The $76 billion network build still needs power and paying users

GSMA expects mobile operators across Africa to invest more than $76 billion in network infrastructure between 2024 and 2030. The figure describes capital over several years and markets. It doesn't say where each dollar lands, when equipment enters service or whether a new site earns its cost.

Follow one unit of capital. It can fund spectrum-related network deployment, radio equipment, towers, fibre, core systems or upgrades. After installation, the asset needs electricity, maintenance, security, backhaul and customers. The investment headline is the first line of that chain.

The site has to earn.

Capital reaches a site in stages

A network plan begins with demand and coverage engineering. The operator secures rights, permits, equipment and a tower or other location. It connects backhaul and power, integrates the site, tests it and places it into service.

Committed capital is not commissioned capacity. Supply delays, site disputes, currency movement and civil works can move completion. Report planned, contracted, installed and live sites separately.

Equipment is often imported, creating foreign-currency exposure between budget and payment. A weaker local currency can reduce the number of sites completed under a fixed local budget. Hedging and vendor finance have costs and eligibility rules.

Shared towers and fibre can reduce duplication. They also create lease obligations and service dependence. The operator's capital spending may fall while operating payments rise. Compare total cost across the asset life.

Electricity becomes recurring cost

The IEA estimates that about 600 million Africans lack reliable electricity access. A mobile site can operate in that environment with batteries, diesel, solar or a combination. Each option changes capital, maintenance and uptime.

Grid connection doesn't guarantee continuous service. Record outage hours, voltage quality, tariff and the fuel or battery response. A generator needs delivery and security. Batteries degrade and require replacement. Solar needs space, equipment and protection.

Energy cost per site varies with load and design. An operator should report the method before publishing a “green site” saving. Reduced diesel volume is measurable; avoided emissions require an emissions factor and boundary.

Power also reaches the user. A covered customer who struggles to charge a phone may use less data. Network energy and household energy sit on opposite sides of the revenue calculation.

Backhaul decides usable capacity

A radio can advertise a generation label while congested backhaul limits speed. Fibre offers capacity and can be expensive to deploy or repair. Microwave and satellite solve different terrain and volume problems.

Measure busy-hour throughput, latency, packet loss and availability, not peak laboratory speed. A site may carry little traffic because demand is low or because service quality drives users away. The diagnosis changes the investment decision.

International capacity and local interconnection affect services hosted outside the market. Caching and local hosting can reduce repeated transit, subject to enough demand and data-centre availability. The network plan should identify where traffic goes.

Resilience needs route diversity. Two fibre contracts that share the same trench do not provide independent paths. Operators and enterprise buyers should ask for physical-route evidence under confidentiality.

Coverage meets the usage gap

GSMA reports that 63% of Africans live within mobile-broadband coverage but do not use mobile internet, compared with a 9% coverage gap. Additional coverage remains necessary for the uncovered population. In already covered areas, device, data, skills and relevant service may constrain revenue more than radio reach.

The commercial case should separate new coverage, extra capacity and service-quality upgrades. New coverage adds potential customers. Capacity protects experience where traffic already exists. Quality investment can reduce churn or support enterprise services. Each has a different payback.

Average revenue per user can hide affordability. A small high-spending segment may support the average while most covered people remain offline. Use distribution, active data users and consumption bands.

Taxes and fees influence retail price, but pass-through is not automatic. Competition, currency and device inventory affect what the customer pays. Any affordability intervention needs price monitoring after implementation.

Enterprise revenue requires more than connectivity

GSMA reports that 79% of surveyed African operators identify becoming a digital-transformation partner as a primary enterprise objective. Selling cloud, security, identity or network APIs requires product integration, service levels and support beyond a data plan.

Enterprise contracts can improve revenue predictability and demand resilience. They can also impose penalties and custom work. Allocate sales engineering, onboarding and support to the product margin.

A network API may monetise an existing signal such as number verification. It still needs authentication, consent, developer support and liability rules. Revenue does not appear because the network owns data.

Public-sector connectivity can anchor deployment, subject to procurement and payment risk. A signed contract and funded budget matter more than an announced partnership.

Read investment against returns

For each deployment programme, publish capital placed in service, incremental operating cost, active users, traffic, service quality and revenue or public-service outcome. Use cohorts so old sites do not mask weak new ones.

Payback depends on financing and currency. A dollar-funded asset earning local-currency revenue carries another risk. Long asset life can support long finance, but technology and competitive change may shorten economic use.

The $76 billion expectation shows the scale of operator commitment GSMA anticipates. It should not be divided by population and called a benefit. Some capital replaces old equipment; some expands reach; some creates capacity customers may not buy immediately.

A tower becomes infrastructure when it goes live. It becomes a sustainable service when power, backhaul and customer cash arrive often enough to keep it there.

Sources

• GSMA, 16 June 2026: The Mobile Economy Africa 2026
• International Energy Agency, 6 October 2025: Stepping Up the Value Chain in Africa