

How geocoding failures appear in delivery operations

When a geocoder cannot resolve an address, the outcome depends on how the tool handles the failure.

?
no coordinate returned
01
Null result

The geocoder returns nothing. The system marks the parcel unroutable and a human has to intervene.

status: unroutable
district centroidactual location
02
Centroid fallback
MOST DANGEROUS

Returns the district centre. It logs as a success, then fails when the driver arrives.

geocode: success · delivery: failed
right neighborhoodwrong building
03
Incorrect partial match

Finds a related entry and returns it with apparent confidence. Right general area, wrong building.

match: partial · confidence: apparent
All three produce the same outcome — a failed delivery, a re-delivery, a frustrated customer. Aggregate success rates hide it; segment by district and the gap appears.
Three failure modes — only the null result looks like failure in the log.

All three outcomes produce the same result: a failed delivery, a re-delivery attempt, and a frustrated customer. Centroid fallback is the most operationally dangerous, because the failure is invisible in the log — it only surfaces in delivery outcomes, or when you audit results against confidence scores.

Standard geocoding success rates aggregate across all address volume. Strong performance on formal urban addresses masks the failure rate in informal districts. Segment your results by neighborhood and the gap becomes visible.

Standard vs. MENA-specific geocoders: how they handle informal addresses

The two approaches diverge at every stage — from the reference data they draw on to how their coverage grows over time.

Approach A
Standard global geocoder
Approach B
Purpose-built MENA geocoder
Resolution method
Official address databases & postal registries
Landmark indexing, community names & delivery feedback
Informal-area coverage
Null result or centroid fallback
Partial-to-full via alternative reference data
Failure output
Null, centroid, or incorrect partial match
Coordinates + confidence score; low scores flagged for review
National Address / formal coverage · KSA
High in formally surveyed urban zones
Formal zones, plus peri-urban extension data
Coverage growth over time
Depends on external database provider updates
Grows with delivery volume in each specific market
The two approaches diverge at every stage of resolution.
See how Blassa resolves the addresses your current geocoder can't — in your own coverage area.

How purpose-built MENA geocoders resolve informal addresses

Geocoders built for MENA use approaches global tools do not have — drawing coordinates from the data that does exist in informal areas.

Ard al-LiwaImbabaBoulaq el-Dakrour
01
Community & neighborhood indexing

Community names are meaningful location identifiers even where street names are absent from any registry.

mosque · anchorschoolpetrolwater tower
02
Landmark anchoring

Mosques, schools, water towers and petrol stations are indexed as coordinate anchors for nearby addresses.

each delivery → a georeferenced point
03
Delivery history feedback

Every successful delivery is a data point. Coverage accumulates from real outcomes and grows with volume.

The result: coordinates with a confidence score where formal systems return nothing — anything below threshold is flagged for human review.
Three sources of reference data that exist where formal addresses do not.

What this means for delivery operations in MENA markets

In Cairo, a large share of delivery addresses in Giza, Shubra, Imbaba, and Ard al-Liwa fall in categories that standard geocoders handle poorly. In Jeddah, neighborhoods beyond the National Address system's formal survey reach face the same problem. For operators working across both markets, this is a routine occurrence, not an edge case.

The practical implication: if your geocoding tool is returning null results or centroid fallbacks on a noticeable percentage of your address volume, the problem is probably not data-entry quality. The addresses your customers submit are the actual addresses they have. Customers in Ard al-Liwa do not have a National Address system code to provide. The question is whether your geocoder resolves the format in which those addresses actually exist.

Check your geocoding failure rate by district, not in aggregate. That breakdown will show where informal-area failures are accumulating — and the aggregate success rate is not.

Ready to see your real geocoding failure rate across Cairo and Jeddah?
Frequently asked
Do informal settlements in Cairo and Jeddah have no formal addresses at all?

Some do. Cairo's older ashwiyyat, such as Imbaba and Ard al-Liwa, have partial street naming from retrospective mapping efforts. Coverage is uneven and incomplete, and a geocoder built on official datasets will still fail on a significant share of addresses there even where some formal naming exists.

Doesn't Saudi Arabia's National Address system solve this in KSA?

The National Address system has improved formal address coverage in Saudi Arabia's urban centers significantly. The limitation is reach. Newly developed peri-urban zones and areas settled before the National Address system rollout often have not been assigned codes. An address must be in the system to be resolved through it — and many informal and peri-urban areas in Jeddah and other KSA cities are not.

Why does my geocoder sometimes return a result for an informal address instead of failing?

Most geocoders fall back to district-level or partial matches when they cannot resolve a specific address. These outputs appear as successful geocodes in the log but produce incorrect delivery coordinates. The failure only surfaces when you audit delivery outcomes against confidence scores rather than resolution rates.

Can better address input from customers fix this?

In most cases, no. Customers in informal areas submit the actual addresses they have — landmarks, local names, and proximity descriptions rather than formal street numbers. There is no formal address to provide. The fix is a geocoder capable of resolving how those addresses are actually written.

How would I know if my current geocoder is missing addresses in Cairo or Jeddah?

Segment your geocoding results by neighborhood for your actual delivery address volume. Compare null rates and confidence scores in informal-area districts (Ard al-Liwa, Shubra, Imbaba in Cairo; older peri-urban zones in Jeddah) against your city-wide average. That segmented view shows what the aggregate hides.