iStock

TSB IT meltdown has the makings of an epic

TSB’s very public IT problems will send shivers down the spine of IT teams at large banks that are yet to migrate to new core banking systems

This article can also be found in the Premium Editorial Download: Computer Weekly: Making music with AI

TSB has experienced the difficulties of moving from age-old trusted legacy core banking systems to a brand new platform designed for the digital banking that today’s customers want – and all this in public.

Over the past week, TSB moved millions of customer accounts from the systems of Lloyds Bank, which has hosted them since TSB was separated from Lloyds, to a new core banking platform from its current owner, Spanish bank Sabadell.

As a result of yet-unspecified problems, over a five-day period, customers were locked out of their accounts and experienced money disappearing from accounts. Some were even able to see other customers’ accounts.

TSB customers are angry, regulators and government are now involved, and banks that are yet to migrate from legacy systems are probably very worried.

Proteo4UK, as TSB’s new core banking system is known, is a UK-specific version of Sabadell’s existing core banking system, Proteo. If all goes to plan, Proteo4UK will support TSB in the digital banking age and enable it to challenge bigger banks by offering fintech services like some of the UK’s digital challenger banks do.

TSB also said the move to Proteo4UK would cut its costs by £160m a year. It currently pays Lloyds Bank a couple of hundred million pounds a year for the service.

But the small matter of moving millions of existing customers from legacy systems to the new platform is the task that has put banks off modernising for years. TSB has bitten the bullet – but did it bite off more than it could chew?

When Sabadell acquired TSB in 2015, it said it would move customers to a new banking platform, and has been planning the implementation ever since. The plan was to move customers across in stages, and Proteo4UK was rolled out to the bank’s staff in November 2017 with a full range of banking services.

The organisation had originally planned to move all customers at the same time, but delayed the changeover until the first quarter of this year, citing the first interest rate rise for years as the reason for the delay. It didn’t want the migration to coincide with the rate rise to avoid potential customer confusion.

But on 20 April, the migration began. Customers were told to expect reduced access to services during that period, but what has unfolded has taken on epic proportions.

Online and mobile banking glitches happen regularly, and they follow a pattern. They happen, there is a small eruption on Twitter where customers vent their anger, the company apologises and promises that it is working hard to get things working. Then the problem is fixed, followed by another apology and sometimes an excuse that most consumers don’t understand and that IT experts doubt was the cause.

But the TSB case will be different. You have to go back to 2012 to find an outage of this scale played out in public. Back then, when the banks’ CA-7 batch process scheduler froze, customers of RBS, NatWest and Ulster Bank could not access their funds for a week or more as the banks manually updated account balances.

RBS was fined ₤56m by regulators for the failure and CA Technologies, which supplied RBS with the software that caused the outage, paid the bank millions of pounds. Apparently, someone in an RBS operation in India had simply pressed the wrong button.

Another major IT problem that went public has echoes of TSB’s crisis. Back in 2007, when another Spanish bank, Santander, integrated customers of its UK acquisition, Abbey, to its in-house platform Partenon, services were disrupted. The bank said only a small number of customers were affected, but staff said at the time that they were under pressure because of difficulties with the platform roll-out.

TSB’s problems seem much larger than those experienced by Santander during the Abbey integration – but that could just be the result of more news and social media coverage.

Twitter explosion

One big difference between Santander’s problems in integrating Abbey to its platform compared with TSB today is the power of social media. Today, if a banking IT meltdown were divided into phases, the Twitter explosion would be one of them.

The TSB meltdown has sparked hundreds of negative Tweets. These are a few examples:

“Disabled daughter left unable to pay rent, etc. Branch wouldn’t help, now 5th day without money.”

“Yes, it’s getting ridiculous now. This was all supposed to be finished with on Sunday. It’s now Tuesday and still no info. Can’t even log in to my internet banking. Keeps coming up wrong log in details even though I’ve checked and double-checked and it’s the right ones.”

“I will definitely be leaving #TSB, going to be opening a new bank account elsewhere on Friday. I have been banking with #TSB since I was seven years old. So disappointed.”

This is not good PR for a bank that is attempting to reinvent itself as a digitally savvy challenger.

Horrendous task

It is hardly surprising that banks still use mainframes that date back 40 years. Which CIO wants to put his or her reputation on the line dealing with a core banking platform migration?

One IT professional who worked in IT at Lloyds Bank before, during and after the original integration of TSB, said: “Migrating to a new core banking system is a horrendous task and there are so many thousands of things that have to happen exactly as expected, or disaster can unfold.”

The professional said Lloyds Bank’s main banking systems were very well built back in the day when quality was as high a priority as cost or time. “Thinking about legacy system migrations and new systems, I would rate the legacy systems as the lowest risk and safest place to be as they have generally been proven over many years, if not decades,” he said.

“New systems are more risky and migrations are extremely risky as well as difficult. The level of care required to migrate a bank onto different systems is perhaps almost uneconomic to perform in a way that keeps the risk small enough.”

Commenting on Sabadell’s decision to move to a new in-house core banking platform, the professional said: “In a world where banks are often trying to outsource IT or buy it as a service, I am always surprised when a firm attempts to do the opposite.”

He said the fee to use the Lloyds Bank systems will probably look very cheap to TSB when this matter blows over. “This issue arises when banks buy, merge or split with other banks, so there is quite a lot of it out there,” he said.

“Migration problems are inevitable, but the trick is to have a plan B, plan C and so on to give yourself the best chance of preventing a mistake becoming the end of the firm. I would consider major migrations as a last resort as they have the potential to kill a bank financially, and destroy its reputation with the regulators as well as customers.”

Potential options for replacing bank legacy systems

  • Forget changing systems and try to remove complexity. This is what often happens when the people making the decisions are near retirement or can’t stomach a multi-year, multibillion-pound project.
  • Buy a modern core banking platform off the shelf, get it working, connect it and migrate everything from the legacy systems onto it.
  • Acquire one of the growing number of new banks with their state-of-the-art IT, and eventually move the whole bank onto these modern systems, which can be tailored to the bank’s needs.
  • Spend money on a state-of-the-art system and make it pay by acquiring other banks and moving them to the platform.
  • Artificial intelligence (AI) could solve complexity issues.

TSB CEO Paul Pester issued a statement apologising to customers. He said: “I’ve just resurfaced after 48 hours with my teams who have been working as hard and fast as they can to get our services back up and running. This isn’t the level of service that we pride ourselves on providing, and isn’t what our customers have come to expect from TSB, and for that I’m truly sorry.

“We’re still seeing issues with access to our digital services. One of the steps we need to take to resolve this is to take our mobile app and online banking down for a few hours. We did this at around 10.30am (Tuesday 24 May) and we hope to be back up later this afternoon. We’ll let customers know as soon as it’s available again.

“Of course, customers can rest assured that no one will be left out of pocket as a result of these service issues.”

TSB reported that its systems were back up and running normally on Wednesday 25 April, five days after the problems began. But at the time of writing, some customers were still reporting issues.

Customers need answers

But Nicky Morgan MP, chair of the Treasury Committee, said that “warm words and platitudes will not suffice. And apologies will also not save the bank from the regulators.

Both the Financial Conduct Authority (FCA) and the Information Commissioner’s Office (ICO) are on its case. An FCA investigation is inevitable when a bank has problems that affect customers in such a way, and the fact that some customers could access other people’s accounts brings the ICO into play.

In a letter to Pester, Morgan demanded to know what had gone wrong, the extent of the failure, and how TSB intends to compensate customers who have suffered a breach of potentially highly-sensitive personal data.

“The reports of unauthorised transactions, access to other customers’ accounts and failures of in-branch services have all the hallmarks of an IT meltdown,” she wrote. “This is yet another addition to the litany of failures of banking IT systems. Potentially millions of customers could be affected by uncertainty and disruption.

“It simply isn’t good enough to expose customers to IT failures, including delays in paying bills and an inability to access their own money.”

Morgan asked for details on what actually went wrong. Banks normally don’t give full details of what has gone wrong, which often raises questions over whether they really know. “TSB customers deserve to know what has happened, when normal services will resume, and how they can expect to be compensated.”

So, from promise to disaster. TSB has been lauding its IT plans over the last couple of years, promoting them as a differentiator.

Perhaps the most worrying thing for the industry is the fact that all the big banks will have to go through a migration process at some point. TSB might be suffering now, but in a year’s time, most people might have forgotten about it, apart from the big banks stretching out their legacy systems for another few years.

They will will have watched their competitor suffer with mixed feelings – they know it might happen to them too one day.

Read more on IT for financial services

Join the conversation

3 comments

Send me notifications when other members comment.

By submitting you agree to receive email from TechTarget and its partners. If you reside outside of the United States, you consent to having your personal data transferred to and processed in the United States. Privacy

Please create a username to comment.

You have used past tense unfortunately. I have two accounts with TSB and can not access my money (Wednesday afternoon). So it looks the saga continuous!
Cancel
Thanks for the update. When I filled the article, which is an analysis, TSB said everything was back. I shall do a news story soon updating that problems still exist.
Cancel

I don’t think agile or DevOps adoption are responsible for the issues taking place at TSB. From my understanding, the issue is that the timescales for implementation were unrealistic and the deadline was not able to be moved for reasons of cost. When this takes place, it’s always easy to look back with hindsight at where problems started. However, you run the risk of sunk cost fallacies affecting decision making – we have already spent this much and committed so much effort here, so we have to continue rather than backing out.

 

From my experience, phased migrations that allow teams to build out a data layer between the customer and the back-end are more likely to help banks succeed. This approach allows banks to continue serving customers with information requests which are typically 80% of customer interactions – how much money is my account, say – while the core banking system can be changed. Rather than a big bang approach, this allows banks to test with a smaller group of customers and check that things are working properly in parallel. This means that banks don’t have to run the risk of an all or nothing migration and run the huge risk of failure.

 

Other banks have had these migration issues recently as they have had to comply with industry changing regulations like PSD2. This phased approach, based on keeping things consistent for the customer experience, has help them successfully change over to these new environments.

 

I’d say that TSB has taken on some of the biggest amount of risk in its migration. However, all banks will have to make changes in their infrastructure over time. Either they have to carry on running mainframes with legacy applications, or they too will have to migrate in the future. At this point, the TSB example will be one that all banks will learn from. In these circumstances, newer approaches like DevOps can  complement more traditional migration planning, as part of a phased roll-out. The important element here is that development and testing phases have to be respected. You can’t ignore the crunch phase in deployment and expect to be lucky.

Cancel

-ADS BY GOOGLE

SearchCIO

SearchSecurity

SearchNetworking

SearchDataCenter

SearchDataManagement

Close