How Nigerian fintechs die: not out of cash, out of usable cash

Most seed models in Nigerian fintech start with headcount and growth, then push compliance, disputes and FX into “later”. The evidence in this brief shows those three lines are often the early cliff edge, because they restrict usable cash and can interrupt operations on a regulator’s timeline, not yours.

The moment looks mundane on a spreadsheet.

Imagine you are a payments founder in Lagos. Payroll is due. A vendor invoice lands with a USD price tag. Meanwhile, the money in your account is not all yours. Some of it is already spoken for by the rules of the game: licensing steps, compliance commitments, and the ugly time lag of disputes you have not even seen yet.

That is the cash-crunch moment founders do not model, because it rarely announces itself as “crisis”. It arrives as admin.

And it is why the most dangerous money in a Nigerian fintech seed deck is the money you cannot touch.

Here is one concrete reminder that this is not theoretical.

On 1 September 2022, Flutterwave, the Nigeria-born payments company, announced it had received a Central Bank of Nigeria Switching and Processing licence.

Flutterwave’s 2022 CBN licence announcement shows how top-tier authorisation is treated as a milestone, not admin. Photo: Opelogbon / Wikimedia Commons, CC BY-SA 4.0.

In their own announcement, Flutterwave called it “Nigeria’s highest payments processing license”. That is their phrasing, not ours, and it matters because it shows how operators themselves frame the authorisation: not as a nice-to-have, but as a top-tier gate.

The simple question that moment raises for every seed-stage payments founder is not “How big a team can we afford?” It is: if licensing is a gate, and gates open on regulators’ timelines, how much should you raise so you do not run out of *unrestricted liquidity* while you wait?

Let’s define that once, plainly, because people keep talking past each other.

Unrestricted liquidity is the cash you can actually spend this week on wages and vendors. Restricted (or committed) liquidity is cash you must keep aside because a regulator, a scheme rule, or an operational reserve effectively earmarks it. Both sit in the bank. Only one is truly yours.

Most seed conversations answer a different question.

They ask, “What did other fintechs raise?” Then they back into 12 to 18 months of burn dominated by headcount, product and growth. Compliance is a line in G&A. Fraud is a risk slide. FX is a footnote.

At first glance, it looks sensible. Most startups fail because they run out of money, and the biggest expense is people.

But the evidence in this brief keeps pointing to a different failure mode for regulated payments rails. They do not mainly fail because they ran out of time to build product. They fail because they ran out of money they were allowed to use, during a regulatory delay, a dispute spike, or an FX step-change that reprices USD-linked inputs.

That is the twist.

Seed size in Nigerian regulated payments is less about “months of burn” and more about financing a sequence of gating milestones while holding explicit liquidity buffers for three traps: (1) licensing and compliance, (2) fraud and chargebacks, and (3) USD-linked vendor costs under FX volatility.

The myth: seed is a headcount problem

The seed model that gets repeated in founder circles is basically a staffing plan.

Build. Hire. Sell. Raise again.

It is not that founders are careless. It is that a lot of “market norms” come from businesses where you can keep operating while you negotiate the hard stuff.

In payments, the hard stuff can stop you operating.

The clause nobody reads: licensing can lock up cash

Nigeria’s licensing regime does not merely ask you to behave well. It can embed material cash commitments that function like ring-fences.

The Central Bank of Nigeria’s “Approved New Licence Categorization Requirements Consolidated” document from 2021 explicitly lists requirements such as escrow deposits, capital requirements and fees across payment system provider categories, including Switching & Processing and PSSP.

The CBN’s 2021 licensing categorisation document spells out escrow deposits, capital requirements and fees that can restrict usable cash. Photo: Aty Jorbes / Wikimedia Commons, CC BY-SA 4.0.

This is where most people stop looking. They see “licensing” and think “paperwork”. The CBN document reads more like a list of practical hurdles: you do not just file forms, you commit real money.

The implication is uncomfortable but clarifying.

Two companies can have the same monthly headcount burn and wildly different survival odds, depending on how much of their cash is effectively committed by licensing requirements or operational reserves.

Flutterwave’s 2022 announcement is instructive because it highlights a decision many seed decks postpone: pursue top-tier authorisation, with its cost and timeline, or remain partner-dependent and hope to “do licensing later”.

We expected this to be a sequencing debate. The interesting part is that in regulated rails, it is often a binary constraint. Licensing is not a nice-to-have upgrade. It is permission to operate at all, or to change your dependence on someone else’s permission.

The hard stop: partnerships are not permission

If Nigeria shows you licensing is a gate, Kenya shows you what happens when you try to walk around it.

In July 2022, the Central Bank of Kenya directed financial institutions to cease dealing with Flutterwave and Chipper Cash, saying they were not licensed or authorised, according to TechCrunch.

In operational terms, this is not a “risk factor”. It is a cliff edge. Revenue can stop or slow. Legal and compliance work speeds up. Payroll remains payroll.

Flutterwave later said it entered Kenya via partnerships. That is a common playbook: use banks or other institutions to access the market while you figure out the rest.

The Kenya episode is a reminder that partnerships may help distribution, but they do not reliably solve authorisation. Regulators can still demand direct licensing.

And if you are trying to size a seed round, that detail changes everything. A partnership-led expansion plan without licensing contingencies is not just optimistic. It is underfunded.

The invisible drain: disputes hit later, and they hit cash

Fraud and chargebacks are often described like weather: unfortunate, unavoidable, hopefully mild.

But the reason disputes are dangerous is the timing. You process volume today. A dispute lands later. Cash moves now, while resolution drags.

The brief points to an industry controls source, the Payments Controller Handbook (2025), hosted at paymentscontroller.com, as emphasising fraud/chargeback reserves and the lag between transaction and dispute resolution.

We can cite it as a source in this brief, but we are not going to hang any extra “many fintechs die” claims on it. The evidence we have is narrower and stronger: disputes create delayed cash leakage and the need for reserves, and that directly constrains usable liquidity.

If you have never run a payments operation, picture a shop where customers can return a month later and demand the money back, while your supplier insists on being paid today. That is not a growth problem. It is a cash timing problem.

The credibility bet: Paystack raised for durability, not just an MVP

Paystack’s early numbers are easy to gloss over because they are old history now. But they matter because they show a different instinct: raise enough to build durability in a trust-bound market.

Paystack publicly stated it raised a $1.3m seed in 2016 and then an $8m Series A announced in 2018 led by Stripe with Visa participating.

The story is not “raise big”. It is “raise for what can kill you”. In regulated payments, under-raising does not just mean fewer hires. It can mean postponing controls, compliance capability, and the buffers that keep you alive when the system does what the system does.

The fundraising trap: regulation can slow your next cheque

Another assumption baked into 12 to 18 month seed planning is that if you hit metrics, follow-on capital will arrive quickly.

Kuda’s story complicates that.

TechCrunch reported in January 2024 on Kuda’s attempted 2023 fundraise at a flat valuation, and noted Kuda’s statement that as a regulated entity it could not share certain numbers until audit or regulator approval.

You do not need to take a view on Kuda’s growth projections to take the lesson: regulation can affect disclosure cadence, audits, and diligence timelines. That means capital access can slow exactly when you want it to be fast.

So your seed runway cannot assume your fundraising runway will behave like a SaaS startup’s.

If you have ever watched a “reasonable” bridge become a future investor’s headache, GI Network has written about this dynamic in The bridge looked “reasonable”. The next lead called it uninvestable..

The world tour: Nigeria is not special, the pattern is

It is tempting to treat Nigeria as uniquely chaotic, therefore uniquely hard to model.

But put Brazil next to Nigeria and the pattern sharpens.

The IMF’s 2023 fintech note on Latin America discusses that fintechs, including Nubank, need to comply with capital requirements, and it references regulation updates first issued in 2013 and evolving thereafter.

Different regulatory architecture. Same basic reality: once you are inside a prudential regime, capital planning is not optional.

Then look back to Kenya’s hard stop. Or to the time lag of disputes.

Different countries, same lesson: the constraint is not ambition. It is permission, liquidity, and timing.

The pattern the decks miss

Put these cases side by side and a pattern appears that none of the individual stories states outright.

Seed-stage founders talk about runway as if it is one number: cash divided by monthly spend.

But regulated payments runway is two runways.

1. 1.Unrestricted runway: money you can spend freely on people, product and growth.

1. 2.Committed runway: money tied up by licensing requirements (including escrow deposits, capital requirements and fees as set out in the CBN’s 2021 categorisation document), plus reserves and other obligations that are not discretionary when you need them.

When the second runway eats the first, you get the classic “everything looked fine” failure. The spreadsheet says 14 months. The bank account says you have six weeks.

That is how strategy drift happens post-close. A founder raises seed “for growth”. Then licensing takes longer, disputes rise, or costs reprice, and suddenly the plan is not growth but survival. Investors call it missed execution. Founders call it reality.

GI Network’s view: In regulated payments, the seed question is not “How many months of burn?” It is “How many gating milestones can you clear without ever touching money you will later be forced to ring-fence?”

When the opposite is true

This is not a universal law about “fintech”. It is a specific pattern about regulated payments rails.

If you are a pure software layer and do not carry the licensing obligations, capital requirements, or dispute liability, your committed cash pile may be smaller.

Even then, the Kenya episode is the cautionary note: “partners hold the risk” is not something you assume. It is something you verify, jurisdiction by jurisdiction, with the regulator’s posture in mind.

For businesses: how to turn milestones into a seed size that holds

Stop starting with “how much are people raising?” Start with “what gates must we pass, and what cash must we hold while we pass them?”

1) Write the gating milestones like a regulator would

Not “launch v2”. Not “expand to Kenya”.

Write milestones that can be verified and that unlock operations.

Use the CBN’s categories as the backbone in Nigeria, because the 2021 consolidated requirements explicitly tie categories to items such as escrow deposits, capital requirements and fees.

2) Separate spend from buffers (do not blend them)

Present execution burn separately from:

• licensing and compliance commitments (including cash requirements set out in the CBN’s categorisation requirements)
• dispute and chargeback reserves (because timing alone can squeeze you)
• USD-linked vendor exposure under FX volatility (because costs can reprice while headcount stays flat)

If you blend these, you will overestimate what you can deploy when something slips.

3) Build a timeline that assumes delays

Kuda’s disclosure constraint is a reminder that regulated status can slow diligence and fundraising.

Model for it. Do not let “we will raise quickly” be the silent assumption holding up your whole plan.

4) Make the use of funds legible to sceptical investors

One page, four buckets:

• build and operate
• licensing and compliance
• disputes and fraud controls
• FX-linked cost buffer

If you need a reference for how investors think about trust before anything else, start with Capital moves on trust: what funders expect before they look at anything.

For investors: what experienced cheques are really buying at seed

When a seasoned investor pushes for longer runway in regulated payments, they are often underwriting constraint, not indulging burn.

Three things they are quietly checking:

1) Is the round buying permission, not just progress?

Flutterwave’s licensing milestone and the CBN’s explicit requirements make the point: permission is a funded activity.

2) Is there a plan for the hard-stop risk?

The CBK action in 2022 is a diligence prompt: show me the authorisation strategy market by market. “We have partners” is not the same as “we have permission”.

3) Is the model honest about cash timing?

Disputes are not only losses. They are timing mismatches.

Investors have seen too many “14 months runway” decks that secretly assume authorisation stays on schedule and the next raise is instant. When either slips, you get the forced bridge. And then the next lead asks hard questions.

For how story coherence affects capital velocity, see Why Fundraising Stalls, and What Serious Companies Do Differently.

What GI Network would do in this situation

GI Network would start by stress-testing whether the business can survive the three liquidity traps this brief highlights: licensing and compliance gating, dispute and chargeback timing/reserves, and USD-linked cost shocks under FX volatility. Concretely, we would rebuild the use of funds around verifiable milestones taken from the CBN’s 2021 categorisation requirements, separate unrestricted operating cash from committed buffers, and rehearse the investment committee objections that typically kill these rounds: what happens if authorisation slips, what is the reserve logic for disputes, and how exposed are core vendors to FX.

The takeaway tool: the Minimum Survivable Raise (MSR)

The Minimum Survivable Raise (MSR) is not “18 months of burn”.

It is the smallest seed you can raise that still allows you to clear your next gating milestones without losing unrestricted liquidity.

A quick MSR checklist drawn from the cases above:

1. 1.Permission Gate: What licences or authorisations must be secured, and what explicit cash requirements (escrow deposits, capital requirements, fees) are tied to them in the CBN’s 2021 categorisation requirements?

1. 2.Delay Reality: If authorisation or diligence takes longer than planned, how long can you survive without changing strategy?

1. 3.Dispute Timing: If disputes spike and cash is tied up while they resolve, what happens to your ability to pay wages and vendors?

1. 4.FX Exposure: Which critical inputs are effectively USD-priced, and what happens to your buffer when FX moves?

If you can answer those four with numbers you are willing to show a sceptical investor, you are no longer guessing seed size. You are underwriting it.

And in Nigerian regulated payments, underwriting is the difference between “we’re out raising” and “we need a bridge by Friday”.