South Africa finally gets a real broadband map showing last-mile gaps—and what it means for investment
For the first time, South Africa has a detailed view of where broadband infrastructure is actually available and where it isn’t. The breakthrough comes in the form of a geospatial information system (GIS) built to reflect real-world connectivity, down to extremely small areas—so planners can see the “last-mile” gaps that often get hidden in broad national averages.
Developed as part of the South Africa Digital Infrastructure Investment Study (SaDIS) 2025, the GIS work was commissioned by the Development Bank of Southern Africa (DBSA) together with the National Planning Commission. At a launch event in Midrand this week, digital economy strategist Pieter Grootes, from Networks Anonymous, described how the tool turns scattered infrastructure information into one unified, map-based dataset.
The key point is that this is not just a general coverage poster. It’s an analytical base that can show, in practical terms, who has access and who doesn’t—down to a scale that can reflect challenges within city blocks.
Broadband map in focus: turning connectivity data into investment decisions
Grootes said the mapping approach uses a geospatial indexing method that divides the country into more than 1.5 million hexagonal cells, each covering about 0.76 square kilometres. That design is what makes the dataset unusually granular for understanding broadband reach at the neighbourhood level, rather than simply counting towers or fibre lines at a high level.
“We’ve built a GIS database that formed the analytical baseline of everything we have,” Grootes said. In his explanation, the dataset links where households are to where key infrastructure sits—so it becomes possible to work out connectivity access within small geographic areas, not just across provinces or municipalities.
He added that the study could essentially identify the people or communities inside a dense urban environment who have service and those who still don’t. That level of precision matters because broadband roll-out planning often wastes money when infrastructure build decisions are based on incomplete or overly averaged information.
The goal of SaDIS is ambitious: determine what investment is needed to connect South Africa’s population with 100Mbit/s broadband, supporting a digital economy where more citizens can fully participate. But to invest effectively, planners need to know where the current network landscape leaves holes—particularly in the final stretch from existing networks to homes, schools, clinics and other services.
A major reason the GIS is being treated as a serious planning tool is that it helps avoid the type of fibre overbuild that can happen in places where networks are already present in theory, but not optimised for real demand or realistic deployment needs.
To build the map, the study overlaid multiple data sources. Coverage information from regulator Icasa for mobile networks formed one layer, while fibre mapping data from the International Telecommunication Union helped inform another layer. The GIS also incorporated information from fixed-wireless providers, fibre operators and publicly available network data.
The resulting product is a layered broadband map capturing different technology realities at once: mobile coverage across 2G through 5G, fixed-wireless access deployment, and fibre node proximity. Instead of treating space as one distance, it breaks down fibre closeness into four bands—within 5km, 5–10km, 10–25km and beyond 25km—to help planners understand where fixed-line connectivity is feasible without subsidies.
What the broadband map suggests about mobile access
One of the clearest findings highlighted at the launch is the state of 4G mobile coverage. According to the analysis, 4G reaches 98% of the population. That compares with about 80% coverage in emerging-market peer Brazil, suggesting South Africa’s mobile reach is relatively strong.
Even so, the map points to persistent deprivation. The estimate is that roughly 400,000 households still have no access to 4G or better. That means mobile coverage alone does not automatically translate into usable broadband access for every community.
The study also created a composite scoring model for local government. Each of South Africa’s 213 municipalities was assigned a weighted “broadband access score,” designed as an early triage indicator for where universal service support should be prioritised. In this scoring system, 80% of the score comes from fibre node proximity, 13% from 4G/5G coverage, and 7% from proximity to wireless access nodes.
Municipalities that connect to neither mobile nor fibre broadband were identified in Mpumalanga, the Northern Cape and KwaZulu-Natal. Grootes linked the pattern partly to terrain and topographical constraints that can make infrastructure deployment significantly harder and more expensive.
Still, there are reasons for optimism. The study found that 74% of households live within 10km of a fibre node. That is important because it suggests a large share of the population could potentially receive fibre-like speeds—either through fibre connectivity to locations where mobile towers exist or by using mobile-based approaches that deliver higher-speed service than older solutions.
The schools and clinics angle: gaps are also structural
Where the map becomes especially revealing is in its analysis of public services. Grootes said more than 6,700 of South Africa’s roughly 15,000 public schools sit between 10km and 25km from a fibre node. In the same distance band, the study indicates more than a thousand health clinics fall within reach of potential upgrades.
Those figures do more than show “opportunity.” They also expose structural exclusion. They suggest many public institutions remain located far enough from fibre that commercially viable fixed broadband deployment may not reach them without intervention.
The map also highlights a more difficult hurdle: 12.2% of households are more than 20km from the nearest fibre node. In practical terms, Grootes described this as effectively outside the reach of any commercially viable fixed-line rollout unless subsidies or public support are introduced.
A missing piece in policy: data that investors and government can trust
The study’s authors criticised government for not making relevant broadband location data available in a way that supports investment decisions. While South Africa has built broadband mapping as part of SA Connect policy work in the past, the report says there is no current public database that government uses for policy planning with updated information.
Grootes argued that missing or outdated data is not just an ICT problem—it reflects a broader national capability issue around keeping statistics current and clean. He said the country needs a competence base to maintain accurate mapping and geo-referenced information.
The study also criticised Icasa, saying it only publishes mobile coverage data despite having legal authority to compel licensees to submit the necessary infrastructure information. In its recommendations, the study urges the communications department to prioritise completion and operationalisation of a national GIS database that allows broad stakeholder contributions. It also calls on Icasa to open a formal regulatory inquiry to require licensees to submit infrastructure data in standardised formats through an online portal.
For the public and private sector, the message is straightforward: you can’t plan universal broadband properly without knowing precisely where the gaps are—and now South Africa finally has a tool designed to reveal them with real geographic detail.
The next step will be what happens with the dataset itself. While the GIS data currently belongs to the DBSA and isn’t publicly available, Grootes said the DBSA will eventually decide whether to open-source the maps. If that move happens, South Africa could gain a powerful planning resource—and possibly reduce wasted fibre spending by targeting investment where broadband access still truly falls short.