The verdict in three sentences
For a Kenyan marketplace, the buyer/platform/seller split is the most critical operation: a 1-shilling error repeated across thousands of orders destroys seller trust. The 2026 platform commission sits between 10 and 15%, and the M-Pesa B2C payout adds Safaricom's per-transfer fee. The right architecture isn't a real-time split at every collection but a deferred batch payout that pays 200 sellers in under 5 minutes with a 1-3% failure rate auto-replayed.
Real-time split or batch payout?
Both models have merit. Here's the 2026 decision comparison.
| Criterion | Real-time split | Batch payout (weekly) |
|---|---|---|
| Payout fee | per transaction | per grouped transfer |
| M-Pesa operations | 1 per order | 1 per seller/week |
| API load | very high | controlled |
| Platform float | none | weekly float |
| Reconciliation | complex | 1 batch to verify |
| Seller payout delay | T+0 | T+0 to T+2 (batch day) |
| Failure rate to replay | 1-3% scattered | 1-3% grouped |
Real time pleases impatient sellers but multiplies fees and API load. Weekly batch is cheaper, simpler to reconcile and covers 90% of marketplaces.
Kenya (M-Pesa Daraja) vs Cameroon (Orange Money)
The two corridors differ on caps and delays. 2026 ballpark figures.
| Parameter | M-Pesa Daraja B2C Kenya | Orange Money Cameroon |
|---|---|---|
| Platform commission | 10-15% | 10-15% |
| Payout fee | Safaricom B2C tariff | 1-1.5% |
| Transfer cap | 250,000 KES/transaction | 1,000,000 FCFA |
| Payout delay | T+0 (near-instant) | T+0 to T+2 |
| Batch 200 sellers | <5 min | <5 min |
| Payout failure rate | 1-3% | 1-3% |
| API | Daraja B2C | OM Business |
M-Pesa Daraja shines on instant transfers; Orange Money Cameroon forces you to manage a T+0-to-T+2 window and per-transaction caps that sometimes require splitting a large payout.
Mini case study
Wanjiru runs a crafts marketplace in Nairobi: 200 active sellers, 4,000,000 KES/week in volume, 12% platform commission. She keeps 480,000 KES in commission and pays out 3,520,000 KES to sellers. Batching weekly instead of real time cuts API operations by ~30x and reconciles in a single pass. The 1-3% failures (2-6 transfers) are auto-replayed the next day, avoiding any manual seller complaint. Any transfer above the 250,000 KES cap is automatically split and traced for reconciliation.
Need a professional website?
Kolonell builds websites that attract clients, optimized for the Sénégalese market. Free quote in 2 minutes.
FAQ
Should I split in real time or batch?
Weekly batch in most cases: fewer fees, less API load, simpler reconciliation. Real time is only justified if your sellers demand instant payout as a recruitment argument.
What platform commission is reasonable in 2026?
Between 10 and 15% depending on category and value added (delivery, marketing, guarantee). Below 10% you barely cover payout fees and support.
How do I handle a payout failure?
Use an automatic retry queue: the 1-3% of failed transfers are replayed the next day, with an alert if a seller stays blocked after two attempts. Never hand-fix hundreds of lines.
Does the M-Pesa cap block large sellers?
Above 250,000 KES per transfer you must split into several operations. A good payout engine handles that chunking automatically and traces it for reconciliation.
Can I test the split with real amounts?
Yes, and it's mandatory before production. We validate the commission/seller calculation to the shilling on multi-item orders and partial refunds before opening to real sellers.
Let's talk about your project. We build your M-Pesa Daraja marketplace split with batch payouts tested to the shilling. WhatsApp +221 77 596 93 33.
Mohamed Bah
Fondateur, Kolonell
Passionate about digital and entrepreneurship in Africa, Mohamed has been helping Sénégalese businesses with their digital transformation since 2020. Founder of Kolonell, he believes every SME deserves a professional and accessible online présence.