Who Gets to Build Nigeria's Next N3 Trillion Lending Engine?
The federal government just cracked open the airtime credit monopoly. Here is why the real war is about APIs, data pipelines, and surviving the scaling nightmare of USSD.
Sapa is a universal experience, but there is a specific kind of panic that hits when you are trying to order a ride in the middle of a sudden Gbagada downpour and your data plan breathes its last breath. You quickly dial the emergency credit code. Within five seconds, you are back online, matching with a driver.
That micro-transaction is part of an absolute monster of a market—roughly N3 trillion in annual transaction value. For over a decade, this entire pipeline was basically a one-house show run by Optasia (formerly Channel VAS). Now, the Federal Competition and Consumer Protection Commission (FCCPC) has recommended nine local tech companies to break up this monopoly.
As a developer, my mind instantly skips past the high-level policy talk. I am thinking about the sheer engineering chaos, the database queries, and the infrastructure it takes to run a system like this in Nigeria.
The Brutal Reality of USSD Latency
If you have ever built anything that connects to Nigerian telecom infrastructure, you know that integrating with telcos is not for the weak. This is not like spinning up a Stripe or Paystack integration in an afternoon. You are dealing with legacy telecom protocols, SMPP gateways, and APIs that sometimes feel like they were written during the Olusegun Obasanjo administration.
When a user dials a USSD code to borrow N200 airtime, you have exactly 15 to 20 seconds before the session times out. In that tiny window, your system has to:
- Receive the request from the telco's gateway.
- Query your database to see who this user is.
- Run a risk assessment engine (do they pay back? how active is this SIM?).
- Hit the telco's provisioning API to credit the line.
- Send a confirmation SMS.
If your database takes more than two seconds to resolve a query because your indexes are messed up, the session drops. The user gets an "Error: Connection problem or invalid MMI code" message, and you lose them. Scaling this to handle millions of concurrent hits on a Friday evening when everyone is prepping for the weekend requires serious backend engineering. We are talking Redis caching on steroids, event-driven microservices, and robust queue management using tools like RabbitMQ or Kafka.
Who Scores the Guy in Akure?
The real battle for these nine newly licensed local companies isn't just about getting the license; it is about who has the smartest risk engine.
Think about a trader in a market in Onitsha or a student in Akure. They do not have a formal bank statement or a traditional credit score. But they have used their SIM card for three years. They make calls every morning, receive funds via mobile money, and browse mostly at night.
That raw behavioral data is gold.
If you can build a machine learning model that ingests these raw telecom logs and accurately predicts—in milliseconds—whether a user will repay a N500 loan when they next recharge, you win the market. If your model is too conservative, you lose transaction volume. If it is too loose, "no gree for anybody" becomes the user's motto, and bad loans will wipe out your capital before you can even complain to the FCCPC.
The Infrastructure Struggle is Lived-In
Let us be honest about building tech here. We are not writing code in cozy, air-conditioned Silicon Valley offices with fiber-optic lines that never blink. We are writing code while keeping one eye on the inverter battery and the other on the AWS billing dashboard.
When you are integrating with local telcos, things break. APIs go down without warning at 2:00 AM. A fiber cut somewhere in the ocean can spike latency across the entire country.
The local players entering this space cannot just be good at writing clean code. They have to be masters of chaos engineering. You need self-healing systems, automatic retries with exponential backoff, and deep fallback systems. If MTN’s API fails to respond, can your system gracefully queue the request without crashing the entire service?
The Real Prize Lies Beyond Airtime
The airtime is just the bait. The actual trophy is the credit history database.
If these nine local companies can successfully build profiles on 50 million Nigerians who previously had zero credit footprints, they are sitting on an absolute goldmine. With data portability on the horizon, this data can be fed directly to credit bureaus and traditional fintechs.
Imagine a developer building a micro-insurance app or a device financing startup. Instead of asking a user to upload a bank statement they don't have, they can query one of these new airtime lenders: "Has this number consistently repaid their N1,000 weekly data advance for the last six months?"
That is how we build real financial inclusion. It is not by writing fancy press releases about "underserved demographics." It is by doing the dirty, unglamorous work of building stable APIs, indexing databases correctly, and keeping the servers running when the national grid decides to take its weekly nap.
I am watching this space closely. The licensing is done. Now, let the best engineering team win.
Related from Business
Let's build your next big product.
Accepting project-based freelance, remote engineering roles, and hybrid positions.