T843 TMA01 article from 1998
Bank learns from computer failure
Peter Roney, chief executive of retail and investment bank Save and Prosper, has disclosed lessons learnt from his companies failure to go live with a multi-million pound unit trust system, Fast2000.
The technology is a key part of the company's strategic development of systems, believed to be costing between 20m and 35m. Co-written with US software specialist DST, Fast2000 was due to be delivered last year but is now expected to go live in two stages in February and July 1997.
It is rare for organisations to admit they have had serious computer problems, and occasionally the subject is taboo even within the company, but this can make it difficult for staff or the computer industry to learn any lessons from the disaster.
Roney is less interested in internal politics than in making sure the mistakes of Fast2000 are never repeated. 'One of the major problems we have had with Fast2000 is interfacing it with our corporate systems' said Roney. 'We probably underestimated the project's complexity and therefore DST and ourselves admit that it is something that we didn't get right.'
He added that save and Prosper has also learnt a lesson well known to IT disaster victims such as the London Stock Exchange, London Ambulance Service and Performing Right Society.
'It is necessary,' said Roney, 'to break down a large project into very small modules and it is ideal if you can have relatively short timescales for the development as a whole.'
He added, 'Fast2000 is a project which has been running for three years. It is a very big project and it has been made much bigger by the need to interface with our systems. That's been the course of the difficulty. We're approaching a new frontier in terms of what the company is trying to do and with what our systems are trying to do.'
'It sounds easy with hindsight to say that we should have done things differently but clearly we should have, because we have had a time over-run on the project.'
On how best to approach the future, Roney said: 'I think the lesson for us going forward is trying to keep IT projects to manageable sizes and if you have a large project, break it down and have six-month milestones where you can post the achievement and you can see the progress of the project.'
He continued: 'Clearly if you get a project of this size over this length of time, one of the key issues is change control. The user requirements change over three years and therefore it is not easy to tie the project down and this is what has happened. Ideally a project should be over in one or two years.'
'It is unrealistic, certainly in the financial services world, to expect the user to be static. If you are building IT solutions, you have got to deliver them fairly quickly according to specification if you are not going to have significant change control problems. We have learnt that lesson and we will apply it more stringently in the future,' Roney concluded.
Source: Computer Weekly, 21 November 1996, page 24
Home page | Pauline's Pages | Howto Articles | Uniquely NZ | Small Firms | Search
Most recent revision:
17th May, 1999