Secrets of your software success

Might a more flexible, ad hoc approach shine a ray of hope for IT software delivery? Fleur Doidge reports

Creating new software is no use to business unless it meets needs as well as budgets and time frames

With many business software delivery projects still being cancelled or seen as failures, it can be argued that – despite improvements – deliverers of new applications must raise their game still further around project management.

Julian Holmes, co-founder at IT project consultancy UPMentors, said current research indicates that about 45 per cent of software delivery projects fail. Furthermore, he said, if you ask different questions and seek to find out whether the stakeholders themselves felt the project was a success, the failure rate tends to be far higher – 65-70 per cent.

“So the big question is how do you perceive success?” Holmes said. “Is it about meeting dates and sticking to a budget, or is it about actually delivering the value required by a business?”

A project might be delivered on time and under budget, but if the software is no longer needed, or the company has evolved in another direction during the course of the project, the project as a whole cannot possibly deliver. The operation may be ‘successful’, even if the patient dies. And that makes for unhappy customers.

Holmes believes that the answer lies not in ditching best-practice techniques and methodologies around PRINCE2 and the like but in reassessing them with fresh eyes to get IT project management to a point where it can, for example, change the goals of the project ‘on the fly’ over time, based on new facts as they arise.

The worst way to begin a software project is to define all the requirements – because they will almost inevitably change, sometimes beyond recognition, he said.

This is by no means easy, Holmes suggested, but it is probably the only way if projects – and the IT industry as a whole – are not to continue to suffer from disrepute as a result of frequently disappointing customers one way or another.

“People start out by saying, ‘we are looking for this’, or ‘we need this’, and ‘we expect it is going to cost this much’. They start down some sort of path -- and as they progress, they realise maybe that is not what they really needed. Maybe what they really want is not what they thought they needed, or things change,” said Holmes.

UPMentors, formed in 2007, aims to help organisations such as Capgemini, ING Direct, and HMRC to deliver and cope with complex software projects, combining consultancy with specific practices such as knowledge transfer, mentoring and training techniques. It has specialisms in software development processes such as Agile and Unified Process.

“With my clients, I say, ‘you have got to do budgets, you have to get competitive timelines, and you want to make sure you get the most value possible from your spend as quickly as possible’,” he said.

“I am trying to encourage a slightly different mentality in how they drive their projects.”

According to Holmes, although the situation around IT project failures has improved, traditional project management roles remain under threat unless the standard of software delivery projects can be raised. This means work on getting the right culture, education, and collaboration is desperately needed for projects to truly succeed.

This doesn’t mean, however, a rush to the next latest ‘greatest’ certification. For some, Holmes said, while its iterative and incremental approach to software projects is a welcome advance, Agile has become a religion. Users need to see past the rhetoric if Agile is not to be just another restriction on truly nimble project management.

And of course, IT needs to better understand its customers’ businesses, from many perspectives, Holmes noted.

The Standish Group released its first Chaos Report on IT project failure in 1994, finding that 31.1 per cent of IT projects were cancelled before completion, with 52.7 per cent costing 189 per cent more than estimated. Only 16.2 per cent were completed on time and within budget – and in larger companies, only nine per cent could say the same.

“One of the biggest reasons bridges come in on time, on budget and do not fall down is because of the extreme detail of design. The design is frozen and the contractor has little flexibility in changing the specifications. However, a frozen design does not accommodate changes in business practices,” the Group stated.

Its latest Chaos Report in 2009 found that 32 per cent of IT software projects are now considered successful. However, project failures have been increasing again through the recession as budgets tightened, according to the Boston based analyst.

Magic conjures up four iBOLT partnerswww.channelweb.co.uk/2268631