Local payments layer for e-commerce and gaming in Africa

Global e-commerce, gaming, and creator platforms entering Africa keep arriving with a similar problem: their global PSP isn't getting it right on the continent, the volume they're missing is large enough to matter, and they don't know how to fix it without rebuilding the stack they spent years standardising on.

Global PSPs like Stripe, Adyen, Checkout.com were not built to meet the needs of digital commerce in African markets today, and the gaps are invisible on the dashboards merchants use to measure their performance on the continent.

While their dashboards show authorisation rates, conversion rates, and transaction volume, they don’t go deep enough to unearth the reasons that some of these payments are not converting. For example, a customer landed on the payment page, saw only card as an option, and left because they don't have one; a Nigerian shopper’s cross-border card was declined; a South African subscriber couldn't complete recurring billing because debit orders aren't supported. Or, for example, a Kenyan customer tried to pay via M-Pesa, but the transaction failed because the global PSP couldn't normalise their phone number into the format the rail accepts.

None of these show up as failures on the dashboard; they are simply customers who never existed. And in aggregate, they represent the majority of digital commerce in African markets.

Cards alone aren’t enough in African markets

The standard architecture for cross-border merchants entering Africa is often to use their existing global PSP that already powers checkout everywhere else. They’ll then enable cross-border card acceptance for African shoppers, and consider the market covered.

This works, in the sense that some volume gets processed. However, it leaves most of the available volume on the table, for three reasons:

• Cards are a minority payment method across most of the continent. Africa has 1.2 billion registered mobile money accounts and 347 million monthly active users, which is more than any region in the world, and accounts for roughly two-thirds of global mobile money transaction value. A checkout that only offers card does not address the full customer base.
• Cross-border card acceptance underperforms badly even when the customer does have a card. Authorisation rates on cross-border card transactions in African markets routinely sit 20-40 percentage points below domestic acquiring rates for the same card and the same customer. Issuers decline cross-border attempts more aggressively, FX-flagged transactions get blocked, and 3DS flows behave inconsistently across local issuers.
• Regulated markets actively penalise cross-border flows. South Africa is the clearest example. Exchange control regulations there make it genuinely difficult for global PSPs to settle cross-border transactions at scale in ZAR, and the resulting decline rates reflect that. As well, South Africa is structurally a debit-order market for recurring billing, and global PSPs cannot access bank-debit recurring without local licensed infrastructure.

In each of these instances, the merchant might see numbers that look workable. However, the customer sees a checkout that doesn't work, doesn't work well, or doesn't exist at all.

The volume is in the methods you're not accepting

The opportunity for global merchants in African markets is not to abandon their global PSP, but to add a layer underneath that handles the markets and methods the global PSP was never built to handle. This should be done through a single integration rather than the unrealistic alternative of stitching together fifteen+ local providers market by market.

The infrastructure to make this work has to deliver several things at once:

• It has to process cards domestically in the markets where domestic acquiring beats cross-border by a wide margin. Egypt, South Africa, Nigeria, Ghana, and Kenya, who collectively comprise the bulk of African digital commerce, are markets where local processing produces authorisation rates 20-30 percentage points above cross-border for the same customer base, with the same cards.
• It has to support mobile money across more than twenty markets with the local execution detail to actually get transactions through. Each Mobile Money Operator (MMO) has its own phone-number formatting requirements, validation rules, transaction limits, and timing windows. One reason global PSPs fail mobile money at scale is that they aren't doing the format-normalisation, retry-timing, and amount-splitting work that local payments actually require.
• It has to handle bank rails, including, in South Africa, the recurring debit order infrastructure that defines the local subscription market. Streaming, gaming, SaaS, and creator platforms operating in the market either offer ZAR bank-debit recurring or they accept materially higher churn than their global average. CrissCross supports cross-border ZAR recurring bank debits via Stitch Group, utilizing our Treasury Outsourcing Company (TOC) license, which currently makes us the only payments provider operating this structure at scale.
• Importantly, it has to deliver all of this — domestic cards, mobile money, bank rails, recurring billing, refunds, and reconciliation — through a single API, a single contract, and a single settlement layer. The alternative is what most merchants have today: a global PSP for the easy markets and a half-built patchwork of local providers for everything else, generating the operational drag and reconciliation complexity that caps how much African revenue a finance team is willing to underwrite.

When merchants add this layer, and once locally relevant payment methods are turned on properly, they can achieve double-digit improvements in completed-transaction rates.

A bank wire is not a refund

We recently onboarded a merchant who had been using a major global PSP for collections in African markets. When customers needed refunds for mobile money payments, the PSP had no path to return funds to the original payment method because most mobile money operators don't natively support refunds. The rail itself doesn't allow value to reverse the way a card refund does, so the workaround landed on the merchant.

That merchant turned to international bank wires. Each wire cost them at least twenty dollars in banking fees and took up to two weeks to settle in the customer's account. They also arrived in a currency and on a rail the customer hadn't asked for. Each generated, on average, roughly twenty customer support tickets as the customer chased status, queried the FX, or asked why the refund hadn't appeared on their original payment method.

Multiply that across a few hundred refunds per month. This had a direct impact on cost-of-goods, support headcount, and customer-satisfaction scores trending in the wrong direction.

The underlying rail constraint is real and not going away in the near term. Most African mobile money networks were built for one-directional consumer-to-merchant flows, not for reversal. But the constraint should be absorbed by the infrastructure layer, not pushed onto the merchant. Where the underlying MMO doesn't support native refunds, the payments layer executes the refund as an outbound payout back to the customer's original wallet. The customer receives the funds in the same currency, on the same rail, in the same account they used to pay, and in minutes, not weeks.

From the customer's perspective, it looks and feels like a refund. From the merchant's perspective, it is a single API call. The underlying complexity, including payout rails, reconciliation against the original collection, and FX where relevant, sits inside the infrastructure rather than on the merchant's operations team. This then needs to work consistently across African markets with materially different underlying rails.

Two adjacent points are worth noting: mobile money transactions, once completed, do not get charged back. Stablecoin payments are final. For merchants who have absorbed the operational and financial cost of card chargebacks in markets with elevated fraud rates, the absence of chargeback exposure on a meaningful share of volume materially changes the unit economics of operating in the region.

Customers are choosing to pay in alternative currencies

The biggest shift in digital commerce across African markets over the past three years is that the customer's preferred currency is no longer always their domestic fiat.

Workers earning income from global platforms increasingly hold a meaningful portion of that income in USDT or USDC for a few reasons. This acts as a hedge against local currency volatility, as a more convenient unit of account for cross-border digital purchases, and as a deliberate financial strategy in markets where local currency has lost real value over recent years. The natural follow-on is that when those workers buy things online, a growing share of them prefer to pay in the currency they already hold.

A recent multi-country survey of stablecoin holders and prospective holders found that nearly 80% of Nigerian and South African respondents already own stablecoins, with more than 75% of those holders planning to increase their positions over the coming year. In Nigeria specifically, 95% of respondents said they would prefer to receive payments in stablecoins rather than in naira. Across emerging markets more broadly, intent to adopt stablecoins among non-holders is roughly twice as high as in high-income economies.

For a global merchant, a customer who holds USDT and can pay in USDT is a customer who converts more reliably than the same customer routed through a cross-border card. The transaction settles in minutes. There is no decline risk from an issuer who doesn't recognise the merchant. There is no FX uncertainty from the customer's perspective. There is no chargeback exposure from the merchant's perspective. The customer pays the way they prefer to pay; the merchant settles in their treasury currency of choice — USD, EUR, ZAR, NGN, or whatever they specify; the conversion happens cleanly in between.

What this requires is adaptive currency presentation across the entire checkout. A merchant can quote a price in USD; a Nigerian customer can pay the equivalent in naira via card, in USDT via wallet, or in mobile money via local network; the merchant settles in their preferred currency. None of this should be the merchant's problem to engineer. It should be what the payments layer does by default. And it should be done compliantly — which, in the African regulatory landscape, is a meaningful operational discipline rather than a checkbox.

How e-commerce, gaming and creator platforms can improve their payments stack

African digital commerce has outgrown the architecture global merchants were sold a decade ago. A global PSP is one layer of the payments stack. It needs another layer underneath it to reach the volume that's actually there.

A few principles for global merchants looking honestly at their Africa stack:

• Audit the gap, not the authorisation rate. The authorisation rate measures what happens to attempted transactions but not the transactions that were never attempted because the right payment method wasn't on the page. The right metric is the completed-purchase rate from intent to settlement, broken out by market.
• Process domestically where it matters. In Egypt, South Africa, Nigeria, Ghana, and Kenya, domestic card acquiring outperforms cross-border by a margin wide enough to justify the operational work to enable it.
• Match the methods to the market. Mobile money is not optional in East and West Africa. Debit order recurring is not optional in South Africa. Stablecoin acceptance is not optional for the segment of African consumers who increasingly transact in dollar-denominated wallets.
• Treat the operational layer as primary. Refunds, recurring payments, reconciliation, and the support load behind all three determine whether a merchant can actually operate a healthy business in a market vs just process some transactions.
• Choose a partner with people on the ground. African payment markets are operationally specific in ways that don't translate into documentation. Knowing which Nigerian issuer flags which transaction type, when a Kenyan rail performs best in the day, and how a South African debit order mandate gets structured to maximise authorisation are the details that determine whether an integration converts.

Merchants that are able to capture the growing volume in African markets are the ones that will win.

CrissCross provides local acquiring and payment infrastructure across 20+ African markets. To learn more, visit crisscross.money or get in touch with our team.