Connecting African Schools: Procurement, Power and Service Accountability for Digital Education

EXECUTIVE THESIS

A connected school is an operating system, not a cable endpoint

On 3 September 2026, UNICEF invited 96 companies to bid for long-term agreements covering internet connectivity, on-site networks and solar power for schools and health facilities across Africa. The procurement spans all 54 countries. Its stated universe exceeds 860,000 schools and 260,000 health facilities; more than 500,000 African schools are already recorded on Giga Maps.[1]

This is not simply a large telecommunications tender. It is a test of whether public procurement can join several fragmented systems: backbone and last-mile networks; school electricity and local-area networks; teacher capability and digital content; cybersecurity and child safeguarding; repair and maintenance; recurrent budgets; and public performance data.

The distinction matters because coverage is not use, and connection is not learning. In 2024, 86% of Africa’s population lived within mobile-broadband coverage, but internet use in sub-Saharan Africa was only 38%, compared with 68% globally. In rural Africa, a quarter of the population remained outside mobile-broadband coverage.[2] A school can therefore sit inside a nominally covered area while lacking affordable service, reliable electricity, functioning equipment, trained staff or the institutional capacity to keep the service alive.

The procurement architecture is promising. UNICEF says contracts will specify speed, uptime and repair requirements, require safeguarding and cybersecurity standards, and monitor performance through real-time and open data. Solar power is included where grid electricity is insufficient.[1] These features move the unit of procurement away from equipment delivery and towards an observable service.

They do not, by themselves, guarantee educational value. A low-cost megabit that fails during the school day is expensive. A functioning network without a teacher-use plan is infrastructure without an operating model. A donor-funded connection without a recurrent budget is a future disconnection. ITU’s work on digital public infrastructure in education reinforces the need for interoperable, scalable systems rather than isolated technology purchases.[11]

STONECOMMS ORIGINAL SYNTHESIS

StoneComms proposes a Connected-School Accountability Contract built around five linked tests: Reach, whether the right technology reaches the right school; Readiness, whether power, local networks, devices and safeguards make it usable; Reliability, whether speed, uptime and repair meet an enforceable standard; Use, whether teachers and learners can turn the service into educational practice; and Renewal, whether funding, competition and maintenance can sustain it beyond the first contract. The weakest test determines the value of the connection. Reporting installations without reporting all five can overstate progress.

WHY THIS MATTERS NOW

The procurement window can shape a market, not only buy a service

Proposals for the new UNICEF long-term agreements are due on 6 November 2026, with the first agreements expected in 2027.[1] The immediate transaction is a framework through which governments and partners can buy connectivity, local networks and solar power more consistently. The wider opportunity is to establish common expectations for performance and transparency across an otherwise fragmented market.

Africa’s demographic and educational stakes are unusually high. The World Bank reports that 251 million children and young people were out of school globally in 2024, while more than half of ten-year-olds in low- and middle-income countries could not read and understand an age-appropriate text. Eighty-seven per cent of those children were in sub-Saharan Africa.[3] Globally, around 1.3 billion school-age children—roughly two-thirds of the total—lack internet access at home.[3]

Digital infrastructure cannot solve these education problems on its own. It can, however, change what education systems can distribute, observe and support. A reliable connection can give teachers access to training and curriculum materials, allow education ministries to manage schools and examinations, widen access to adaptive or assistive resources, and help learners reach content unavailable locally. During disruption, it can also provide continuity. These benefits depend on pedagogy, language, inclusion and institutional capability as much as bandwidth.

The tender arrives after several years of Giga work on school mapping, connectivity modelling, procurement and financing. Giga reported engagement across 47 countries and territories in 2025, with mapping and modelling completed in nine and under way in 21.[4] That accumulated information can reduce one of the most basic public-sector failures: procuring for facilities whose locations, power status, existing service and technical options are not known accurately.

The timing also creates an accountability risk. Large numbers make persuasive announcements. They can encourage progress to be counted as schools contracted, schools passed by infrastructure or schools with equipment delivered. None is the same as a school receiving usable, dependable service during teaching hours. The design should therefore lock performance evidence into the contract before scale makes weak definitions hard to change.

THE UNIT OF PROCUREMENT IS CHANGING

From equipment purchase to a measurable public service

Traditional school-technology programmes often separate connectivity, power, devices, training and maintenance into different projects, budgets and contractors. When something fails, each component can appear compliant while the classroom remains offline. The internet service provider delivered a router; the solar contractor installed panels; the ministry distributed devices; the training programme held a workshop. The combined service still may not work.

The UNICEF procurement begins to correct this by grouping three interdependent service categories: internet connectivity, on-site network infrastructure and solar power.[1] It also introduces requirements that should make service quality visible: minimum speeds, uptime, fault-resolution expectations and monitoring. Long-term agreements can give buyers standard commercial terms while preserving competition among qualified providers.

Standardisation has economic value. It can reduce repeated tender design and due-diligence costs, make service offers easier to compare and allow demand to be aggregated. It can also create a stronger signal to satellite operators, mobile-network operators, fibre providers, equipment firms and energy-service companies that school connectivity is a recurring market rather than a sequence of isolated pilots.

But standardisation must not become technological rigidity. The appropriate solution for an urban school near fibre is different from that for a remote school beyond terrestrial coverage. Some sites require fixed fibre; others may use microwave, mobile broadband or satellite. Some have stable grid power; others need solar generation and storage. A common framework should standardise outcomes and evidence while allowing the access technology and operating model to match local conditions.

The most useful contract therefore defines what a school must experience, not only what a supplier must install. At minimum, it should state service availability during teaching hours; actual rather than advertised throughput; latency where it affects applications; performance within the school, not merely at the external handoff; time to acknowledge and repair faults; energy availability; security responsibilities; and the data that will be published.

AFRICA’S CONNECTIVITY GAP IS ALSO A POWER AND USAGE GAP

Network coverage is necessary, but it is an incomplete readiness measure

The International Telecommunication Union’s 2025 broadband assessment illustrates the gap between infrastructure reach and meaningful use. Mobile broadband covered 86% of Africa’s population in 2024. Yet internet use across sub-Saharan Africa remained at 38%, and 25% of rural residents were outside mobile-broadband coverage.[2] Affordability, skills, relevant services, devices and electricity help explain why a coverage map is not a participation map.

Schools intensify these dependencies. A network may reach the boundary while internal wiring or Wi-Fi does not reach classrooms. Devices may be too scarce, insecure or difficult to maintain. Power may be intermittent. The available data bundle may be exhausted before the end of the month. A head teacher may avoid using a system because replacing damaged equipment would be impossible. These are not secondary implementation details. They determine whether the public asset produces a service.

Solar power can make the package materially more useful, particularly where grid extension is slow or supply is unreliable. UNICEF describes examples in which solar systems support connected classrooms and digital learning in remote settings. In Eritrea, 12 schools across four remote regions received systems with solar panels, batteries and computer equipment, serving about 6,500 students. Madagascar’s solar-powered Eco-Villages combine learning and community services, reaching thousands of young people and women.[5] These are programme examples rather than controlled impact evaluations, but they show why energy belongs inside the service design.

Energy also creates lifecycle obligations. Batteries degrade; inverters fail; panels need inspection; theft and physical damage occur. Procurement should specify maintenance, replacement responsibility and end-of-life handling rather than treating solar installation as a one-off capital item. E-waste and battery disposal are part of the contract’s environmental integrity.

The same principle applies to local networks. Buying internet access without designing secure, manageable Wi-Fi or wired access inside the school can leave bandwidth concentrated in an office. A service-level agreement should measure the user environment where feasible, with clear responsibility for routers, access points, configuration, filtering, authentication and fault diagnosis.

POOLED PROCUREMENT CAN CHANGE THE ECONOMICS

Comparable demand and performance data can improve bargaining power

Giga reports that bulk procurement and improved market information have reduced school-connectivity costs by 30–60% in several programmes.[6][12] In Rwanda, a procurement covering 63 schools reportedly reduced the price per megabit from US$20 to US$9. In Kenya, the use of the National Optic Fibre Backbone Infrastructure and improved performance monitoring contributed to a reported 60% reduction. In Botswana, Giga reports a 93% fall in cost per megabit while 700 schools serving around 400,000 learners were connected by the end of 2024.[6]

These figures are significant but should be interpreted carefully. They are institution-reported results from specific procurements, not guaranteed savings for every market. The starting price, service specification, contract duration, public infrastructure, technology mix and level of competition differ by country. A percentage reduction can also conceal weak absolute service if the baseline product was poorly specified.

The underlying mechanism is nevertheless credible. School demand is usually dispersed across thousands of sites and multiple public agencies. A provider asked to serve one remote school faces high acquisition, installation and maintenance costs. A provider offered a mapped portfolio, standard requirements and longer contract can plan capacity, combine routes, negotiate equipment and reduce commercial uncertainty. Buyers can compare bids and monitor a common service definition.

Rwanda also shows the inclusion value of explicit portfolio design. Giga reports that the 63-school package included schools in the Eastern Province and GS Nyabiheke, which serves about 800 refugee students.[7] Bundling hard-to-serve sites with a wider programme can stop remoteness or refugee status from becoming an implicit exclusion criterion. The contract must still disclose whether remote sites receive comparable performance and repair times.

Sierra Leone offers a financing lesson. Giga reports that evidence from a 42-school pilot contributed to a US$5 million Islamic Development Bank loan for wider connectivity.[6] A well-measured pilot can therefore do more than prove technology; it can create the cost, demand and performance evidence needed for a larger financing decision.

PRICE IS ONLY THE FIRST PERFORMANCE TEST

The cheapest compliant bid may not be the lowest-cost education service

Public buyers need competitive prices. They also need to distinguish savings created by aggregation and infrastructure access from savings created by reducing service quality, maintenance or resilience. A school connection that works for the acceptance test and deteriorates afterwards can produce a low procurement price and a high public cost.

Performance-based contracts should make payment depend partly on independently verifiable service. Giga’s proposed Connectivity Credits illustrate one possible mechanism: providers earn credits only when a specified performance threshold is met, with pilots explored in Malawi, South Africa, Kenya and Mexico.[8] The model is not yet a universal solution, but it sharpens an important idea. Public or concessional support should reward delivered service, not merely network presence.

Monitoring needs careful governance. Real-time data can reveal outages, actual speed and chronic underperformance. Open data can help regulators, ministries, communities and competing suppliers see where markets fail. Yet measurement equipment can malfunction, schools can lose power and providers can dispute responsibility. Contracts should define measurement points, data ownership, audit rights, exceptions and the process for resolving contested results.

Cybersecurity and child safeguarding are equally substantive. A connected school introduces risks involving personal data, harmful content, account compromise, surveillance and unauthorised access. Suppliers should not determine these protections alone. Education authorities need standards for data minimisation, retention, access control, content governance and incident response. Teachers and school leaders need practical procedures, not only policy documents.

Vendor concentration is another concern. A continent-wide framework can expand competition by admitting many qualified providers, but scale can also favour firms with large balance sheets and cross-border reach. The procurement should preserve routes for national and regional suppliers, transparent subcontracting and interoperability. Schools should not become captive to proprietary equipment or management platforms that make future competition costly.

CONNECTIVITY MUST REACH TEACHING PRACTICE

Educational use depends on teachers, content and classroom routines

The World Bank’s Digital Pathways for Education framework argues that digital investments need enabling conditions, human capacity and safeguards, with teachers central to design and implementation.[3] It also notes that evidence on the effectiveness of many education technologies remains limited. This is a useful restraint: connectivity is an enabling layer, not a learning outcome.

A 2026 UNICEF blended-learning pilot in Somaliland shows both readiness and constraint. The pilot began in five schools with 15 teachers and was being expanded to 56 schools. In a baseline survey of 178 teachers and school leaders, 98% believed digital tools could improve teaching and learning, yet more than half had never used digital tools in their work despite an average of more than 13 years’ experience. Only 7% of surveyed teachers and leaders were women.[9]

The pilot combined a tablet, a solar-powered projector, teacher support and grouped teaching at the learner’s level.[9] That package is important. The technology was placed inside a pedagogical method rather than offered as a self-executing intervention. It also surfaces gender and inclusion questions: who is trained, who controls equipment, which languages and learners are represented, and whether girls, learners with disabilities and displaced children gain comparable access.

For procurement, the implication is not that a connectivity supplier should own teacher development. It is that the connectivity programme needs an explicit interface with the education programme. School activation should trigger a readiness and use plan: designated staff, training and support, curriculum-aligned resources, safe-use procedures, device management and a small set of educational-use measures.

Those measures should avoid crude surveillance or vanity counts. Time online is not learning. More defensible indicators include teacher participation in supported practice, reliability during scheduled lessons, access to approved resources, administrative processes successfully completed, and learning measures designed independently of the supplier. Qualitative evidence from teachers and learners should sit alongside network telemetry.

COUNTRY SYSTEMS WILL DETERMINE WHETHER SCALE LASTS

A continental framework still lands in national budgets and institutions

Africa-wide procurement can create common terms and a qualified supplier pool, but implementation remains national and local. Ministries of education own school policy. ICT ministries and regulators shape spectrum, licensing, quality and infrastructure sharing. Universal-service funds may subsidise underserved areas. Finance ministries determine recurrent affordability. Utilities and energy authorities shape power access. Local governments or school boards may maintain buildings and equipment.

Giga’s January 2026 Government Exchange Programme brought officials from Ghana, Kenya, Malawi, Uganda and Zambia together across education, ICT, finance, regulation, research networks and universal-service funds to develop country roadmaps.[10] That multi-agency composition reflects the real delivery system. No single ministry controls all the conditions of school connectivity.

The most important country decision may be the recurrent funding route. External finance can map schools, install equipment or subsidise early service. Giga’s financing work includes public budgets, universal-service mechanisms, development-finance loans, blended structures and performance-linked instruments, but each route still needs a durable payer.[13] If the national budget, universal-service mechanism or another reliable revenue source does not assume the continuing cost, successful connections can expire. Procurement should therefore require a total-cost view covering service, power, maintenance, monitoring, replacement and contract management over the expected term.

Countries should also decide who can use school connectivity after hours. Community access or local resale can improve utilisation and sometimes revenue, particularly in underserved areas. It can also introduce security, safeguarding, congestion and governance risks. The choice should be explicit, locally governed and technically separated from protected school systems.

STONECOMMS ORIGINAL SYNTHESIS

The Connected-School Accountability Contract

The evidence supports a five-test model for evaluating whether a connectivity programme has produced a durable education service.

1. Reach: connect the right institution with the right technology. Every target school needs a verified location, enrolment and service baseline. Technology should reflect distance to infrastructure, terrain, expected use, expansion needs and whole-life cost. Progress is a school receiving service, not a cable passing nearby.

2. Readiness: make the connection usable and safe. Electricity, backup, on-site networks, device access, secure configuration, physical protection and safeguarding must be in place. Where any component is delivered separately, a named integrator should remain responsible for end-to-end acceptance.

3. Reliability: pay for observed performance. Speed, uptime, latency where relevant and repair time should be measured consistently. Contracts need transparent thresholds, remedies, audit rights and protection against manipulated or unrepresentative tests. Public reporting should identify persistent geographic or provider-level gaps.

4. Use: connect infrastructure to teaching and administration. Teachers and school leaders need supported practice, locally relevant content, safe-use procedures and help when systems fail. Measures should capture meaningful use without confusing screen time with education quality.

5. Renewal: finance and govern the service through its lifecycle. A recurrent funding source, maintenance arrangement, replacement schedule, interoperability plan and future competition route should exist before the initial subsidy ends. Suppliers should provide portable data and configurations so renewal does not become lock-in.

The five tests operate as a chain. Strong reach with weak readiness produces idle infrastructure. Readiness without reliability produces intermittent service. Reliability without use produces a functioning network with little educational value. Use without renewal produces a temporary success.

Confidence in the model is moderate. Procurement and programme evidence strongly supports the dependencies, while public evidence connecting specific service designs to comparable learning outcomes remains limited. The model should be tested through longitudinal school-level data and independent evaluation rather than adopted as a scorecard without validation.

A PAN-AFRICAN SCALABILITY TEST

One framework needs several delivery archetypes

A credible 54-country approach should distinguish at least four site archetypes.

Fibre-reachable urban and peri-urban schools can often use competitive fixed service, provided last-mile construction, wayleaves and in-school distribution are resolved. Here the priority may be competition, capacity and reliable repair rather than technology subsidy.

Mobile-covered rural schools in a usage gap may require business tariffs, external antennas, power, pooled demand and performance enforcement. The network exists nearby, but ordinary retail products may not offer the capacity, affordability or support a school needs.

Remote schools outside terrestrial coverage may require satellite or new wireless backhaul, solar power, resilient equipment and different repair logistics. Higher costs should be visible rather than hidden through a lower specification. Portfolio subsidy can make inclusion explicit.

Fragile, conflict-affected, refugee-serving or highly mobile settings need service designs that can tolerate disruption, protect sensitive data and support continuity. Local maintenance, portable equipment and offline functionality may be as important as maximum speed.

Performance comparison should therefore control for archetype. Publishing one continental average could conceal systematic under-service in remote schools. A useful dashboard would report reach, readiness, reliability, use and renewal separately by country, provider, technology, rurality and inclusion category.

IMPLICATIONS FOR DECISION-MAKERS

Governments should procure an accountable service

Education and ICT authorities should establish a verified school registry, baseline power and connectivity, and assign end-to-end responsibility. Tender evaluation should use total service cost and lifecycle performance, not equipment price alone. Recurrent funding and contract-management capacity should be approved with the award.

UNICEF, Giga and development financiers should protect the market-shaping function

The long-term agreements can reduce repeated procurement cost and set better norms. Their value will increase if bid and contract data are published in comparable form; savings claims state the baseline and specification; and underserved schools are not silently excluded. Concessional support should target genuine viability gaps and pay for verified performance.

Regulators and universal-service funds should use school data as market intelligence

School monitoring can show where coverage maps overstate usable service, where infrastructure sharing could lower cost and where persistent service failure requires enforcement. Universal-service finance should be linked to observable outcomes and coordinated with education budgets.

Suppliers should compete on reliability and adaptability

Providers that combine appropriate technology, local maintenance, transparent telemetry and interoperable equipment can create a stronger proposition than firms selling bandwidth alone. National firms and community networks may be particularly valuable in fault response and local conditions; contracting structures should make their accountability visible.

Education leaders should retain ownership of educational purpose

Connectivity procurement should not determine pedagogy by default. Ministries, teachers and learners need authority over content, use, safeguards and evaluation. Network measures should inform education management without allowing the supplier’s dashboard to become the definition of success.

RISKS, COUNTERARGUMENTS AND LIMITATIONS

Scale can amplify weak design as easily as strong design

A framework agreement is not a contract award, a funded rollout or a connected school. The September 2026 invitation establishes a procurement route; country participation, budgets, call-off contracts, supplier performance and implementation remain future decisions.[1] The potential facility counts should not be reported as commitments.

Cost-reduction examples from Botswana, Kenya and Rwanda are relevant but not directly comparable. They reflect different baselines, contract structures, infrastructure access and reported time periods. Independent publication of tender prices, full service specifications and realised performance would improve confidence.

The strongest evidence reviewed concerns mapping, procurement, connectivity prices and implementation capability. Evidence linking particular connectivity models to learning gains is thinner and context-dependent. Schools with stronger leadership, electricity and resources may both adopt technology more successfully and produce better outcomes, making attribution difficult.

There are material risks of vendor lock-in, market concentration, weak cybersecurity, inappropriate content, child-data misuse, equipment theft, e-waste and unequal access within schools. Solar components can fail without maintenance; monitoring devices can misreport; remote sites can face slower repairs. Open performance data needs privacy and security controls.

This paper uses public evidence available through 24 September 2026. Institution-reported programme results were retained with attribution and not treated as independent impact evaluations. It does not compare bids, endorse a supplier or estimate the full cost of connecting every African school. The health-facility component of the UNICEF procurement is acknowledged but not analysed in depth because the paper’s substantive focus is digital education.

COMMISSIONABLE RESEARCH AGENDA

Follow the service from tender to classroom

A next-stage commission should create a multi-country Connected-School Accountability Observatory. It would combine facility mapping, procurement records, network telemetry, power availability, repair logs, teacher-use evidence, budget flows and independent education measures for a representative panel of schools.

The design should compare the four delivery archetypes and at least four contracting models: individual school contracts, national aggregation, regional or cross-border frameworks, and performance-linked subsidy. It should record total lifecycle cost rather than price per megabit alone, including solar maintenance, local-network equipment, monitoring, training and replacement.

Fieldwork should centre school leaders, teachers and learners, with deliberate inclusion of women educators, rural communities, refugees and learners with disabilities. Supplier and ministry interviews should identify where responsibility fragments during faults. Cybersecurity and child-protection audits should test whether written safeguards operate in practice.

The central evaluative question is causal and practical: which combinations of reach, readiness, reliability, use and renewal produce sustained educational value at an affordable public cost? Answering it would turn an ambitious procurement into reusable institutional knowledge.

CONCLUSION

The contract must keep working after the connection ceremony

UNICEF’s continent-wide procurement is important because it treats school connectivity as a service that can be specified, aggregated and monitored. It can lower transaction costs, expand supplier competition and make solar power and local networks part of the same delivery conversation. It can also create common performance evidence across very different markets.

Its success should not be judged by the size of the framework or the number of facilities eligible to use it. The meaningful unit is a school that can access a safe, powered and reliable service; teachers who can use it; learners who are not excluded by geography, gender, disability or displacement; and a public institution able to fund and govern the service over time.

The practical opportunity is to make every call-off contract pass five tests: Reach, Readiness, Reliability, Use and Renewal. That standard is more demanding than counting installations. It is also more likely to turn connectivity spending into lasting education infrastructure.