Chaos continues at TSB as IT woes grind on

Digital business resilience is an area in which retail banking in the UK continues to be challenged, with adverse effects on its reputation, jointly and severally.  Because the risks and  complexity of major data migrations of this kind are very considerable, it is all the more surprising that this incident has occurred.

Was the data migration outsourced? If yes, to whom?

This stinks to me of some eejit manager pushing to get a migration done by using people unfamiliar with both systems (whether outsourced, local contractors or even permy staff) and making a specific timeframe as to when it will be completed.

One of the biggest problems in IT (specifically finance, but also other sectors) are unrealistic time lines with testing that is done using systems that are not an exact replication of what is actually in production.

There should always be a roll back process along with enough time to do it and significant testing post upgrade work.

Looks like the management decided to:

a) push it out too quickly

b) decided not to roll back despite several issues appearing or a conversion process that took longer than expected and therefore instead of calling a halt and roll-back, pushed on regardless.

Its very possible this was also outsourced and/or offshored - this kind of work usually is and unfortunately, this is the usual mess folks get into.

data migration to any new system is always a big risk and with muliple systems all interlinked further complicates any migration. For such fundamental errors to be found immediately post migration shows that TSB hadn't planned to have any staff testing post-migration before it was roled out to the public which is a significant management failure on effective planning.

This is a disaster of huge proportions. Even the RSB disaster of 2012 did not compromise Customer data. The fact that Data has been shared with other customers without their permission is a breach of som many rules and TSB are lucky GDPR is no in effect yet or they would be faced with a much larger fine than the one they are likely to get.

The mess that we can see and who knows what things we cannot see have all the hallmarks of a date driven project (most usually are) based on cost control rather than putting the Customer at the heart of the work. Testing must have been done really badly and presumably by people unfamiliar with the older systems. Typical these days where experienced people are not valued by the Banks they work for.

One has to expect these problems to happen more frequently as experienced people whi have worked in Banks all their lives leave and are replaced by outsourced staff with minimal experience and no experience of the BAnks home grown systems they now have to support.

With ICB, FCM, Open Banking and other regulatory work watch this space for more problems, hopefully not as bad as the TSB disaster but they will cause problems nonetheless.

Named account migrations just dont work. Naming conventions vary Mr S Smith, Steve Smith, Mr Steve Smith, Mr SF Smith. There is no consistent primary field. When will programme directors just realise the mark one human eyeball is the only way to mitigate risk in these enourmous data moves where straight data matching cannot take place. Even allocating a customer number to every account would work. Ive worked on three can you tell. All failed because of incorrect allocation to their new home.