In an economy where profit margins and infrastructure budgets are currently being squeezed to a sliver, the cost of "near enough" location data by itself is leading to dozens of massive store closures and groups of people running the risk of being excluded from e-commerce.
Sophie Hasell, CRO of AfriGIS, argues that the failure to start to see things differently is costing the country billions in aborted logistics, failed retail expansions, and infrastructure projects that get delayed before the first servitude is even signed.
“Re-asserting South Africa's economic momentum doesn’t require a new sort of figurative map,” she asserts. “What is needed is moving from seeing locational data as points that relate to economic realities, and rather seeing them for what they are: an unrivalled base for contextual insights that can bridge the gap between abstract data and real-world decision intelligence. And for that you need a strategic partner.”
The high price of the "last mile"
The crisis of the "last mile" in e-commerce is rarely about a customer not being home. “It’s more common for a delivery to fail because the driver couldn’t find the listed address than because the receiver wasn’t available,” says Hasell.
“Now the e-retailer loses part of a paper-thin margin when they have to try to deliver again at a later date. But, if the issue is the difficulty of finding an address on a map, it’s not like it will be easier next time.
“So, quite often, the e-retailer then loses even more money because of the even higher costs of having to return the item not to their own distribution warehouse, but eventually to the original supplier.”
This spatial friction is also present – if less visible – when it comes to infrastructure. For instance, as informal settlements expand, the window to secure a servitude for a substation or a transmission line is incredibly narrow.
If planners rely on locational data that is inaccurate or doesn’t contain all the relevant information, they may find a settlement has moved in a different direction, rendering the "ideal" spot for infrastructure a suddenly missed opportunity.
The multimillion-rand geography of failure
The retail sector offers perhaps the most expensive lesson in spatial intelligence.
Last year, a major national retailer was forced to shutter approximately 20 stores because they were fundamentally mismatched with their surroundings. These are multimillion-rand bad decisions born of the assumption that being "in the ball-park" is sufficient.
Successful trade mapping requires a single, definitive answer to a complex set of needs.
A fast-food franchise, for instance, operates on a razor-sharp set of variables that have to come together just right: it must sit on a primary taxi route, be within sight of a high-traffic grocery anchor, and target a specific income demographic – all while maintaining a calculated distance from existing stores and competitors.
Without this level of precision, an investment becomes a gamble.
“If simply ‘most’ of the variables are accounted for in the trade mapping exercise it is not a guarantee that a venture will work out. Was the locational data mix missing the taxi route insights? That could be the make-or-break one,” says Hasell.
And this local nuance is even more so where global tech giants can stumble.
Hasell recalls a three-month process of trying to explain to a major e-commerce entrant why the global postal code model is a non-starter in South Africa. While a postal code might be a pinpoint in London, in a local context, it is a blunt instrument.
A single South African code can mask dozens of vastly different economic realities, from wealthy hubs to informal zones, or contain anything from a few city blocks to 49 suburbs, making it a dangerous metric for high-stakes decision-making.
“It comes down to realising that real commercial and economic challenges relating to the physical world – and the locations in it – can’t be solved with ‘more data’,” Hasell explains. “It’s insights that matter. Strategy is built on insights, not more data.”
Another example Hasell refers to is around how insights can only come from data that is verified to the maximum degree – everything else is guesswork.
"Mapping the informal economy, for instance, requires clarifications and classifications that a high-level approach simply cannot see," she notes.
