Questions people actually ask.

A permanent payment identity for your business — a number like NP-847291 that you give customers instead of a wallet address. Underneath it we maintain payment endpoints on each supported asset and network. Those endpoints can rotate for security or operational reasons; your account number does not change.

A defined supported set — USDC and USDT on a small number of networks at launch, with the current set shown at the point of payment so a payment page never offers a rail that would fail. We will not say “any chain”: it is not technically or legally true, and nobody can honestly promise it.

No, and that is the point. You send one line — “please pay invoice INV-1034 to Nomapay account NP-847291”. They pick from what is supported. The account resolves the rest.

What you are expecting, declared before the money arrives: invoice reference, expected payer, expected amount, accepted assets and an expiry. Without one, all we can tell you is that money arrived. With one, the arrival matches itself to an invoice and a payer — and a payment that lands $50 short inside your tolerance is flagged as matched with variance for you to accept or query, rather than silently settled or silently held.

No. You should never have to know what a network is. If our product requires you to learn, we have built the wrong product.

No — not on-chain and not in fiat. The account identifies where incoming value belongs; it does not mean we store it. Digital asset custody sits with a licensed or authorised partner, and conversion and payout sit with a licensed partner under their own permissions. We orchestrate: identity, detection, screening, matching, routing, ledger and record. No pooled customer wallets, no indefinite balances, no lending, no yield.

No, and we do not intend to build one. Cross-chain transport is infrastructure that already exists; we consume it and own the identity, detection, screening, routing and settlement instead. Where a route is needed we prefer native or canonical asset movement over wrapped representations, and we never pick a route merely because it is cheap when its operational or security risk is materially higher.

It becomes unknown, which is a first-class state rather than a failure. If a provider times out after we send an execution instruction we do not know whether it executed, and blindly retrying a $500,000 conversion sends it twice. So we query the provider, reconcile, and retry only once the original is definitively confirmed not to have happened. Every external operation carries an idempotency key.

Only to the verified corporate bank account on your profile. A payment cannot be redirected to an arbitrary account, and changing the settlement account requires authentication, approval, audit logging and — where appropriate — a cooling-off period.

It can happen in Nigeria, particularly where a counterparty is later linked to fraud. It is one reason we screen every sending wallet before settlement and keep the provenance ourselves, so you hold documentation rather than an assurance from a desk. We cannot promise it will never happen, and anyone who does is overselling.

Never. We need the verified settlement account for the payout and nothing more. Nobody from Nomapay will ever ask for a password or a one-time code.

One fee on the conversion, charged once, in a band from 0.25% to 0.75% depending on transaction size, corridor, asset, network, volume and operational complexity. That is our entire revenue on a payment. It does not have a second, quieter half hidden inside the exchange rate — and if we ever do earn from FX spread, that gets disclosed.

Network cost, normalisation cost and the liquidity or execution cost are shown as separate lines at what they cost. We mark up none of them. Marking up the network cost would quietly make the route choice ours to profit from, which is the opposite of what we are selling.

Quotes are indicative unless the underlying provider guarantees execution. A locked rate means somebody carries FX risk between quoting and executing, which needs pre-funded balances we do not hold. We could promise it anyway and it would hold most of the time — the times it did not would be the volatile days, which is to say the days it mattered.

No, and we will not claim it. We cannot see every provider or every transaction, so nobody honestly can. What we will do is show you exactly what it cost to receive and settle the payment.

No — not custody partners, not liquidity venues, not conversion desks. A route chosen by someone the route is paying is worth nothing, and the routing decision is the product. If that ever changed we would state it here before it took effect.

Often not. Network costs are largely fixed, so they matter at $2,000 and barely register at $100,000. There may be a size below which neither of us comes out ahead, and we would rather put that on the page than discover it with you.

We do not use the word instant. Once there is real history we will publish a median, a 95th percentile and a failure rate measured from detection to bank credit, rather than a number chosen by marketing.

Because those are different things. A reference rate is a public benchmark at a moment in time, useful for judging what happened but not something you can transact at. An executable rate is what a partner will actually trade at, at your size, right now. We show both, with the reference source named.

The record says so. We do not fill the gap with an estimate — an estimate presented as a benchmark is worse than no benchmark.

Two reasons, one of which benefits us. Keeping the network private lets us move volume the moment a partner stops being competitive, without you having a relationship that makes that awkward. It also protects the settlement performance data we are accumulating. You still see every figure that affects you.

Every sending wallet, before settlement: sanctions, known illicit addresses, mixers, ransomware and stolen-funds exposure, darknet exposure and fraud indicators. Clear, review or blocked. An adverse result stops the payment and escalates to a person, and the result is retained as part of the transaction record either way.

We are in private beta, by invitation. The first payments are run by hand — monitored, routed and settled manually — so that what gets built is what those businesses actually needed. Open signup comes after that works.

If fewer than five of thirty interviewed businesses have hit a meaningful mismatch, we do not build the routing product. And frequency alone would mislead us: a 20% mismatch rate on $2,000 payments is irrelevant, while a 5% rate on $500,000 payments is an excellent business — so we measure affected volume, not just how often it happens.

Unclear, and it is what worries us most. McKinsey and Artemis put genuine stablecoin payment activity at roughly $390bn in 2025 once trading and internal transfers are filtered out, with B2B the largest slice at about $226bn. The IMF estimates Nigeria received around $59bn in crypto-asset inflows between July 2023 and June 2024, with stablecoins more than 65% of 2024 inflows. That establishes the rail exists. It does not establish the size of our addressable receiving-account market.

Eventually. The same job through a consumer interface, and deliberately not first — consumer acquisition, support and fraud handling are expensive things to learn on while the core question is unproven.

That is the plan and it may be the largest channel: give your own users a receiving account while you keep the customer relationship entirely. The shape is here; the API is designed, not shipped.

We are onboarding our first businesses now. Publishing a number this early would be marketing rather than information; when it is meaningful we will publish it and explain how it was counted.