A licence doesn't launch a betting business. It's only the beginning.

1. Licensing is not launching

The pattern repeats across the continent: an operator secures approval, then discovers the licence was the easy part. Industry coverage of African market entry keeps returning to the same three stall points. First, platform localisation — European-built platforms routinely underperform on African usability and bandwidth realities (iGB, "What works in Africa: the features global platforms overlook"). Second, payments — each market runs its own rails (M-Pesa in Kenya, agent networks and bank transfer in Nigeria, MoMo in Ghana and Uganda), and every wallet is its own integration project. Third, regulatory volatility — the rules move while you build: Kenya froze applications in October 2025; Nigeria's Supreme Court re-drew the entire licensing map in November 2024; Uganda's president vetoed part of the 2026 tax bill in July 2026. Track360's 2026 operator guide advises treating Nigerian licensing as "a moving target" and tax as "a live political variable" — advice that currently applies to most of the region.

The launch partner's job is to make those three stall points somebody's day job. Ours.

2. The regulatory map — as of July 2026

Every figure sourced; conversions indicative at mid-2026 rates. This table is maintained — the detail sections below carry the citations.

3. Kenya: the freeze, and what to do with it

Kenya is mid-transition. The Gambling Control Act 2025 (in force August 2025) replaced the BCLB with the Gambling Regulatory Authority (GRA), which formally took over in February 2026. The BCLB had already halted all new licence applications and renewals on 2 October 2025 pending the handover. Then the new Gambling Control (Licensing) Regulations 2026 — which set an online bookmaker/casino fee of KES 50 million (~$385k) and raised some renewal fees by 200% to 49,900%, plus a 6% approval fee on advertising budgets — were suspended by the High Court on 20 July 2026 in a constitutional challenge, with a mention date of 21 September 2026 (FocusGN; iGB).

Under the Act: 30% Kenyan shareholding applies to land-based operations but not to online-only foreign operators (Bowmans); online gambling requires a KES 100M bank guarantee; minimum capital for hybrid online licences is KES 50M. The 2025–26 tax stack: 15% on GGR, excise cut from 15% to 5% — but re-based from stakes to deposits, and withholding restructured to 5% on wallet withdrawals (a 20% rate is proposed in the draft 2026 Finance Bill). Despite lower headline rates, the deposit/withdrawal base is projected to more than double total take (Yogonet, Oct 2025).

4. Nigeria: fifty markets pretending to be one

The November 2024 Supreme Court judgment voided the National Lottery Act and made gambling regulation an exclusively state matter — the NLRC's remit shrank to the Federal Capital Territory. The practical routes in 2026: license per state, or use the Federation of State Gaming Regulators of Nigeria (FSGRN) Universal Reciprocity Certificate — one application covering ~22 member states at ₦100M (~$65k) per year per category (framework signed 7 May 2025; fees payable from January 2026).

Lagos — the anchor market — licenses online sports betting at ₦150M plus a ₦500k application fee with ₦100M annual renewals (Mondaq 2026; older sources cite ₦50–100M — figures have moved), a 2.5% levy on sales revenue, and since February 2026 a 5% withholding tax on net winnings paid out by Lagos-licensed platforms. Nationally, the FSGRN framework sets a flat 11% GGR tax from 1 January 2026, and the Nigeria Tax Act 2025 exempts stakes from VAT while applying full corporate income tax to operator profits (KPMG Nigeria). Lagos's published clock is among the region's fastest: ~4–6 weeks after complete documentation, 10–12 weeks end-to-end (Koriat Law).

5. Ghana and Uganda: opposite directions on the player, same lesson for operators

Ghana runs the region's most structurally partner-driven regime: Class 2 (sports betting) licences at roughly $40k application + $50k annual (advisory-firm figures), minimum stated capital of US$2 million, and mandatory Ghanaian ownership — a JV needs a local partner at ≥10% equity with the foreign investor injecting at least US$200,000. Operator tax is 20% of GGR. The headline 2025 move went the player's way: the 10% withholding tax on winnings was abolished effective 2 April 2025 — a demand-side stimulus in the region's fourth-largest online market ($883M gross win 2025, H2GC).

Uganda went the other way. From 1 July 2026: a harmonised 30% GGR tax (up from 20% for betting) plus 15% withholding on net winnings — though President Museveni returned the Income Tax (Amendment) Bill in July 2026 over a casino-exemption clause, leaving the WHT element back before Parliament. The 15% excise on stakes continues. Entry costs remain modest (application UGX 20M residents / UGX 50M foreigners; minimum capital UGX 250M; UGX 500M bank guarantee), but the operating math is now among the region's tightest — which punishes heavy cost structures and rewards lean channels. This is where USSD-led distribution, with no app-store tax and operator-controlled session costs, earns its keep.

6. Tanzania, Zambia, DRC: the fast, the cheap, and the deadline

Tanzania is the region's fastest, cleanest entry: Internet Sports Betting Licence at TZS 90M (~$34k) with a TZS 1M application fee, a Gaming Board that aims to decide within 30 days of a complete file, and 1–3 months typical end-to-end (Gofaizen & Sherle, 2026). The 2026/27 budget added a 5% excise on all betting stakes on top of 25% GGR and ~12% winnings withholding — model it before you price bonuses.

Zambia pairs one of the region's cheapest licences (online sports betting around ZMW 20,000 ≈ $750, per secondary sources — verify institutional detail before filing) with one of its heaviest tax stacks: 10% excise on stakes (September 2025), then the Betting Levy Act 2025 adding 5% on deposits and 5% on withdrawals from January 2026, plus 15% WHT on winnings. Cheap to enter, expensive to run — unit economics decide winners here, not licence budgets.

DRC is a ~$1.7B/yr market in regulatory transition: licensing via the Ministry of Finance with state lottery SONAL as gatekeeper, an interim supervisor (CSJA) created January 2026, and a permanent regulator (ARJA) pending in Parliament. Live compliance reality: all licensed operators must complete CSJA electronic registration by 31 July 2026, with the digital-platform phase closing 30 August 2026 (Yogonet, Jul 2026). High risk, thin enforcement, first-mover economics — a market you enter with local counsel and a compliance calendar, or not at all.

7. What launching actually costs

Honest framing: there is no rigorous public dataset on African sportsbook launch costs — what exists is consultancy estimates. Published ranges: white-label platforms from the tens of thousands of dollars; custom builds $150k–600k+; all-in market entry estimated at $200k–800k for Kenya (6–12 months) and $300k–1M for Nigeria (9–15 months) (SourceCodeLab, 2026 — single source, unverified methodology; the same firm reports ~9.5 months average planning-to-launch across its 8+ operator launches).

What we change in that equation: the platform, USSD stack, telco integrations and mobile-money rails are running infrastructure, not a build project. Your budget concentrates on the licence, the market, and the launch — the parts only you can own.

8. What the partnership covers

9. Licence-holder FAQ

Can I get a Kenyan licence right now? Effectively no — applications frozen since October 2025, new fee regulations court-suspended July 2026 (mention: 21 September). Existing licensees run on current terms. Use the window to prepare the file (see §3).

Fastest market in? Tanzania: 30-day decision target, 1–3 months typical, ~$34k licence. Lagos: ~10–12 weeks. Ghana: 4–8 months.

Do I need local shareholders? Ghana: yes (≥10% local equity in a JV, US$2M stated capital). Kenya: 30% local — land-based only; online-only foreign operators exempt. Uganda/Tanzania/Zambia: no published quota.

I already have a platform — do I need this page? Then you need the USSD channel, not a rebuild. If you're live and growing, see Scale.