TSB Bank's Failed Data Migration: Lessons on Testing, System Complexity, and Operational Risk

In 2018, the UK‑based TSB bank attempted a massive data migration, moving over 5.4 million customers' records from its legacy system to Banco Sabadell's Proteo4UK platform. The project, announced with great confidence by Banco Sabadell executives, involved more than 2,500 person‑years of effort and was scheduled for 18 months.

During the weekend of the migration, the old system was taken offline, and data was transferred to the new environment. Although the new system went live on Sunday night, within 20 minutes customers reported missing funds, incorrect balances, and even seeing other people's account information.

The failure was quickly noticed by the Financial Conduct Authority (FCA) and the Prudential Regulation Authority (PRA), leading to a series of investigations that identified a lack of rigorous testing as the root cause. IBM’s post‑mortem highlighted that the combination of new micro‑service usage, dual‑active data centres, and insufficient regression testing introduced multiple production risks.

Experts emphasized that modern banking IT systems are highly complex, interconnected, and must meet stringent availability targets (the “four‑nines” of 99.99% uptime). Without robust testing, change‑management, and redundancy, even well‑planned migrations can result in severe outages, financial loss, and reputational damage.

The incident underscored the importance of comprehensive test plans, risk‑based change control, and investment in resilient infrastructure. It also prompted regulators to issue guidance on operational resilience, urging banks to balance innovation with the stability of existing services.