Songbird Business Plan

Africa has the steepest cost-of-connectivity curve in the world. A school in rural Niger pays five to ten times what a school in Marseille pays for the same megabit, and the marginal cost of that megabit hasn't changed materially in five years despite a wave of new subsea cables. The gap is not bandwidth — it is the wholesale layer that turns subsea bandwidth into school-priced retail capacity. Songbird builds that layer.

Songbird is a wholesale first-mile telecom infrastructure venture, originated within UNICEF/Giga and now structured as an independent commercial entity owned by a Swiss Foundation. It operates 20 Points of Presence across 11 African and Mediterranean countries, terminating subsea capacity from cables including 2Africa, Equiano, Medusa, and EASSy, and selling aggregated capacity at substantially below-market prices into Giga-mediated education procurement.

Core thesis in five points

• ▸Songbird is wholesale infrastructure, not last-mile retail. It does not compete with telcos; it gives them — and Giga — better wholesale economics on the routes they need most.
• ▸The Foundation governance and Giga firewall are the moat. They guarantee discounted school pricing cannot be arbitraged into competitor markets — which is what makes donors willing to fund the build.
• ▸20 POPs is the long-term shape, not the launch shape. A 3-POP Phase 1 (Kenya + Uganda + Lagos) costs $8–10M peak funding to prove the model.
• ▸Mission-aligned revenue alone is insufficient. It must be combined with commercial revenue (hyperscalers, CDN peering, banks, sovereign data) — with strict capacity-firewalling so school routes are never displaced.
• ▸Public capital (EU, AECID, AFD, EIB Global, EFSD+) does most of the heavy lifting. Patient private capital (Convergence Partners, EAIF, Proparco) provides the equity layer enabling commercial flexibility.

2.1 The first-mile gap

Subsea cable capacity into Africa has multiplied 10x since 2019 — Equiano, 2Africa, Medusa, and EASSy upgrades have collectively added >200 Tbps of theoretical capacity. Yet retail prices have barely moved. The bottleneck is the wholesale aggregation layer between cable landing and last-mile distributor, and the absence of a price-disciplined buyer of school connectivity.

2.2 The addressable market

• Schools (Giga TAM)
  ~700,000 schools across SSA + Maghreb mapped by Giga
• Clinics & hospitals
  ~30,000 facilities (WHO / Africa CDC)
• Public universities
  ~1,200 institutions
• Hyperscaler peering
  AWS, Google, Meta, Microsoft seek diverse African routes
• EU strategic premium
  Trusted non-Chinese routes carry political premium post-2022
• CDN & content
  Akamai, Cloudflare, Netflix want POP density

Songbird is structured as three deliberately separated layers, each with its own governance and pricing logic.

How the firewall works

• ▸Constitutional Foundation clause prohibits diverting school-priced capacity to commercial use.
• ▸A capacity reservation system at each POP guarantees a minimum percentage of throughput is allocated to verified Giga traffic at all times.
• ▸Independent annual audit by a third party (Big-4 or telecom-specialist) verifies allocation compliance.
• ▸Commercial customers contract with the wholesale entity; Giga contracts with the Foundation directly.

Geographic rollout is phased: Phase 1 launches with 3 POPs (Mombasa, Kampala, Lagos) at $8–10M peak funding to prove the wholesale model and unit economics. Phases 2–4 expand to the full 20 POPs across 11 countries by 2029, conditional on Phase 1 validation.

All ~25 model inputs — equipment capex, IRU vs lease share, school price, penetration ramp, IPT decline, staff and overhead, working capital, FX, tax, exit multiple — are documented in the Super Report and live in the Simulator. The five inputs driving 80% of outcome variance are: Giga discount, commercial revenue mix, IRU vs lease share, schools penetration ramp, and IPT price decline.

The plan identifies 20 direct revenue channels across four families:

Combined with Tier-1 ideas active in the simulator, payback shifts from Never to ~5 years; 2029 revenue rises from $2.24M to ~$5–7M.

Incumbents on the routes Songbird covers — Liquid Intelligent Tech, Bayobab/MTN GlobalConnect, Seacom, WIOCC, Orange — are all conflicted between wholesale and retail. None can honestly price below their own retail tariffs. None has access to the public capital pool that Foundation governance unlocks.

• ▸Foundation governance + Giga firewall: unique, no incumbent has this. Unlocks public capital that incumbents cannot access.
• ▸Non-retail conflicted: MTN and Orange cannot honestly price below their own retail tariffs. Songbird can — because it has no retail business.
• ▸Maghreb-EU corridor: Songbird is one of the few wholesale operators positioned on Medusa from day one. Liquid/Bayobab/MTN are weaker on this axis.

Total raise: $25–28M over 18 months. Phase 1 minimum: $8–10M. See the Fundraising Strategy for the full two-track plan.

1. M0–3Foundation entity formed; AECID/FEDES proposal submitted; Wilson Sonsini structuring memo finalised
2. M3–6Phase 1 site surveys (Mombasa, Kampala, Lagos); FEDES letter of interest; EIB Global pre-screening
3. M6–9First IRU contracts signed; investor deck circulated; Convergence indicative term sheet
4. M9–12Phase 1 POP equipment installed; AECID disbursement; Convergence-led Series A close
5. M12–15First commercial customer live; EIB Global drawdown; Nordic + Italian closes
6. M15–18Phase 1 + Phase 2 POPs operational (12 of 20); EAIF senior debt term sheet; Phase 2 funding open