For merchant acquirers across Africa, fraud prevention has long focused on minimising chargebacks, reducing scheme exposure, and protecting merchants from losses. But there is another threat hiding inside many fraud strategies: false positives.
Legitimate transactions wrongly declined as fraud are not just a customer experience issue – they are a major revenue drain. In Africa, where digital payment adoption is accelerating but trust in electronic payments is still being built, an unnecessary decline can do more than block a sale. It can push a customer back to cash, an alternative provider, or another merchant altogether.
The scale of the opportunity makes this especially important. According to GSMA, mobile money accounted for more than $2 trillion in global transaction value in 2025, with Sub-Saharan Africa continuing to sit at the centre of mobile money adoption and usage.
For merchant acquirers, that means fraud strategy is not just about protection. It is also about enabling growth.
When fraud controls suppress growth
Now don’t get us wrong, fraud management is essential. Excessive fraud leads to chargebacks, operational costs, and scheme scrutiny. Across African markets, acquirers also have to manage a complex mix of risks across cards, mobile money, bank transfers, wallets, USSD, instant payment rails, and cross-border transactions. The pressure to tighten controls is understandable. In South Africa, for example, SABRIC reported that card-not-present transactions made up 85.6% of gross fraud losses on South African-issued credit cards in 2024.
But overly conservative fraud controls can quietly suppress merchant revenue. Static risk rules, rigid thresholds, and blunt controls often err on the side of caution, declining legitimate customers alongside fraudsters.
This is particularly important in African markets, where payment behaviour does not always fit the assumptions of card-led fraud models. A diaspora buyer, a cross-border shopper, a shared device, or a customer switching channels after a failed payment can all trigger risk signals. Under rigid rules, those signals may look suspicious. In context, they may be entirely legitimate.
In competitive digital markets, customers do not always abandon immediately after a failed payment. They often try an alternative card or payment method first. But when legitimate transactions are repeatedly declined – particularly due to merchant- or acquirer-side controls – many will abandon the purchase altogether or switch to another merchant.
For subscription services, digital goods providers, marketplaces, travel merchants, and cross-border retailers, even a small increase in false decline rates can significantly impact conversion and customer lifetime value and translate into major revenue losses. In more severe cases, acquirer-side controls may escalate beyond individual transaction declines to restricting or blocking a merchant altogether based on perceived risk – temporarily preventing payment acceptance and amplifying the revenue impact.
Take a merchant processing 100,000 transactions a month with an average order value of $80. If unnecessary declines reduce approval rates by just one percentage point, that means 1,000 additional lost sales – or $80,000 in lost monthly revenue. Over a year, that equates to nearly $1 million lost, before accounting for customer churn or lost lifetime value.
In this environment, fraud strategy becomes conversion strategy.
Moving beyond static rules
Many fraud systems still rely heavily on static rules and predefined risk signals. But modern African commerce is far more complex. Acquirers need tools that can interpret behaviour across payment methods, markets, devices and merchant types, rather than applying the same threshold to every transaction.
Forward-looking acquirers are addressing this by taking a hybrid approach to prevention and detection, combining machine learning with adaptive risk rules.
Machine learning models analyse behavioural signals across large transaction datasets to better distinguish legitimate customers from fraudsters. Rather than preventing every fraudulent transaction outright, they enable faster and more accurate identification of suspicious activity – allowing acquirers to take targeted action and limit further exposure from specific merchants or emerging fraud patterns. This improves overall accuracy and reduces unnecessary declines.
As a result of using this AI-driven detection, some organisations have reported reductions in false positives of more than 80%. When layered with dynamic risk scoring, risk rules become more precise – helping block high-risk transactions while allowing legitimate payments to proceed.
From fraud prevention to revenue protection
For merchant acquirers in the region, this represents a strategic shift. Management around it can no longer be judged solely by rate reduction. It must also be measured by its impact on approval rates, customer experience, and merchant revenue.
False positives may not generate chargeback disputes, but they quietly erode revenue, weaken customer loyalty, and limit growth.
Effective strategies are not about eliminating risk entirely, but about balancing it. Overly conservative controls can suppress legitimate revenue, while more adaptive approaches accept that some fraud may occur – but prioritise rapid identification and targeted response, particularly in cases of repeated or concentrated fraud activity at the merchant level.
Stopping fraud is necessary, but stopping legitimate customers from paying is far more expensive.
