The Green Sheet Online Edition
May 11, 2026 • 26:05:01
AI can close the hidden revenue gap in payments
The payments industry has long measured success in authorization rates, processing volumes and interchange economics. But for the small and midsize businesses that form the core of the merchant base, a different set of numbers tells a more troubling story, one that rarely surfaces on any single report, and one the industry has been slow to address.New research from Access PaySuite, drawing on surveys of hundreds of management and finance professionals at SMEs in the UK, quantifies the gap in striking terms.
Nearly half of respondents (49 percent) are losing between $6,000 and $125,000 annually to failed transactions, payment-related customer churn, and the administrative burden those failures generate. Around 8 percent report annual losses exceeding $1.25 million. On average, 3.4 percent of transactions fail, and 55.8 percent of those failures are never recovered.
For payment service providers, ISOs, and merchant acquirers, those numbers represent more than a merchant pain point. They represent a structural gap in the value that the payments ecosystem delivers to its customers—and a significant opportunity for providers willing to address it.
Where the money goes
Tony Craddock, director general of The Payments Association, framed the underlying problem as one of fragmentation. "It's a whole series of typically quite small failures," he said. "Because all of these small little pieces are in different functions within the company, the size of the overall problem is often unknown. It's almost like a hidden loss of revenue."
The mechanics will be familiar to payments professionals. Authorization failures at the issuer level, checkout abandonment at the merchant level, chargebacks, refunds, and silent subscription churn each contribute to revenue leakage, but rarely trigger a coordinated response. Nearly half of businesses surveyed reported checkout abandonment, with an average rate of 7.8 percent. More than one in five reported losing customers to competitors offering a smoother payment experience.
Sandra Mianda, founder and CEO of Paypr.work, identified a cultural dimension to the problem that compounds the technical one. "Traditionally, payments have been seen as a cost center," she said. "The KPIs tracked are fees and approvals. But somewhere between intent and settlement, declines can have many different reasons that account for a lot more than what the data shows. There's a real hidden opportunity in those failed transactions."
The payments value chain itself amplifies the damage. Issuers decline transactions without full context. Acquirers and card schemes apply independent risk frameworks. Merchants manage refunds, disputes, and chargebacks through disconnected back-office systems with no unified view of cumulative impact.
Chris Jones, managing director of PSE Consulting, identified the opportunity squarely: "If you can move them into a single user experience and allow those exceptions to be handled in a much shorter activity focused on edge cases, that's transformative," he said.
For ISOs and agents working with SMB merchants, this fragmentation creates a difficult conversation. Merchants see transaction fees on their monthly statements; they rarely see the compounding cost of failed payments, churned customers, and staff hours consumed by reconciliation. The result is a value proposition that undersells what better payments infrastructure could actually deliver.
The operational toll
The revenue gap carries a companion problem: the administrative cost of managing failures at scale. More than 70 percent of organizations surveyed spend between five and 20 hours per week on payment failure management and related administration. Fewer than four in ten have full visibility into the broader revenue impact of those problems across their business.
Jones quantifies the burden for smaller merchants directly: three to four days per month in reconciliation between different financial systems. "If you ask small businesses what causes them pain," he said, "it's those exceptions of uncollected payments." For a business owner managing operations alongside finance, that is time that cannot be spent on customer acquisition, product delivery, or growth.
The administrative drag is not limited to retail. In organizations dependent on recurring revenue—subscription services, membership organizations, charities reliant on regular giving—failed payments generate follow-up calls, lapsed agreements, and difficult conversations with customers who may not know anything went wrong. The technology built to collect payments, as one senior finance professional observed, is "geared towards bringing payments in, rather than actually focusing on those failures
AI as the recovery layer
The industry is responding. Ninety-five percent of the SMEs surveyed are actively exploring AI-based systems to identify and recover hidden revenue, a figure that signals significant demand for technology-led solutions from payment partners. For providers evaluating how to serve the SMB segment more effectively, the implications are concrete.
David Birch, an international adviser on digital financial services, described the core value of AI in payment optimization with characteristic directness. "It will find patterns. It will uncover connections," he said. "It will spot trends that you wouldn't necessarily see yourself. You only need to take decline rates down by a small amount to add a lot."
The practical applications are already emerging across the industry. AI-driven retry logic can determine dynamically whether a failed transaction should be retried, rerouted through an alternative acquirer, or flagged for a different authentication approach, decisions that today are largely rule-based or handled manually.
Predictive models can identify subscription customers at risk of lapsing before their next payment fails. Natural language interfaces can allow non-technical users to query payment performance data without relying on analysts or custom reporting tools.
Mianda cautioned that realizing the potential requires a strong data foundation. "The data is one thing," she said. "But being able to interpret what that story tells you in your environment and process that in a way that drives intelligent decisions, that's the next step." Without clean, consolidated data, AI tools produce noise rather than insight, a lesson that points to the importance of integrated platforms over point solutions. The shift toward integrated payment intelligence is already shaping new product development across the industry. For smaller businesses in particular, the goal is not simply automation but visibility, turning fragmented payment data into actionable insight that can improve recovery rates and reduce the operational effort required to manage failures.
Craddock pointed to a broader set of applications: AI-assisted journey design, smarter fraud prevention that reduces false declines at source, and feedback loops that improve authorization rates upstream. Over time, he argued, programmability and micro-payment capabilities may reshape the revenue models available to businesses operating on digital payment rails.
Jones raised a note of caution alongside the optimism. AI-driven commerce creates new dependencies on the platforms that control customer relationships, and merchants that benefit from AI-generated demand must weigh data ownership and disintermediation risk carefully. The technology reshapes commercial relationships as well as operational ones.
For the payments industry, the opportunity is clear. SMB merchants are carrying a significant and largely invisible revenue loss, spending time they cannot afford on problems they cannot fully see.
The businesses best placed to capture this opportunity are those that can bring integrated data, intelligent automation, and a unified platform view to a merchant base that has managed fragmentation for far too long, helping merchants not only process payments, but to understand and optimize them.
SideNote:Failed payments create hidden CX problem
When payments fail, the damage extends beyond lost revenue. Failed transactions erode customer trust and loyalty in ways that are difficult to measure but increasingly important.
Today's consumers expect subscriptions to renew automatically, ecommerce checkouts to work instantly and stored payment credentials to function without interruption. When a payment fails, especially without clear communication, the customer experience can deteriorate quickly.
In subscription businesses, failed payments may lead to canceled services, interrupted memberships or declined access to digital platforms. In ecommerce, checkout friction and payment declines can drive customers directly to competitors. For small businesses, losing an existing customer can be significantly more expensive than recovering a failed payment.
Also, risk systems designed to stop fraud sometimes block legitimate transactions, frustrating customers who may never attempt the purchase again. Because these declines often occur between issuers, acquirers and fraud tools, merchants may not fully understand why conversion rates are suffering.
This is one reason why improving authorization rates, reducing false declines and streamlining recovery processes can directly affect customer retention and lifetime value.
AI tools may help address the problem by identifying patterns behind failed transactions and automating retry logic or routing decisions. But merchants also need better visibility into where failures occur and how payment friction affects the broader customer journey. Payment performance is increasingly becoming part of the customer experience itself, not just a financial process operating in the background.
Jonathan Reynolds is head of product at Access PaySuite. To learn more about how AI-driven payments insights can help uncover hidden revenue and reduce payment failures, please visit https://www.accesspaysuite.com/the-ai-advantage-closing-the-hidden-revenue-gap/. To contact Jonathan via LinkedIn, see https://www.linkedin.com/in/jonrreynolds.
Notice to readers: These are archived articles. Contact information, links and other details may be out of date. We regret any inconvenience.
