The verdict in three sentences
Each operator adds a different statement, format and reference: from 3 wallets, manual reconciliation becomes unmanageable. In Lome in 2026, aggregating Moov Money, T-Money and bank transfers into a single ledger via matching by reference divides discrepancies by 4 and saves 4 to 5 hours a week. In Tanzania, cross-checking M-Pesa (Vodacom), Tigo Pesa and Airtel Money via API/statement brings non-matches below 2 %.
The problem grows with the number of wallets
With a single wallet, a spreadsheet is enough. With three, references don't follow the same format, timestamps diverge, and the same customer may pay through two channels. Reconciliation time doesn't grow linearly: it explodes, and so does the non-match rate.
| Number of wallets | Reconciliation time/wk | Non-match rate | Gain after automation |
|---|---|---|---|
| 1 wallet | 1-2 h | 1-2 % | Low |
| 2 wallets | 3-4 h | 3-5 % | Medium |
| 3 wallets | 5-7 h | 6-10 % | High |
| 4+ wallets | 8-12 h | 10-15 % | Very high |
The Lome pipeline: one ledger by reference
Automation means normalizing each statement (Moov, T-Money, bank) into a common format, then matching by reference + amount + date. Whatever matches is marked paid; the rest falls into an exception queue. Result in Lome: discrepancies drop from about 8 % to under 2 %, and weekly time falls sharply.
| Metric | Before (3 manual wallets) | After automation | Gain |
|---|---|---|---|
| Time per week | 5-7 h | 1-1.5 h | ~75 % |
| Non-match rate | 6-10 % | < 2 % | /4 |
| Close time | 4-6 days | 1-2 days | -60 % |
| Automation cost | - | ~800,000-1,500,000 FCFA | paid back 6-10 mo |
| Frequency | Monthly | Daily | x20 |
Mini case study
Kodjo keeps the books for a distribution SME in Lome that collects via Moov Money, T-Money and bank transfers: about 1,200 operations a month. Before, he spent 6 hours a week reconciling, with 8 % non-matches dragging on until close. After a normalization and reference-matching pipeline, he drops to 1 h 15 a week and 1.8 % non-match. Over the year he frees about 250 hours and firms up his monthly close, cut from 5 days to 2 days.
FAQ
Need a professional website?
Kolonell builds websites that attract clients, optimized for the Sénégalese market. Free quote in 2 minutes.
From how many wallets should I automate?
From 3 wallets, automation becomes worthwhile: reconciliation time and the non-match rate climb too fast to stay manual.
How do I match references of different formats?
You normalize each statement into a common format (reference, amount, date, channel), then apply a matching rule tolerant of minor timestamp gaps.
What non-match rate should I target?
Manually across 3 wallets, 6 to 10 % is common. With an automated pipeline, aiming below 2 % is realistic.
How does it work in Tanzania?
You cross-check M-Pesa, Tigo Pesa and Airtel Money via API or statement, with the same single-ledger principle, bringing non-matches below 2 %.
How much does setup cost?
2026 order of magnitude: 800,000 to 1,500,000 FCFA depending on the number of wallets and volume, paid back in 6 to 10 months by the time saved.
Let's talk about your project. We build your single ledger and reference matching for Moov, T-Money and your transfers. 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.
