Funding for telecoms and connectivity businesses in South Africa

Why connectivity raises on better terms than almost anything

A subscriber base paying monthly with low churn is annuity revenue, and lenders treat it that way: debt sized against ARPU, churn and penetration rather than hard assets alone — leverage most sectors cannot access at any price. South Africa’s open-access fibre model sharpens the opportunity: ISPs ride national networks without owning them, shifting the raise from capex-heavy infrastructure to customer-acquisition economics, fundable with far lighter structures. The sector’s capital question is rarely whether money exists; it is whether the subscriber data survives the modelling.

The four numbers that decide diligence

Churn first — it is the number that turns subscriber growth into value or vanity, and funders model it monthly by cohort, not blended. Then ARPU trajectory against price competition in the node; uptake rates in passed premises for network builds, because homes passed is a cost and homes connected is a business; and enterprise contract quality for B2B-focused providers — tenor, SLAs, concentration. A provider with honest cohort churn and defensible nodes raises debt comfortably; one with blended churn raises expensively or not at all.

The consolidation reality

Sub-scale ISPs are being acquired for their subscriber bases in an actively consolidating market — which cuts two ways. For operators with capital access, acquisition growth funded with debt against combined subscriber economics is the fastest route to scale. For sub-scale operators facing a hard raise, a sale process is a genuine alternative: subscriber bases command real multiples from acquirers building density, and the honest comparison is the raise’s dilution against the acquisition offer, run side by side. International infrastructure investors are active across African connectivity at both ends of that trade.

The capital map and what kills telecoms raises

Debt sized against subscriber economics for network and acquisition capex; growth equity for footprint expansion; infrastructure capital for owned-network builds with uptake proof; and mezzanine where senior leverage maxes before the plan does. What kills raises: blended churn concealing cohort decay; homes-passed pitched as achievement without uptake; ARPU forecasts ignoring node-level price wars; enterprise concentration in one anchor; and infrastructure money requested for what is a customer-acquisition business. Start with the readiness check or growth funding.

A worked example: what subscriber economics fund

An ISP with 8,000 subscribers at R550 ARPU and 1.2% monthly cohort churn is generating roughly R4.4m monthly of annuity-quality revenue — a base against which lenders will typically size meaningful acquisition or network capex, covenanted on churn and ARPU floors. The same revenue with 3% blended churn supports a fraction of the leverage at worse pricing, because the book replaces itself every three years instead of seven. That spread — identical topline, different durability — is why cohort-level churn data is the single most valuable document a connectivity business owns.

The questions connectivity funders will ask

• Show me churn monthly, by cohort and by node — not blended.
• What is ARPU trajectory where competitors have entered your nodes?
• For builds: uptake rate in premises passed, at 6, 12 and 24 months?
• What drives your churn — debit-order failures, installer quality, price — and what is being done about each?
• What is enterprise concentration and contract tenor?
• What are the wholesale network terms you ride on, and their escalation clauses?
• What does subscriber acquisition truly cost, fully loaded, by channel?

Cohort churn with causes attached is the exhibit that moves terms — it proves management sees the book as it is.

Node economics: where connectivity is actually won

Blended metrics hide the truth that connectivity is a street-by-street business: a node with 40% penetration, sticky debit orders and one competitor prices differently from one at 12% with three rivals discounting. Funders increasingly model node-level unit economics — penetration, ARPU, churn causes, cost-to-connect — and providers who bring that granularity control the conversation. Churn causes deserve equal granularity: debit-order failure churn is a collections problem with collections solutions; installer-quality churn is an operations problem; price churn is a node-selection problem. Treating them as one number leads to one (wrong) answer; separating them is usually worth several points of monthly churn — and several turns of leverage.

The wholesale layer deserves its own scrutiny: margin per line after network costs, the escalation clauses in wholesale agreements, and what happens to unit economics if the network operator raises pricing at renewal. Providers who negotiate multi-year wholesale terms with capped escalations — and can show the margin bridge per product — remove the quiet risk that erodes every forecast built on today’s input prices.

Last, protect the covenant headroom: subscriber-economics debt comes with churn and ARPU covenants, and the operators who suffer are those who leverage to the maximum the base supports today. Sizing debt against a stressed case — competitor entry in the best nodes, a point of extra churn — keeps the facility an asset rather than a trap.