INTERVIEW: Why Can’t Africa’s Digital Infrastructure Keep Pace With Its Digital Economy?

In this exclusive interview with Impact Newswire, Katie Hill, Partner and Associate Director at Boston Consulting Group discusses how emerging markets can close persistent gaps in digital infrastructure without replicating the inefficiencies of earlier development cycles, why digital infrastructure has become a critical economic foundation, why investment is still lagging behind demand, and what it will take to structure capital, regulation and delivery models that can finally close the gap at scale.

Digital infrastructure has become the central layer of modern economic systems because most value creation, from payments to logistics to government services, now depends on always-on connectivity, compute capacity and data exchange. In Africa, this shift is especially pronounced because digital platforms are rapidly substituting for legacy infrastructure in financial services, public administration and commerce. 

Yet despite strong demand growth, the underlying infrastructure base remains thin. The World Bank estimates Africa holds about 14% of global internet users but less than 2% of global digital infrastructure capacity, underscoring a structural mismatch between usage and underlying systems .

That imbalance is most visible in connectivity. As of 2025, only about 37–38% of Africans are online, meaning roughly two-thirds of the continent remains digitally excluded, largely due to cost and infrastructure constraints . Even where users are connected, the quality of service is often constrained by limited fibre backbones, congested mobile networks and high international bandwidth dependency. 

Sub-Saharan Africa has also seen fast growth in mobile subscriptions, but subscriber density per tower remains significantly higher than in other regions, signalling persistent capacity stress on existing infrastructure . This translates into higher prices, lower speeds and uneven access between urban hubs and rural economies.

The data centre and cloud layer illustrates an even sharper gap. Africa accounts for only a small fraction of global data centre capacity, estimated at around 1–2% despite its large and growing internet user base . At the same time, demand for digital services is accelerating rapidly: mobile data traffic across the continent is growing at roughly 40% annually, nearly double the global average. This divergence between demand growth and local hosting capacity means that much of Africa’s digital activity is still routed through infrastructure outside the continent, increasing latency, costs and regulatory dependence.

The investment response has not matched the scale of the gap. Annual digital infrastructure spending needs in sub-Saharan Africa are estimated at about $7–8 billion through the second half of the decade, covering fibre, mobile networks, cloud systems and data centres . However, investment is heavily skewed toward mobile radio networks, which absorb around 60% of capital expenditure, while data centres account for less than 7% of total spend in recent years . This creates a structural imbalance: connectivity expands faster than the compute and storage layers required to fully utilise it, leaving the digital stack incomplete.

This is where financing structure becomes as important as volume. Pure private-sector models struggle because returns on core infrastructure are long-dated, capital intensive and often dependent on anchor tenants such as governments or large platforms. As a result, Public-Private Partnerships are increasingly being re-evaluated as the default mechanism for scaling digital infrastructure. 

In particular, Build-Operate-Transfer (BOT) and Build-Own-Operate-Transfer (BOOT) structures are being revived for fibre networks, national data centres and government digital systems, because they allow private capital to finance upfront costs while governments retain long-term ownership or control. These models are not new to Africa; they have already been used extensively in energy generation, transport corridors and telecom infrastructure, where concession-based financing has long underpinned large-scale asset development.

The key constraint is no longer conceptual but executional. BOT and BOOT models succeed in digital infrastructure only when demand is contractually anchored, typically through government digital services or regulated utility-like usage; when regulatory regimes are stable enough to support 10–25 year concession horizons; and when interoperability standards ensure that systems are not fragmented across competing platforms. 

Equally important is the ability to align data governance and sovereignty requirements with private financing structures, since governments are increasingly unwilling to host critical systems outside national jurisdiction. Where these conditions are met, BOT-style models can close financing gaps at scale; where they are not, they tend to replicate fragmentation rather than resolve it.

Editor Faustine Ngila spoke with Katie Hill, Partner and Associate Director at Boston Consulting Group, Nairobi, where she focuses on large-scale digital transformation, infrastructure strategy, and public sector modernisation. Here is the interview excerpt:

Q1. Why is digital infrastructure now being treated as critical infrastructure alongside roads, ports, and energy systems? What specific economic functions does it now enable that make it foundational to growth and competitiveness?

Digital infrastructure is now as foundational to growth as roads, ports, and power. It enables commerce, payments, health, education, public services, AI, and regional trade. Additionally, AI has shifted digital infrastructure from a useful enabler to a competitive necessity for countries and businesses. Globally, data centre capacity is set to roughly double by 2030 with around 70 percent of that demand driven by AI, against an estimated USD 3 trillion of investment required. In Africa, the network grew mobile-first rather than fixed, and subsea cables have driven a sharp drop in bandwidth costs into hubs like South Africa. A way to think of it is that towers and fibre are the roads of the digital economy, subsea cables are its ports, and data centres are its industrial estates. A key difference is that once the digital layer exists, value scales faster and more broadly than with physical infrastructure, because it moves data, services, capital and knowledge much more freely.

Q2. Where are the most binding infrastructure bottlenecks in Africa’s digital economy today, across fibre backbones, subsea connectivity, data centres, cloud, and last-mile access, and what do current capacity and access metrics reveal about the scale of the gap?

The most urgent bottlenecks are fixed long-distance connectivity (reliable cross border connectivity, national fibre backbones), in-region data centre capacity and cloud infrastructure and fixed broadband, along with the cost to communicate. Africa sits at under 1 MW of data centre capacity per million people against a global range of 29 to 89, with mobile penetration at 44 percent against 71 globally. Most African markets now have reasonable thin mobile coverage, but devices and tariffs still price out lower-income users, and the last-mile business case for those segments is fragile. Investment lags because demand is fragmented (despite adoption increasing rapidly across populations, value is still trailing with low ARPUs), regulation and political environments can be uncertain, and many projects need long-tenor capital before revenues become predictable, coupled with uncertain FX environments.

Q3. Despite rising demand for connectivity and digital services, why has investment in African digital infrastructure not scaled proportionately? What are the main structural and financial constraints discouraging capital deployment?

The headline number is the gap itself: UNECA puts the unfunded shortfall across Africa’s digital infrastructure portfolios at up to USD 108 billion, and on current trajectory the gap is widening, not closing. Capital that does flow concentrates in a few hub markets and a narrow set of asset classes. Projects struggle to reach close because of currency volatility, capital controls, regulatory fragmentation across 54 jurisdictions, shallow local debt markets, and a thin technical talent pool in several countries. Capital remains cautious despite plausible macro demand as actual value has taken longer than expected to manifest (e.g., with early data centre builds outside SA). Closing the gap will likely require blended finance, sovereign anchor demand, and risk-sharing instruments calibrated by asset class.

Q4. Which parts of Africa’s core digital infrastructure stack remain most underdeveloped, and what are the systemic risks of continued dependence on offshore or external infrastructure for data, compute, and digital services?

The four layers that define the modern stack are data centres, subsea cables and national backbone, fixed broadband, and cloud and AI platforms, and Africa is short across all four. In terms of data centre capacity, its about having sufficient sovereign grade data centres, sufficient national backbones to drive down the cost to communicate within and also cross-border terrestrial fibre, additionally last-mile fixed broadband for local access, and in the longer term, sufficient domestic AI compute. The economic and sovereignty risk is that high-value workloads, especially AI training and inference, are processed offshore, exporting both data and surplus value and anchoring Africa’s dependence on foreign jurisdictions for critical national workloads. The West African subsea cable outages of the past two years showed how directly this exposure plays out, with multi-day disruption across several economies. This was reinforced by the European cable outages seen, and attacks on data centres in the Middle East additionally highlighted the strategic nature of these assets.

Q5. What Public-Private Partnership models are currently being used to finance digital infrastructure in Africa, and in which specific asset classes have they been most successful or least effective?

Typically there are ~3 distinct pieces: concessions for long-life assets with contracted, predictable revenues such as fibre backbone; co-investment or anchor models where a government or hyperscaler underwrites demand to crowd in private capital; and BOT/BOOT structures for the build-fund-operate-transfer of new assets. There are a number of African precedents in adjacent infrastructure that prove the mechanics with the common success traits being transparent licensing, anchor demand, open-access obligations, and credible dispute resolution. Failures usually trace back to opaque award processes, mismatched tariffs, or weak public-side counterparty capacity, rather than to the model itself. Within digital, the model has worked best so far on passive fibre, where asset lives are long and technology-cycle risk is low.

Q6. Why are Build-Operate-Transfer (BOT) and Build-Own-Operate-Transfer (BOOT) models re-emerging as relevant structures for digital infrastructure, and what has changed in markets or technology to make them viable again?

Three things have changed at once. First, capital intensity has outgrown most African public balance sheets, with AI-ready data centres now built at scales and costs that no fiscus can carry alone, while demand profiles are finally predictable enough to underwrite. Second, BOT and BOOT resolve the sovereignty tension directly: the private party builds, funds, and operates the asset, then transfers ownership, preserving a defined public reversion. Third, lenders and governments are already familiar with the model from power, ports, airports, and roads (e.g., Azura-Edo, Lekki, Blaise Diagne, N4), which shortens the bankability curve materially. However BOT models may not be a match for all types of digital infrastructure and should be considered judiciously, they make most sense for things like fibre, but are less relevant for classes that require ongoing innovation and distinct operational capabilities.

Q7. What are the minimum technical, financial, and institutional conditions required for BOT or BOOT models to successfully deliver digital infrastructure projects in African markets without creating long-term inefficiencies or stranded assets?

Four conditions determine close-ability in digital infrastructure specifically: a long-lived asset whose useful life exceeds financing tenor, a bankable off-take contracted or backed by a credible counterparty before financial close, proven and predictable demand, and a slow enough technology refresh cycle that core kit will not need major reinvestment mid-concession. I would layer four African market preconditions on top: an independent and technically credible regulator, access to forex or genuinely local currency revenue, pre-cleared land and rights of way, and a capable public-side counterparty. Where these are present, capital prices are at single-digit risk premia. Where two or more are missing, projects either stall or price beyond widespread adoption.

Stay ahead of the stories shaping our world. Subscribe to Impact Newswire for timely, curated insights on global tech, business, and innovation all in one place.

Dive deeper into the future with the Cause Effect 4.0 Podcast, where we explore the ideas, trends, and technologies driving the global AI conversation.

Got a story to share? Pitch it to us at info@impactnews-wire.com and reach the right audience worldwide