Hefty bills for failed projects

New Zealand organisations are losing tens of millions of dollars a year through failed software projects, a software tool maker claims.

Borland New Zealand bases the figure on the rate of failure observed by researchers in the United States.

Chris Gray, Borland Australia and New Zealand sales director, said the Massachusetts-based Standish Group recorded two-thirds of software projects as failures. Another US research group, NIST, put the cost of project failure in that country at US$60 billion ($84 billion), Gray said.

If anything, New Zealand had a worse record than the US, said Gray, describing the typical approach to projects here as "a cottage industry".

"In New Zealand we have some of the best talent in the world working on these projects but they're not doing it in any managed business process."

While development methodologies abounded, said Gray, projects often came to grief because of lack of integration between the efforts of those involved. "There's a big difference between having a methodology and having a repeatable, managed business process that gives you predictability and control."

Gray would not back his claim of widespread failure by identifying specific projects.

"I'm certainly not going to mention any individual names because that's not my role but if you are involved in the software development industry, or even as a taxpayer, you don't have to look very far to see some high-profile failures. They're continuing."

The rate of failure is not disputed by professional project managers. Michael Ryan, who chairs the Auckland branch of the Project Management Institute, agreed the industry's record is "dismaying".

But he pointed the finger at a different cause than Gray. "I believe in the Standish report the number one reason [for failure] is believed to be lack of executive sponsorship. That's a communication issue as well but it's not communication between members of the project team."

Ryan calls project management "the hardest job in the world" - delivering the "highest quality" results to tight time frames at the least cost, while bringing together a disparate group of stakeholders.

"It's a difficult problem. One of the objectives we have in the institute is to turn business on to professional project management. There are a hell of a lot of project managers in the industry who aren't affiliated to the institute and who don't hold a [project management professional] qualification." "

Gray's - and Borland's - prescription is a tool, Core SDP, designed to get those involved in the development process working together.

"You've got lots of different skills involved in the software development process. The various roles speak different languages, almost; they certainly use different tools. There's no communication, very little visibility and certainly not a lot of collaboration when you step through the lifecycle.

"Up until now you have the different roles buy a different product. They get lots of functionality that they need and lots that they don't need. With Core SDP what we've done is carve out the particular pieces of functionality required over the whole lifecycle needed to meet each role," said Gray.

"It's a starting point in terms of what Borland is going to do to provide a managed business process which gives predictability and control around software development."

Haydn Easte, Auckland chairman of the Computer Society and a project manager at the University of Auckland, said lack of deep understanding between the different roles was the key issue. He said there were already tools available designed to do what Core SDP was billed as being.