The Financial Times' Digital Business editor, Peter Whitehead, interviews David Elton of PA Consulting Group about how he came to the counter-intuitive conclusion that, for an IT project to work, it should be run as anything but an IT project.
David Elton says those who consider hearts and souls before numbers will do better when they embark on technology changes
Everyone knows IT projects don't work. Twenty years ago the Standish Group produced its shocking conclusion that 80 per cent of such projects failed.
Our own research, with Harvey Nash, showed little improvement - a 63 per cent failure rate. Over the past two decades pretty much every other piece of research has reached similar conclusions.
Not that anyone really needs data on this point. Who hasn't been involved in an IT project where frustration got the better of them and, head buried in hands, they wondered what it would take just to get it all to work?
The irony is that the reason IT projects fail is because they are IT projects. I worked with a client recently who was looking after a project and experiencing the usual delays and endless re-planning. Try as she did, she met resistance at every turn. She then came up with the obvious but brilliant idea: why not get the business to take the lead?
The effect was remarkable. The project became the new baby for the business. Previously insurmountable barriers became challenges to resolve. Vital requirements became the subject of innovative workarounds. On cutover day, the operations team came in early, relishing the prospect of helping the baby take its first steps.
Of course, IT was a big part of this project, but people played a bigger part.
No amount of technology will overcome resistance to something people don't want. Attitudes to IT projects are far from being a "right brain" rational calculation. Success is often about emotion - the "left brain" instinctive reaction
Make a compelling emotional case and most technical problems can be overcome. Do it the other way round and solutions are elusive.
Why should this be? After all, surely IT is not about people. It is about automation, illustrated in that classic picture of a data centre with the lights out and the machines left to get on with the work. In fact, IT doesn't eliminate people from the process. It changes what they do, usually allowing them to do more. So IT projects, far from eliminating people, are all about them.
There are plenty of failures that illustrate this basic truth. The first attempts to set up online retail businesses in the 1990s are good examples.
At the time, I was working with an airline that set itself a 50 per cent online sales target. It did not come close the first time round, but in 2002-2003 it tried again and achieved 70-80 per cent of sales via the internet.
The differences between the first and second attempts were about people. First, there was increased availability of secure payment methods that consumers trusted. Second, vendors started to understand that savings do not come from automation, but from enabling customers to do more for themselves. These were not internet projects, but customer enablement projects.
Another illustration of the importance of people in IT projects can be seen in a global call centre consolidation project I was involved in: it was based on worldwide call-flow technology. The macroeconomic case was compelling and board approval resounding. Regional managers were unimpressed.
However, behind the parochial territorialism, there were good reasons for modifying the proposal to improve it. A new, more rational version emerged, which still included some consolidation of operations and, critically, a shared view of how things would work. Costs of implementation fell.
I know from my own business, how important changing behaviour can be to obtaining value from technology. We recently implemented a new contact management system that aimed to create a more integrated approach to selling. The theory was sound and the technology effective but the biggest obstacle to securing the benefits was getting people to use the system.
When we understood this and started working on changing behaviour rather than implementing systems, we started to see the improved co-ordination of sales we had sought.
These lessons apply to just about every IT project I have worked on - changing the way people work, not just the technology, has to be at the core.
I've seen this idea in action when I worked with a CIO on a large upgrade and outsourcing project.
He started by defining the change he wanted to achieve. His questions were: what would the new way of working feel like? Will it feel better than today? Does it feel like it should generate value?
The key is in the verb, to "feel", the emotional, not the rational was the starting point. Of course, he then moved on to the numbers to validate the judgment, but they were not the starting point.
So what are the questions boards should ask when the CIO presents an IT project, to make sure they do not repeat others' mistakes. The first is, have you appointed the right leader for the project? That person needs to be an influencer who can change the way people think.
The second question is: are you clear about what you are doing and why?
If you want to change the way people work, then they will need to understand these things.
Finally, the board needs to ask whether there is a meeting of minds on the need for change and, if not, how it is going to win people over? Projects where everyone agrees are do-able. Projects where people disagree are a nightmare.
If there are credible answers to each of these questions, there is a good chance the project is heading out of the abstract world of IT, into the real world and so, towards a greater chance of success.
David Elton is an IT and change management specialist at PA Consulting Group and a member of PA's management team