Hard lessons for major IT programmes

Despite BT's own problems with the NHS' national programme for IT (NPfIT), Patrick O'Connell, the managing director of BT Health, offered attendees of the Smart Healthcare expo some advice about how to succeed with a major public sector IT programme.

Although he admitted BT's own NHS project "got off to a bit of a slow start," the programme has since rebounded, he claimed. O'Connell cited the N3 broadband rollout and the Spine database as examples of successful projects being delivered ahead of schedule by BT - discounting the slow start, of course.

The experience taught BT and O'Connell some lessons on large-scale project management.

The first was to realise that there are big differences between projects, programmes and major programmes, and it's not just about size, he said.

"It's not just a car of a different size, it's something different - as cars are to airplanes," said O'Connell.

Because of the extra complexity, major public programmes are difficult to complete without trouble - and government, media and citizen complaints.

He said any such large-scale project requires the right people organised in the right way. O'Connell said public sector programmes need experienced suppliers who know how to develop and implement technology as well as users who understand exactly how the IT is going to be used. But they also must be organised in an integrated way under the same umbrella, with clear, coherent leadership, O'Connell said.

Once the people are in place, they and their processes must be flexible. Planners should assume that technology will evolve during the development process and take advantage of it.

And managers must learn as they go and realise that as the project carries on, they'll gain a deeper understanding of what's needed and how to go about achieving it. "As they go along, they can change the process and improve things by another order of magnitude," he said.

He also warned against too much optimism. No matter how well-planned a project is, unexpected problems will occur. "Nobody can foresee everything over a 10-year period," he said.