When Mobile Money Meets Overdraft: How M-PESA Built Fuliza
When Safaricom launched M-PESA in 2007, it transformed shipping money across Kenya—allowing millions to send, receive, top up, pay bills, and transact digitally through agents and phones. For many years, payments, transfers, and wallet balances were the core offering. But as M-PESA grew, Safaricom began encountering friction from one corner: failed transactions due to “insufficient funds.”
These daily failures—users wanting to send money, buy airtime, or pay bills but lacking enough balance—were a signal. They were not just annoyances. They revealed a deeper opportunity for a product innovation that could improve user satisfaction, reduce friction, and unlock incremental revenue. Enter Fuliza, a native overdraft facility built into the M-PESA ecosystem.
What Is Fuliza & Why It Matters
Fuliza is Safaricom’s overdraft / short-term credit service that allows users to complete M-PESA transactions even if their account lacks sufficient funds. In effect, when a user tries to perform a transaction and doesn’t have enough balance, Fuliza “tops up” the transaction by covering the shortfall, making the wallet balance go negative. Then, once the user receives a deposit (or makes a top-up), the negative balance + fees are automatically settled.
The product was officially launched in 2019. It relied heavily on Safaricom’s data science and big-data analytics teams, which analyzed transaction history, frequency, and wallet behaviour to assess eligibility, set limits, and price the overdraft facility. Because M-PESA had already amassed a large user base with rich transaction data, Safaricom was able to build a risk model with much less friction than a lender starting from scratch.
Key Features & Design Parameters
Here are some of Fuliza’s design choices that have made it effective:
Opt-in activation: Not all M-PESA users have access by default. Users need to accept the terms and opt in (e.g. via USSD or *234#).
Dynamic limits: A user’s Fuliza “limit” depends on their historical M-PESA usage, transaction behaviour, repayment discipline. Users with more frequent activity and good repayment may get higher overdraft caps.
Automatic repayment: Once there’s incoming funds to the wallet, M-PESA automatically clears the overdraft + fees. No need for manual effort on the user’s part. This helps reduce default risk and simplifies UX.
Fee structure: Fuliza charges include an access fee (for opting in) plus daily maintenance or usage-based fees for the overdraft portion. Users must be cautious as fees can build up if the negative balance persists.
Outcomes & Impacts
Reduced friction in payments. Users no longer bounce when they don’t have enough in their M-PESA wallet—this improves trust and platform stickiness.
Increased usage. Because Fuliza “fills the gap,” users are less likely to abandon transactions or defer bill payments. Many use it for everyday needs: completing purchases, paying bills at end of month, emergency expenses.
Revenue generation. Fuliza adds incremental fees and interest-like charges. For Safaricom, this product became another margin stream beyond the traditional transaction fees. It helped monetize the “last-mile gap” when wallet balance is low.
Data insights & risk calibration. Fuliza’s usage generates rich data about defaults, repayment behaviour, user sensitivity to fees, etc. This strengthens Safaricom’s ability to refine credit scoring, set dynamic limits, and fine-tune risk thresholds.
Lessons for Platforms Considering Overdraft / Negative Balance Products
For platforms (wallets, e-wallets, mobile money operators) considering adding similar offerings, here are lessons from Fuliza that are especially relevant:
Leverage existing transaction history. If you already have frequent wallet or payment usage, that data becomes your foundation for building credit risk models without needing costly alternative credit sources.
Start small with clear guardrails. Fuliza worked because the product design included opt-in, dynamic limits, auto repayment, and disciplined fee structures. Starting with modest limits and conservative risk appetite means you can gather experience.
Design for transparency and trust. Users need to clearly understand what fees will be, how repayment happens, and what the consequences of non-repayment are. Lack of clarity breeds distrust, default, and reputational risk.
Governance, oversight & risk control. Overdrafts carry more risk than simple payments: DPD (days past due), adverse selection, misuse, etc. Platforms need governance structures (credit committees or steering committees), monitoring dashboards, risk triggers.
Use the overdraft as a stepping stone. Fuliza isn’t meant to replace traditional credit but to complement it, generating data and user behaviour insights that can support larger products (term loans, salary advances, etc.).
What This Means for Operators and Platforms
Fuliza’s success demonstrates three critical lessons for mobile money operators, e-wallets, and payroll-card providers:
Timing is everything – credit at the moment of transaction drives adoption.
Data is destiny – frequent transactions feed better credit models, fast.
Revenue diversification is vital – as payments become commoditised, credit is one of the strongest levers for growth.
How Factfin Helps Replicate the Fuliza Effect
Factfin helps payment platforms, mobile money operators, and digital wallets replicate the commercial success of products like Fuliza by embedding credit innovation inside an AI-Innovation Lab.
Through the Lab, clients gain access to digital lending experts who use AI tools to explore, test, and scale new lending models, using the clients’ own transaction data as the foundation.
Data-driven credit modelling: Factfin builds transparent, machine-learning frameworks from wallet, payroll, or transaction data. Unlike vendor “black boxes,” these models are explainable, auditable, and the intellectual property belongs to the client.
Governance & control: We establish steering committees and credit forums so that management teams retain oversight of risk appetite, product rules, and customer eligibility.
MVP-to-scale pathway: The Lab enables quick, low-capital pilots (such as negative balance or pay-advance features) that generate repayment data rapidly. These insights then feed into more advanced loan products once confidence in the models grows.
Portfolio monitoring: Real-time dashboards track repayment performance, credit losses, and customer usage, ensuring data-driven conversations between credit, business development, and executives.
The AI-Innovation Lab ensures innovation isn’t just about launching new lending features — it’s about building the governance, risk control, and learning systems that turn those features into long-term revenue growth.
