DRC has 60 days to formalize a $1.7B betting market that paid $1M in tax.

1. The $1.7 billion vs $1 million paradox

Two numbers anchor every DRC iGaming pitch deck and they barely speak to each other. iGaming Afrika and adjacent coverage put DRC sports-betting handle (total turnover wagered) at roughly $1.7 billion a year. Historical state tax capture from gaming is closer to $1 million annually. That is roughly 0.06% of reported turnover landing on the public ledger.

Two structural issues explain the gap. The first is taxonomy: $1.7 billion is widely cited as turnover, but parts of the press conflate it with gross gaming revenue. The two figures can differ by a factor of seven or more depending on retained margin. Our underlying research flags this as the single weakest data point in the DRC stack. The second is enforcement: in 2022 the government counted around 139 illegal operators, against roughly 16 formal licensed platforms today. The grey market is not a rounding error. It is the market.

Bankable Africa frames the 2025-2026 package as an explicit revenue mobilization exercise. EAGT monitoring, mandatory digital accounts and the 10% WHT are not three independent reforms; they are one closed loop designed to make every bet observable, every payout taxable and every operator auditable.

2. SONAL and the dual-regulator structure

Two state bodies sit between an operator and a legal bet. Societe Nationale de Loterie (SONAL) is the operational regulator: it issues licences for lotteries, sports betting, casinos and digital gaming including SMS and USSD, and supervises game integrity and consumer protection. Direction Generale des Impots (DGI), under the Ministere des Finances, owns tax collection and fiscal policy. Operators must register, file and remit through both. SONAL signs the licence; DGI signs off the tax position.

Licence pricing is steep relative to nearby Francophone markets and structured to discourage casual entry:

A national USSD-plus-online sportsbook in DRC starts with $100,000 in licence fees alone, before any technology spend or carrier setup. Compared to Ghana, where initial Gaming Commission fees can reach the half-million-dollar range for some classes, DRC headline pricing looks lighter; but the $50,000 SMS-and-USSD line is unusual. Most Anglophone regulators bundle channels under a single sportsbook licence; SONAL prices each rail separately. Operators that already paid for online but want USSD reach now have a second cheque to write before March 31.

3. The March 31 2026 compliance deadline

The headline date is fixed. Per mediacongo and Bankable Africa coverage of the Ministere des Finances directives, every betting, lottery and casino operator in the DRC must by March 31 2026 have:

• Settled outstanding tax obligations with DGI
• Renewed or obtained the appropriate SONAL agrement (licence) for each channel
• Migrated every active player to a verifiable digital account
• Connected operational systems to the EAGT central monitoring API
• Demonstrated 10% withholding-tax deduction on payouts

The penalty stack is severe: licence suspension, financial penalties and criminal prosecution under the partnership framework. FocusGN reporting on the digital accounts mandate confirms that anonymous USSD or retail betting, which historically dominated upcountry play, is being explicitly banned. The legal posture is no longer "encourage formalization." It is "anything not on the digital account is unlicensed."

Even allowing for the usual gap between announced deadline and actual enforcement, the direction is clear: an operator without an EAGT-integrated stack on April 1 2026 is operating without political cover.

4. The 10% withholding tax on player winnings

The most consequential fiscal change is the Taxe ad valorem sur les gains, a 10% withholding tax applied directly to player winnings rather than to operator GGR. Before any payout reaches a player's mobile-money wallet, the licensed operator must deduct 10% of the gross win and remit it to DGI. The operator becomes a tax-collection agent embedded inside the betting flow.

This is not the GGR tax structure most international operators are used to. Two implications:

1. Cash-flow impact is on the player, not the operator's P&L. Headline tax burden is technically borne by the bettor; the operator's role is collection and remittance.
2. Tax repulsion risk is severe. The regulated cohort must visibly deduct 10% of every payout while grey-market operators continue paying winnings gross. Punters do the maths: a 10% reduction in realized yield on every winning ticket is a powerful incentive to drift back to unregulated platforms.

Ghana's policy makers ran the opposite experiment by abolishing the 10% WHT on betting winnings to shrink its grey market. DRC is taking the inverse bet: EAGT monitoring plus payment-rail control will keep enough of the market on-platform that withholding becomes self-enforcing. Whether that holds depends on how cleanly EAGT integrates with mobile money payouts and how aggressively SONAL pursues the unregulated tail.

5. EAGT central monitoring system

The enforcement layer of the 2026 reform is technical, not bureaucratic. In mid-2025 the DRC government signed a public-private partnership with the East African General Trade Company (EAGT), a firm originating from Burundi, to deploy a centralised digital monitoring system connecting in real time to the back-end of every licensed operator.

EAGT becomes the regulator's eyes and the tax authority's calculator. Per FocusGN and Bankable Africa coverage of the partnership, the platform sees every transaction the licensed market produces. Practical scope:

• Stakes: bet amount, currency, sport, market, timestamp
• Odds: price taken at the moment the bet is placed
• Settlement: outcome of the wager, gross win figure
• Tax validation: confirmation that 10% WHT has been deducted from any winning payout before it leaves the operator's wallet
• Player identity: the digital account ID associated with each transaction

The mechanism kills two operator workarounds at once. First, anonymous USSD play disappears: every action now has a verified digital account behind it. Second, off-book settlements become structurally harder, because the regulator sees the bet at placement time, not via post-hoc operator filings.

6. The USSD friction risk: API latency in 120-second sessions

EAGT integration is easy in a web stack with multi-second response budgets. It is operationally brutal inside USSD. DRC USSD sessions cap at 120 to 180 seconds. Application timeouts drop a session if the user takes 30 to 60 seconds to respond, or if the operator's server takes longer than 10 to 15 seconds to fetch live odds or, now, to round-trip a tax-validation call to EAGT.

The economic impact compounds with billing. DRC USSD is a flat-per-session model at roughly 97 CDF per dial via aggregator (~$0.04). Every dropped session forces a redial at another 97 CDF. A modest worst-case of 1.5 sessions per completed bet produces roughly 145 CDF in telco cost per placed bet. On an assumed gross profit of 600 CDF per bet (4,000 CDF stake at a 15% retained margin, both UNVERIFIED for DRC), that is roughly 24% of operator gross revenue consumed by USSD access alone.

The numbers above are illustrative, not audited. The point: any architecture that bolts a synchronous EAGT validation call onto the critical path of a USSD payout flow is subsidising the regulator with telco costs. Sensible operators will treat EAGT integration as an asynchronous queue with optimistic settlement and a circuit breaker for partial outages.

7. MNO landscape: Vodacom, Airtel, Orange, Africell

USSD reach in the DRC is gated by a four-carrier oligopoly. ARPTC Q3 2024 data distributes mobile subscriptions across Vodacom RDC, Airtel DRC, Orange RDC and Africell. The mobile-money revenue split inside Services Financiers Mobiles (SFM) is similarly concentrated. Together these splits define what national reach actually costs.

Two operators own the betting wallet. M-Pesa and Airtel Money together account for over 82% of mobile-money revenue, so most successful USSD-born bets clear through one of those two rails. Vodacom, Airtel and Orange between them control roughly 95% of subscriptions; a national shortcode that stops at Vodacom and Airtel still leaves nearly 30% of the country dark. A serious DRC USSD play needs at least three carriers integrated and at least two mobile-money APIs with high availability. Africell is deferrable; Orange is not.

8. The flat-per-session billing model: 97 CDF per dial

DRC USSD economics differ from East African benchmarks. Uganda, Kenya and South Africa typically run a time-sliced model, charging in 20-second increments. DRC aggregator pricing is a flat per-session fee at roughly 97 CDF (~$0.04) per dial on the postpaid side. The cost is the same whether the user is in-menu for 10 seconds or 115, provided the session does not time out.

This flips the optimal design pattern. In a time-sliced market the operator pushes users through menus as fast as possible. In DRC's flat-rate market the operator's job is to keep users in-session and complete the bet inside one dial. Deeper menus, league browsing and accumulator builders are economically viable as long as session timeout discipline holds.

Setup and ongoing costs:

• Dedicated USSD code setup: approximately 2,846,000 CDF (~$1,225) one-off
• Monthly maintenance per telco: approximately 570,000 CDF (~$245) per carrier per month
• Three-carrier presence: approximately $735 per month in fixed code-leasing alone
• Per-session usage: ~97 CDF, flagged as UNVERIFIED at the precise rate because direct-to-MNO commercial agreements are NDA-shielded and aggregator pricing varies

The data caveat matters. Public aggregator pricing is the figure operators see at evaluation. Direct-to-MNO contracts held by larger licence holders almost certainly differ, sometimes substantially. Anyone modelling DRC unit economics off public aggregator quotes is calibrating against a ceiling, not a clearing price.

9. Stripped ASCII French: the encoding hack every Francophone operator uses

French is the official commercial language of the DRC and proper French uses accented characters: e-acute, a-grave, e-circumflex, c-cedilla, o-circumflex. Standard USSD messages run on the 7-bit GSM alphabet, which gives operators 140 to 160 characters per page. The moment a single accented character lands in the payload, the network falls back to UCS-2 encoding, which slashes the per-page limit to 70 characters. Source coverage on this is consistent across IBM developer documentation and Medium technical write-ups on USSD versus SMS encoding.

Seventy characters is catastrophic for sportsbook menus. Listing team names, kickoff times, odds and a navigation prompt does not fit. What used to render on one screen now spans three. More screens means more user inputs, longer time-in-session and a higher probability of triggering the 120 to 180 second timeout, which means another 97 CDF redial.

The industry workaround is universal: strip the accents. Francophone operators in DRC, Cameroon and Cote d'Ivoire write "Evenements" instead of the accented form, "Resultats" instead of the accented form, "Cote" without the circumflex. Grammar suffers; encoding stays in 7-bit GSM and the 160-character budget survives. Any DRC USSD style guide should explicitly forbid accented characters in menu copy. ASCII-only French is not optional. It is the difference between one redial and three.

10. The operator landscape: Winner.bet leads, MBet pioneers USSD

The licensed cohort is small, brand-concentrated and shaped by Pan-African and Eastern European groups rather than indigenous DRC startups. Public sourcing across iGaming Afrika, Mighty Tips, the IJPSAT Kinshasa Winner.bet study and the World Lotteries newsroom paints a consistent picture.

MBet is the operator most relevant to the USSD conversation. World Lotteries reporting describes the largest African turnkey deal of its type with EveryMatrix supplying the platform. MBet has prior form transitioning physical retail into USSD-first acquisition in adjacent markets, which gives it an operational head start on EAGT integration.

With roughly 16 formal operators against an estimated 130-plus including informal players, the licensed market is structurally smaller than the active market. The 2026 reform compresses that gap by design.

11. The CDF dollarization reality

The Congolese Franc (CDF) has depreciated steadily against the US Dollar. As of April 2026 the mid-market rate hovered around 2,317 to 2,323 CDF per USD, with persistent downward pressure on the 12-month trailing window. In practice DRC operates as a heavily dollarized economy: real-estate listings, restaurant prices in Kinshasa and a long tail of B2B contracts are routinely USD-denominated.

That dual-currency reality bleeds into betting. Major operators run dual-currency wallets, list deposit thresholds in both CDF and USD and anchor marketing in dollar equivalents to neutralise CDF inflation in the punter's mental accounting. For international operators that means structural FX exposure: every CDF-denominated stake settled into a USD-denominated parent ledger crystallises a spread loss, and that spread widens when CDF accelerates its slide. Compared to Uganda's UGX, which moves inside a relatively narrow managed band, CDF requires both wallet duality and a buffered FX policy on the operator's books.

12. Why we don't yet ship a DRC USSD revenue calculator

ussdbet.africa publishes USSD revenue calculators for Tanzania, Uganda, Ghana and Zambia. We have not yet shipped one for the DRC. Three reasons, all rooted in data quality:

1. Handle vs GGR conflation. The widely cited $1.7 billion DRC iGaming figure circulates ambiguously across handle and GGR. Until SONAL and DGI publish audited filings post-EAGT integration, any top-down market-size input is noise.
2. Opaque margin assumptions. Operator retained margins of 10-15% are pan-African baselines, not DRC-verified figures. Grey-market operators offering improved odds may compress regulated margins below that band; we do not yet have post-2026 operator filings to triangulate.
3. NDA-shielded direct rates. The 97 CDF per session aggregator price is the public ceiling. Direct-to-MNO commercial agreements held by larger operators will be more favourable, sometimes materially so, but those rates are not public. A calculator that bakes in $0.04 per session as gospel would systematically misprice the cost stack.

Operators sizing a DRC entry, pricing a SONAL licence or modelling EAGT cost-of-compliance can request a tailored model. We treat DRC modelling as a bespoke exercise until enough audited primary data accumulates to meet our public-calculator bar. In the meantime, the four markets we do cover are the right comparable starting points.

13. What every DRC-active operator should be building right now

The reform package is non-negotiable, the deadline is fixed and the technical work falls almost entirely on the operator. A pragmatic engineering checklist for the next 60 days:

• EAGT API connector with retry, timeout and idempotency. Treat EAGT as a third-party payment processor: circuit breakers, exponential backoff, idempotent retries and an exception queue for transactions that complete locally before EAGT confirms. Avoid putting EAGT validation on the synchronous critical path of a USSD payout flow.
• Mandatory KYC via mobile-money provider data. SIM registration data and mobile-money KYC become the cheapest path to a verifiable digital account. Bridge M-Pesa, Airtel Money and Orange Money KYC into your account creation flow rather than asking a feature-phone user to type a 13-digit national ID into a USSD menu.
• Dual-currency wallet with FX-buffered ledger. Settle in CDF at the customer face; ledger in USD with a configurable spread buffer. Re-evaluate monthly against the BCC reference rate and the parallel-market premium.
• ASCII-only French menu copy. No accents. No exceptions. Add a build-time linter that fails the deploy on non-ASCII menu strings.
• Tax-deduction telemetry. Log the 10% WHT deduction at the same point you call EAGT and treat the regulator response as a reconciliation primary key, not a fire-and-forget event.
• Tax-repulsion countermeasures. Bonus design, free-bet allocations and odds-boost campaigns must absorb part of the perceived 10% haircut, or marginal punters migrate to grey-market platforms.
• Carrier-availability monitoring. Vodacom, Airtel and Orange downtime each translate directly to dropped USSD sessions. Build per-carrier dashboards and route around outages where possible.

The reform creates a structural advantage for operators that take EAGT integration seriously this quarter. Once the regulator has a year of clean transactional telemetry, market-share visibility is one report away, and the operators who showed up with clean integrations on April 1 will be the ones SONAL trusts when the next licence window opens.

Data gaps and limitations

• $1.7B handle estimate is widely cited but conflates turnover and GGR across sources. Direction-of-travel only.
• Offline adult population (36.8M-40M) is a linear extrapolation from 30.6% internet penetration applied to an estimated adult cohort. Multi-SIM ownership inflates uncertainty.
• Operator retained margin (10-15%) is a Pan-African baseline; not DRC-verified.
• 97 CDF per session is aggregator postpaid pricing. Direct-to-MNO rates are NDA-shielded and likely lower for larger operators.
• Annual SONAL renewals are mandated but not publicly digitized. Treat the $50K + $50K initial fees as a floor, not a full-life-cycle cost.
• EAGT technical specification is not yet fully public. API surface, latency budgets and SLAs will only be confirmed as licensed operators complete integration.

Compare with neighbouring markets

DRC's reform package sits alongside parallel regulatory shifts across the region. Three useful comparisons:

Ghana abolished its betting WHT to reduce grey-market drift; DRC is imposing a 10% WHT on the same structural problem. Watching how the two countries' regulated market shares move over the next 18 months will be one of the cleanest natural experiments African iGaming has run.