Barclaycard’s website malfunctioned for most of last week, its second outage in a fortnight, leaving its customers unable to login and access their accounts.
Angry Barclaycard customers have contacted Computerworld UK, some saying they cannot carry out basic tasks on the site and others unable to see their financial information at all.
One customer said: "This is now becoming a total disgrace." Another stated: "I am appalled, not only by the fact that the site has been down for such a long time now, but also because Barclaycard can't even be bothered to post bulletins about the fault on their main site."
It is understood that the migration of 1.7 million Goldfish customers onto Barclaycard’s services, after the bank purchased the credit card business in February, is part of the problem. Barclaycard is said to be experiencing significantly higher volumes of customers each day trying to access their accounts online.
But the bank itself has still failed to provide any information to customers on the problem, except for a message all week on its site that stated: “We are currently experiencing high demand, which may affect the speed of the service. We apologise for any inconvenience this may cause.”
Barclaycard refused to say when the problem would be fixed, what caused the error and which suppliers’ equipment was involved. On Tuesday, Barclays said it had increased the capacity of its systems by 50 percent, but the problems are persisting.
The IT industry has reacted with astonishment that Barclaycard has not yet solved the issue, nearly two weeks after the problem emerged. Martin Stern, EMEA manager at web testing firm Keynote Systems, said: “On something as critical as your customers’ experience of the site, you just can’t mess around with that.”
Proper software testing, before changes of demands or processes on the site, is at the heart of avoiding these types of problem, he said. “Many companies just don’t test properly. They test the inside of the IT system. What they need to do is test each layer the customer goes through, including entering the site, finding the logic page, logging in, and carrying out transactions.”
It is important to conduct full load testing and then analyse the results, fix the problem, and test again before going live, said David Chalmers, director of product strategy at performance testing firm Macro 4. “This kind of testing and analysis is absolutely fundamental to knowing things work before they go live,” he said. “Businesses often test functionality and how things will look, but don’t simulate genuinely heavy traffic.”
Trevor La Fleche, senior research analyst at IDC, said it was “very rare” for a bank to allow these problems to persist for “more than a few days”. Serious but shorter problems had been experienced by Santander, for example, when it started to move Abbey’s business onto its in-house banking platform, called Partenon, he said.
Planning customer migration and testing systems as thoroughly as possible was vital, he agreed. But he added: “It is inevitable there will be some issues at banks that have made acquisitions and are migrating customers. Santander will be migrating Bradford & Bingley onto Partenon, and Lloyds TSB will be integrating HBOS’ systems.”