Your platform works. It just can't reach most of the market. Add the channel that can.

1. The market your app cannot see

The coverage problem in Africa is solved; the usage problem is not. The GSMA puts Sub-Saharan Africa's mobile-internet coverage gap at 13% of the population — but the usage gap at 60%: people under a broadband signal who still don't use the internet (GSMA, State of Mobile Internet Connectivity). Its Mobile Economy Africa 2026 report sharpens the point: 63% of people covered by mobile broadband in SSA do not use it — roughly 960 million Africans covered but not connected.

Hardware is one reason. Feature phones accounted for 55% of all mobile phone shipments in Africa in Q3 2024 (IDC), overwhelmingly Transsion devices — Tecno, Itel, Infinix — that cannot run an operator's iOS or Android app. Kenya's regulator counts the installed base directly: as of March 2026 the market held 50M+ smartphones but still 28.5 million feature phones — 36% of all devices on the network (Communications Authority of Kenya, Q3 FY2025/26 sector statistics). Smartphone adoption is rising — 54% of SSA connections in 2024, forecast 81% by 2030 (GSMA) — but an entry-level 4G smartphone still costs about 26% of monthly GDP per capita in the region. The transition is a decade, not a quarter.

2. Even smartphone owners hit the data wall

The second reason is running cost. The World Bank's Global Findex 2025 measured mobile data in SSA at 2.4% of monthly income for 1GB — and 5% for the poorest 40%. The ITU's median for an entry-level 2GB plan in Africa is 4.2% of GNI per capita, the highest of any region and more than double the UN's 2% affordability target. A bettor with a smartphone and an empty data bundle is, functionally, a feature-phone user until the next top-up.

USSD sessions ride the signalling layer: zero data consumed, on any phone, on any network generation. In Kenya the market norm is operator-paid sessions — SportPesa's *790# is free to the user on both Safaricom and Airtel. The channel's running cost to the bettor is nothing, which is precisely why it converts where app funnels stall.

3. What the operators who dominate actually do

The strongest evidence for USSD is where market leaders keep investing. Betika has run *644# alongside its app since 2016 and built Kenya's largest user base. SportPesa launched *790# in November 2016 explicitly "for customers without internet access" — and in November 2025 launched an entirely new USSD casino product on the same shortcode, marketed as data-free gaming for "cities, towns, and remote areas" (Biznakenya, Nov 2025). Operators do not ship new products on dying channels.

4. The 2G floor isn't going anywhere

The "USSD is legacy" argument collapses on the network data. Africa's connection mix at end-2024: roughly half 3G, a third 4G, ~10% still 2G, ~1% 5G (GSMA). The GSMA expects African network sunsetting to be "gradual and more selective" than elsewhere, moving in earnest only from 2030 onwards. South Africa is the single announced full shutdown (December 31, 2027). Kenya, Nigeria, Tanzania, Uganda and Ghana have no announced 2G sunset dates.

Kenya's March 2026 regulator statistics still count 9.7 million 2G and 5.0 million 3G subscriptions — 14.7 million legacy connections in one market (CA Kenya). A betting channel that requires only signalling survives every migration between them.

5. The money is already mobile

Deposits are where African betting funnels live or die, and the rails are mobile money, not cards. Sub-Saharan Africa processed $1.4 trillion in mobile-money transactions in 2025 — 66% of the global total (GSMA State of the Industry 2026). The operator numbers behind that: M-Pesa Kenya at 41.0 million monthly-active customers and KSh 41.7 trillion in value (Safaricom FY2026 results); MTN MoMo at 69.5 million actives and $500.3 billion processed (MTN FY2025); Airtel Money at ~47 million customers and a $245 billion annualised run-rate (Airtel Africa, Jul 2026).

Our USSD flows put the deposit inside the menu: the bettor selects Deposit, and the M-Pesa or MoMo PIN prompt appears on the same phone, in the same session. No redirect, no browser, no data.

6. Shortcode economics, market by market

The commercial layer operators rarely see priced in one place. Representative published figures:

Sources: Safaricom USSD Tariff Guide; Uganda Communications Commission shortcode schedule; Ghana NCA Form AP19; aggregator rate cards (indicative, 2024–2026). Nigeria's end-user billing regime followed the 2025–2026 resolution of the banks–telcos USSD debt dispute.

A structural note on Nigeria: USSD banking volumes there are declining as smartphone users migrate to apps (₦2.19 trillion in H1 2024, down 54.75% year-on-year — NIBSS). That is the channel maturing into its real role: the reach layer for the unconnected majority, not the whole funnel. Betting operators should model USSD in Nigeria as incremental reach, priced at end-user billing rates.

7. What we actually integrate

8. The channel split nobody publishes — except to you

Section 3's honest number cuts both ways: because no operator publishes channel data, most operators don't have it. Every USSD session on our platform emits structured events — dials, menu depth, bet placement, deposit completion, session duration — rolled up in real time and split by shortcode, market and network. You see USSD's contribution next to your app and web funnels, per day, from launch week. When the board asks what the channel is worth, you answer with your own data.

9. Operator FAQ

How much betting volume goes through USSD? Nobody publishes it (see §3). The documented context: 94% of African bettors bet on mobile (GeoPoll 2025), 63% of the covered SSA population is offline (GSMA 2026), and market leaders shipped new USSD products as recently as November 2025. Run the channel; measure your own split.

How fast can we be live? Integration is weeks — the telco stack already runs. Shortcode acquisition sets the calendar: shared codes can be same-week (Kenya via aggregator; Uganda and Zambia immediate), dedicated codes typically 4–10 weeks depending on market.

Does USSD cannibalise our app? The populations barely overlap: the channel's core users hold feature phones or run empty data bundles. Where overlap exists, the dialer shortcode is retention infrastructure — it can't be uninstalled for storage.

Is the channel dying with 2G? No sunset is announced in Kenya, Nigeria, Tanzania, Uganda or Ghana; the GSMA projects African sunsetting from ~2030, and USSD itself is network-generation-independent signalling. South Africa (Dec 2027) is the exception, not the pattern.