In 1994, South Africa made a promise: that no citizen would be excluded from services, opportunity, or dignity based on where they lived.
In the month we celebrate Freedom Day, Geospatial data company AfriGIS is drawing attention to the solutions addressing gaps in the country’s digital infrastructure: millions of South Africans live in homes, communities, and settlements that existing address systems do not fully recognise.
The consequence is exclusion from bank accounts, insurance, municipal service delivery, emergency response, and economic participation. The solution? Inclusion.
“When we talk about address data in South Africa, people assume we’re talking about a street name and a number. But the reality is far more interesting,” says Marna Roos, account manager and geospatial acientist at AfriGIS.
“South Africa has 14 address types and most of the systems that determine who gets a loan, who gets insured, or who gets services are only looking at one or two of them.
“The people who fall outside those types can’t merely be said to have incomplete addresses, because, in essence, they have addresses that the system was never designed to see,” Roos adds.
“When a citizen’s address does not match an expected format, such as a street name, a street number, a postal code, systems flag it as incomplete, unverified, or high-risk. The end result is that the person is turned away as if they didn’t exist.”
South Africa has 14 recognised address types and AfriGIS holds verified data across all of them. From formal street addresses in established suburbs to site addresses, sectional title schemes, points of interest, rural farm addresses, and informal settlement identifiers, each type reflects a distinct way that South Africans occupy and navigate their country.
“Used in isolation, any single layer produces an incomplete picture,” says Roos. “Used together, they reveal the full, complex geography of a nation.”
Roos explains that the deeper problem is not so much missing data, but a misplaced assumption about what an address should look like.
When a system expects a street name and number and receives something different, it does not conclude that the address belongs to a different class. What the system does is conclude that the address is wrong.
That perception of incompleteness triggers rejection. And rejection, repeated at scale across banking, insurance, healthcare, and emergency services, becomes a form of structural exclusion that is invisible precisely because it hides behind process.
“We have seen areas classified as undeveloped (with no development flag attached to them in the data) and on that basis, residents were simply not considered for services. The area wasn’t actually undeveloped – the data was just incomplete,” explains Roos.
“That distinction matters enormously, because when decisions are made on a partial picture, the consequences are very real for the people living there. It happens every day, and it happens because organisations are looking at one piece of the puzzle and treating it as the whole.”
The more complete your address data, the more democratic your decision-making. When a municipality allocates budgets, when a financial institution assesses risk, or when a logistics company plans last-mile delivery, each of these decisions is only as equitable as the data that informs it.
“An organisation that operates on just one out 14 address classes is not making informed decisions,” Roos explains. “The reality is that such an organisation is – unwittingly – just making assumptions. It is precisely those assumptions that consistently disadvantage the same communities that were marginalised before 1994 – most often in the informal settlements that are growing while service delivery lags behind.”
AfriGIS is calling on the government, financial institutions, and the private sector to recognise the multilayered addressing ecosystem that supports the democratic landscape of our nation.