If a platform appears cheaper because it has already been depreciated, but every change requires scarce specialists, manual reconciliation and another workaround, is it really the economical option?
In this recent article, Julian Farley, Sales Director for the UK and EU at BPC, an expert in the payments industry with nearly 25 years’ experience across solutions consulting, speaks to The Fintech Times on the economics of payments modernisation
Legacy technology rarely announces its retirement date. It continues processing transactions, reconciling files and keeping ATMs alive, which is precisely why banks postpone replacing it. The trouble is that “still working” and “still fit for purpose” are not the same thing. In payments, the gap between them has become a balance-sheet issue. Legacy infrastructure absorbs change budgets, slows product launches, fragments data and turns routine regulatory updates into bespoke engineering projects.
The hidden bill is often larger than the maintenance invoice. BPC’s Modernisation Without Disruption guide models a mid-sized issuer processing 10 million debit attempts each month. If avoidable false declines are only 0.50 percentage points, 50,000 legitimate transactions disappear. At an average transaction value of $30, that means $1.5 million of approved spend lost every month; for a European issuer, the guide takes optimistic numbers and estimates roughly $42,000 of annual interchange revenue lost before counting customer frustration, support costs or reduced card preference.
Europe has had repeated reminders that technology resilience is not theoretical. The UK Treasury Committee found at least 158 banking IT outages between January 2023 and February 2025, totalling more than 803 hours across major institutions. Barclays alone suffered an outage in which 56% of online payments failed and expected compensation costs of £5–7.5 million. In February 2025, the ECB’s TARGET Services suffered a major incident after storage hardware failed, suspending payment and settlement processing for hours. Not every outage is a legacy failure, but the ECB has warned that functional legacy systems can pose inherent risks to operational resilience and adaptability.
Modernisation changes that equation. Modular, API-led, real-time platforms allow components to be upgraded independently, connect more cleanly to wallets, instant-payment rails and digital channels, and scale without rebuilding the entire stack. Instead of paying repeatedly to modify ageing interfaces, institutions can configure products, expose standard APIs and add services such as virtual cards, tokenisation, BNPL or new authentication methods without reopening the core. Engineering capacity moves from keeping yesterday’s architecture alive to building tomorrow’s revenue.
BPC SmartVista illustrates this approach. Its open architecture is designed to accommodate different machine-learning engines, proprietary data-science pipelines and third-party AI models, allowing new AI services and autonomous agents to be introduced without rebuilding the platform. It also gives institutions deployment freedom across AWS, Oracle Cloud, Azure, private cloud and on-premise environments, with automated CI/CD, containerised releases and blue-green deployment supporting upgrades and rapid rollback without planned service interruption.
A bank can migrate by product, channel or customer cohort, run old and new platforms in parallel, reuse integrations and certifications, reconcile outcomes daily and retain rollback points until confidence is established. BPC’s guide describes phased or “box” migration as the lowest-risk route for large, complex estates. BPC has completed more than 300 migrations over 25 years, including replacements of platforms from ACI, Tieto, TSYS, OpenWay, HPS, Electra and FIS. Its proven “Pass Through” approach places SmartVista in front of the legacy environment, reuses existing integrations and certifications, and allows
modules to be migrated progressively without downtime for services already in production.
European examples show what this makes possible. Lithuania’s Artea Bank moved issuing to a cloud-native SaaS environment supporting Mastercard products, Apple Pay and Google Pay tokenisation and real-time fraud management. In Romania, Banca Transilvania replaced legacy card management and processing infrastructure; that foundation now supports more than 8 million cards and helped enable Romania’s first payment authenticated through the EU Digital Identity Wallet. In Bulgaria, DSK Bank moved more than 3 million debit and credit accounts to euro-denominated payments without interrupting customer service and has protected all of its channels from fraudulent activities, including reduction in false positives helping the bank to greatly reduce its incurred costs.
Fraud infrastructure deserves equal attention. Old rule engines can be overcautious and under-informed at the same time they block genuine customers because behaviour looks unfamiliar, yet miss fraud that crosses channels. Modern fraud platforms combine real-time transaction monitoring, behavioural profiling, machine learning, case management and cross-channel data so that risk can be assessed in context rather than through isolated rules.
Fraud is universal across the world and examples from other parts of the globe are relevant as well. Malaysia’s Co-opbank Pertama modernised digital-channel fraud controls with behavioural profiling, machine learning and real-time monitoring, while Colombia’s Banco Finandina combined 3DS 2.0 with advanced fraud management across online purchases, targeting stronger security alongside higher approval rates.
Financial institutions that delay will not merely look old-fashioned. They will keep paying for manual work, outages, false declines, fraud leakage, specialist skills and lost product opportunities while faster competitors compound their advantage. Standing still is not neutral. Every month of delay adds cost, widens the capability gap and makes the eventual migration harder.