What mobile money integration costs, the M-Pesa, MTN MoMo, Flutterwave and Paystack routes, timelines, and the licensing and reconciliation that matter.
Connecting your product to mobile money wallets across Africa costs about $2,000 to $6,200 in engineering for a direct single-provider build, or a few hundred dollars and under a week if you route through an aggregator such as Flutterwave or Paystack. Ongoing per-transaction fees range from roughly 0.5% to 2.5% depending on the route you pick. In 2025, mobile money moved $2.1 trillion worldwide, according to the GSMA State of the Industry Report 2026, so for software sold into Kenya, Tanzania, Nigeria, Ghana, or Uganda this is the checkout, not an extra.
| Route | Upfront engineering | Timeline | Per-transaction fee |
|---|---|---|---|
| Direct M-Pesa Daraja | $2,000-$6,200 (KES 250k-800k) | 3-5 weeks | ~0.5%-1.5% (Safaricom tier) |
| Aggregator (Flutterwave, Paystack) | ~$1,500-$6,000 | 3-10 days | 0.5%-2.5% + rail fee |
| No-code plugin (WordPress, Shopify) | $0-$500 | hours | higher blended fee |
| Multi-country custom wallet | $15,000-$60,000 | 2-4 months | volume-based |
These are planning ranges, not quotes.
Businesses integrate for a few reasons, and each maps to a different API. A merchant taking payment at checkout needs STK Push or C2B. A marketplace or gig platform paying out sellers needs B2C. A lending or savings product needs both, plus balance queries. The mix determines your scope, not just the market.
Integration means your app or internal system talks to telecom-operated wallets so payments, refunds, and balance checks happen automatically. The core operations carry names from Safaricom's Daraja platform.
STK Push (Express) is the checkout prompt. Your server sends a request, the customer's phone shows a PIN screen, and you receive a callback when the payment settles. C2B covers customer-initiated payments to your Paybill or Till number. B2C pushes money out for payouts, refunds, and disbursements. Status and balance endpoints let you query a transaction when a callback never arrives.
The happy path is easy. The callbacks are not. Daraja webhooks are not HMAC-signed the way Stripe's are, so your server needs a state machine that treats the callback and a status query as two inputs to the same transaction. If your reconciliation is not idempotent, a retry becomes a double charge or a paid order marked as failed. Settlement timing also varies. M-Pesa settles in near real time, while some West African rails batch, which changes how you time payouts and how you reconcile.
The market splits into direct operator APIs and aggregators.
M-Pesa Daraja is the direct API for Safaricom in Kenya and Tanzania. Registration and the sandbox are free; you pay tiered transaction fees around 0.5% to 1.5% in production. Daraja 3.0 arrived in November 2025 on cloud infrastructure with capacity for 12,000 transactions per second.
MTN MoMo spans 16-plus African markets and leads in West and Central Africa. Airtel Money covers Airtel's markets. For several rails under one contract, Flutterwave (34-plus African countries) and Paystack (Nigeria, Ghana, South Africa, Kenya) are the defaults, with pawaPay and Onafriq for mobile-money-only reach and Africa's Talking for SMS and USSD next to payments.
Direct integration costs less per transaction at volume and keeps the code yours. An aggregator goes live in days but adds a per-transaction margin. Choose by market, not by preference. In Kenya you want M-Pesa directly or through an aggregator that supports it. In Nigeria the same product leans on Paystack or Flutterwave plus NIBSS bank transfers.
A single direct M-Pesa build in Kenya is commonly quoted at KES 250,000 to 800,000. Aggregator integrations start cheaper but add 0.5% to 2.5% per transaction on top of the operator's fee. A product that must take mobile money across three or four countries is a different job: a payment abstraction layer, per-provider adapters, and reconciliation across different settlement cycles. That scope runs $15,000 to $60,000 and two to four months.
The main cost drivers are the number of providers, whether you need B2C payouts and refunds rather than collection only, idempotent reconciliation, settlement reporting, and the compliance paperwork each market demands. Plan for maintenance as well. Operators change endpoints and fee tiers, and your reconciliation rules need updating when they do. A reasonable allowance is 15% to 20% of the initial build per year. For a deeper look at the surrounding payment infrastructure, see our guide on building a cross-border payment platform.
For accepting payments, no. Integrate through a licensed operator or aggregator. Kenya's Central Bank of Kenya licenses payment service providers and e-money issuers, and you still need a Safaricom Paybill or Till number to receive money. Nigeria's Central Bank separates mobile money operators (NGN 2 billion minimum capital) from payment solution service providers (NGN 100 million), which is the licence most software companies work under. South Africa sits under the South African Reserve Bank and the FSCA through the National Payment System Act, and Ghana under the Bank of Ghana.
Holding customer balances or issuing wallets is a heavier licensing burden. If your product does that, treat it as a separate project. For collecting payments, work through a licensed partner and keep your own records clean.
A direct M-Pesa build takes three to five weeks. An aggregator route through Paystack or Flutterwave can be live in under a week. A multi-country product with several rails runs two to four months.
Not for accepting payments. Use a licensed operator or aggregator. You need your own licence only if you hold customer balances, issue wallets, or operate as a payment service provider in a given country.
The Daraja API and sandbox are free to register. You pay Safaricom's per-transaction fees in production, plus an aggregator margin if you route through a third party.
Direct is cheaper per transaction at volume but costs more to build. An aggregator has higher per-transaction fees but near-zero upfront work. Below a few hundred transactions a day, the aggregator usually wins on total cost.
Build reconciliation around an idempotent state machine that treats callbacks and status queries as inputs to a single transaction. Never key a payment record on one webhook alone.
No single API covers all of them. You build a payment abstraction layer and per-provider adapters, or use an aggregator that already spans the markets you need.
GlobeSoft is a software development company based in Shenzhen, China, and fintech and online payment systems are one of our core areas alongside ERP and CRM. We have delivered projects for clients in South Africa, Brazil, Singapore, and the US. When we take on a mobile money build, we fix the state machine and reconciliation rules before the first API call, because that is where these projects quietly fail. If you are planning one, see our fintech development service or how to build a payment gateway.
GlobeSoft builds custom software, AI applications, SaaS, FinTech and enterprise systems for clients worldwide. Message us on WhatsApp for the fastest answer, or email us — we typically reply within 24 hours.
Keep reading
KYC is the gate every FinTech product has to pass before a single payment moves. Here is what it covers, where it lives in the funnel, and whether to build it or buy it.
5 min readConnecting two systems is easy on the happy path. The real work is in retries, idempotency, rate limits, and data mapping. Here's where the hours go.
10 min readWhat white-label software development really involves, how a commissioned build works, what it costs, and how to choose between rebranding and building.
10 min readA practical playbook for working with a China-based software team: communication rhythm, contracts, IP, quality checks, and the mistakes that sink projects.
13 min readA real estate CRM is more than a contact list. Here are the modules, MLS and IDX integrations, and commission handling that decide the build, and what it costs.
9 min readWhat belongs in your first mobile app build, what to cut, and how to ship an MVP that gives you real signal without burning your budget.
