The allure of artificial intelligence (AI) may be captivating boardroom discussions across industries, but for many payments companies, a more fundamental challenge demands immediate attention: transforming vast troves of financial data into actionable insights that finance teams can genuinely trust. This critical theme anchored the latest installment of the PYMNTS Summer School series, a recurring educational program designed to equip finance professionals with cutting-edge knowledge. The session featured insights from Ben Catterall, global head of solutions engineering at Fynapse, a prominent finance Enterprise Resource Planning (ERP) solutions provider. Catterall emphasized that true strategic advantage in the payments sector will not stem from merely accumulating more data than competitors, but rather from building robust financial foundations that meticulously preserve the integrity and meaning behind every transactional event.
"Every payment company has tons of data," Catterall stated during the PYMNTS Summer School session, highlighting a ubiquitous reality. "But data without context is not particularly useful." He elaborated on the imperative to understand not just what the data represents, but crucially, "what happened in the real world to generate that data." This nuanced perspective underscores a growing realization within the finance community: raw data, irrespective of its volume, holds limited value without proper contextualization and validation.
The complexity of modern payment ecosystems exacerbates this challenge. Each payment method – from traditional cards and burgeoning digital wallets to the increasingly popular buy now, pay later (BNPL) schemes, recurring subscriptions, and app store purchases – generates distinct financial records. The global nature of commerce further complicates matters, introducing multiple currencies, settlement variations, and a multitude of payment gateways. This intricate web of transactions, coupled with ever-climbing transaction volumes, presents a formidable reconciliation and accounting hurdle.
For decades, legacy finance systems, designed for a pre-cloud era of batch processing and summarized data, have struggled to keep pace with this escalating complexity. These systems often operate on a delayed basis, providing a retrospective view rather than a real-time understanding of financial operations. The consequences of this data deficiency extend far beyond mere operational inefficiencies.
Catterall recounted a pertinent example of a multinational payments client that processed transactions across nearly 18 countries. While the company’s consolidated financial statements appeared balanced at a high level, a deeper dive into transaction-level records revealed a significant issue: foreign exchange spreads that, on average, amounted to approximately 2%. This seemingly small percentage translated into millions of dollars in lost revenue, impacting roughly $100 million in cross-border payment volume annually. "If you apply that to $100 million of cross-border payments being processed, that could be $2 million a year that’s lost," Catterall calculated, underscoring the profound impact of granular data oversight.
This revelation serves as a stark reminder for finance leaders that transaction-level visibility is not merely an accounting exercise; it is a strategic imperative. It provides the essential lens through which to identify where revenue is being eroded, where payment costs are disproportionately accumulating, and where targeted operational adjustments can yield tangible improvements in financial performance.
Building the Foundation Before Deploying AI
The same principle of foundational data integrity applies directly to the discourse surrounding AI adoption in finance. While many financial institutions have enthusiastically launched AI proofs-of-concept over the past two years, a relatively small fraction have successfully transitioned these initiatives into full production. Catterall identified the underlying obstacle as frequently residing not within the AI models themselves, but in the quality and structure of the data upon which they are trained.
"If you’re relying on batched, aggregated, summarized data, and you put an AI tool on top of that, all that AI tool can learn from is the summary view," Catterall explained. "It doesn’t have enough to go on." This limitation effectively hobbles AI’s potential, preventing it from uncovering nuanced patterns or providing truly predictive insights. Research corroborates this sentiment, with Catterall noting that a significant 46% of AI proofs of acceptance fail to reach production due to poor data quality that undermines their effectiveness.
This philosophy directly informs the design of Aptitude’s platform, which is engineered to capture financial events as they occur, rather than attempting to reconstruct them retrospectively at the close of a reporting period. The primary objective is not merely to expedite the month-end close process, but to empower finance teams with continuous, real-time visibility into margins, payment costs, and overall business performance. This proactive insight enables finance professionals to drive informed, real-time business decisions.
Catterall advocated for a paradigm shift, urging finance organizations to treat financial data as a core infrastructural asset, rather than a mere byproduct of payment processing activities. He emphasized, "Making that available to the business is a differentiator." Companies that meticulously preserve detailed transaction records and make this granular information accessible across treasury, pricing, forecasting, and risk management functions are inherently better positioned to support growth initiatives without compromising financial control. This is the essence of achieving "financial truth" for payments, a capability that Fynapse delivers to a diverse range of clients, including telecommunications giant T-Mobile. The latter now processes an astonishing 200 million journal lines per hour in real time, a testament to the scalability and efficacy of a modernized data architecture.
As the landscape of payment methods continues to diversify and AI assumes an increasingly significant operational role, this discipline of robust data management may well emerge as finance’s most enduring competitive advantage. The firms that proactively invest in modernizing their finance data architecture today will be far better equipped to comprehend, rather than simply record, the complexities of tomorrow’s transactions.
What Finance-Grade Data Unlocks
A modernized approach to financial data fundamentally entails three critical shifts in perspective and practice. Firstly, the data must reflect individual transactions, eschewing the limitations of aggregated totals. This granular view allows for deep dives into specific events, rather than providing only a high-level overview.
Secondly, the detailed attributes behind each transaction must be meticulously preserved and accessible. This includes vital information such as the payment method employed, the currency used, any associated fees or charges, and the parties involved in the transaction. This comprehensive metadata is crucial for accurate analysis, reconciliation, and dispute resolution.
Finally, and perhaps most critically in today’s fast-paced business environment, this rich transactional data must be available immediately. It should not be a resource that is only pieced together later during a laborious close process. Real-time accessibility empowers finance teams to act decisively and proactively, rather than reactively.
The implications of embracing such a finance-grade data approach are far-reaching. For instance, in the realm of e-commerce, granular data can illuminate the precise reasons behind cart abandonment, allowing businesses to optimize checkout processes and marketing campaigns. In subscription services, it can reveal churn drivers by analyzing payment failures, subscription tier changes, and customer engagement patterns. For BNPL providers, detailed transaction data is indispensable for accurate risk assessment, fraud detection, and compliance with evolving regulatory frameworks.
Furthermore, the integration of AI on a foundation of clean, contextualized data can unlock transformative capabilities. AI algorithms can leverage this rich data to predict future cash flows with greater accuracy, identify anomalous transactions indicative of fraud, personalize customer offers based on spending habits, and automate complex reconciliation processes that currently consume significant human resources. The potential for AI to drive operational efficiency, enhance customer experiences, and improve financial forecasting is directly proportional to the quality of the data it consumes.
The PYMNTS Summer School series, by bringing together industry leaders and providing a platform for knowledge sharing, aims to bridge the gap between the aspirational goals of AI adoption and the practical realities of data management in the payments sector. The emphasis on building a solid data foundation before embarking on advanced technological implementations is a recurring theme, underscoring its foundational importance for any organization seeking to thrive in the digital economy.
The complete PYMNTS Summer School interview with Ben Catterall offers a deeper dive into these critical topics, providing actionable strategies and case studies for finance professionals looking to navigate the complexities of modern financial data. As the payments landscape continues its rapid evolution, the ability to trust and leverage financial data will undoubtedly be a defining characteristic of successful and resilient organizations. The journey towards AI-powered finance begins not with the algorithms, but with the data – meticulously crafted, thoroughly understood, and readily accessible.
