Earlier this month, Paul Pester, the CEO of TSB Bank in the UK, spent an uncomfortable session in front of a Parliamentary Select Committee being questioned about the major failures in its core banking migration project back in April. The project had involved the migration of 1.3 billion accounts from the core banking platform that the bank had inherited on its breakaway from Lloyds Bank to Proteo4UK, a platform developed by its new Spanish parent company Sabadell Bank. The project had initially been deemed a success, with Sabadell putting out a congratulatory press release to this effect, but it quickly went sour as it emerged that millions of customers were unable to access online or mobile banking or make transactions, and in some cases it appeared that customers had been given access to others’ accounts.
With remediation taking several weeks, the immediate detrimental effects have been fairly evident. Paul Pester claimed that, to date, the bank has already spent £70m on compensation and additional resources to deal with the aftermath. The bank received close to 100,000 complaints, has already lost 12,500 customers, and is being investigated by the FCA (the UK’s financial regulatory body), with sizeable fines certainly not out of question (and Mr. Pester also has had to give up a £2m bonus tied to a successful migration). Longer term, the bank has clearly taken a significant reputation hit, and the impact on its customer-acquisition ambitions to establish it as a major challenger to the UK big banks is likely to be notable.
While the formal investigation into the migration failures are still ongoing, there are a number of major learnings for enterprises. Firstly, large-scale legacy modernization is rarely a simple project and carries major risks. Indeed, while there have been a few cases of successful “big bang” modernization projects with aims similar to TSB’s migration project (for example, Deutsche Postbank back in 2003), the preparatory run-up period generally stretches to several years. While TSB has stated that significant testing had been conducted and there was no pressure to migrate before they were ready, there were strong financial drivers for the bank to ensure migration occurred sooner rather than later. In contrast, most banks are adopting a progressive modernization approach where smaller pieces are changed over time or, as in the case of Nationwide Building Society, where both old and new systems are run in parallel for several years – a far more expensive, but ultimately less risky approach.
Secondly, it also highlights the importance of business as well as systems testing in legacy modernization. While it is clear, in hindsight, that the testing deployed by TSB and Sabis (the IT arm of Sabadell) was insufficient, TSB has continued to claim that testing was rigorous. Indeed, earlier in the year the bank claimed it had conducted over 80,000 tests in the run up to the migration. The challenge here is that the testing process has become highly automated in such projects. This allows a high degree of scale in terms of test numbers and should allow more comprehensive testing. However, it does mean that testing is largely conducted through IT and/or third parties rather than through business users. Systems may appear to work from an IT perspective, but users will have a better view on whether they are actually delivering correct business outcomes.
It also draws attention to the importance of enhancing data quality during systems migration. Testing may well confirm that data has been accurately migrated from one system to another, but if the data was erroneous, inconsistent, or incomplete in the first place you are merely transferring bad data from one system to another. A key part of legacy modernization is comprehensively understanding and cleaning up the previous system, its processes, and its data, before actually working on the new platform. Most enterprises have developed workarounds to deal with the limitations of legacy systems, and ensuring that the right data is fed into the new platform is as important as ensuring it has been migrated successfully.
Straight Talk is a weekly briefing from the desk of the Chief Research Officer. To receive this newsletter by email, please contact us.