Most change initiatives fail
Most of the research measuring project success and failure is from North America. While figures vary, ‘by most estimates, 50 to 70 per cent of all corporate change initiatives launched in the 1980s and 1990s failed to achieve their objectives’. Specifically, ‘business process re-engineering projects fail at 90 per cent’.
The Standish Group have been exploring the rate of software development project failure in the USA since 1994. While project success rate has been slowly improving – which they attribute to improved project management – the rate of failure remains alarmingly high. In 1998 they estimate that 74 per cent of projects failed. They group their figures into two types of project failure: ‘challenged’, ie projects that are completed and operational, but over budget, over the time estimate and with fewer features and functions than initially specified and ‘abandoned’, ie projects that were cancelled before completion. They subdivide the 1998 failed projects into 28 per cent abandoned and 46 per cent challenged. They reported a small improvement in a similar study in 2003 where they concluded that 66 per cent of projects did not successfully complete, with 15 per cent abandoned and 51 per cent challenged.
Project failure in the UK tends to mirror that in the USA. An equivalent report in the UK examined IT projects in the UK on four dimensions of project performance: variance against budget, variance against schedule, variance on scope/functionality and abandoned. This study concluded that 84 per cent of projects did not successfully complete, with nine per cent being abandoned and 75 per cent being challenged. However, the study also introduced a more ‘common sense’ measure for successful projects, what they referred to as ‘typical’, which included all the successfully completed projects together with those projects where the variance is so small they were statistically indistinguishable from the successes. The variance they found acceptable was when a project was, on average, less than five per cent behind time, just over five per cent under specification and less than four per cent over budget. The reports states that 55 per cent of projects fit this category, i.e. only 45 per cent of projects seriously fail. Whether you accept this additional ‘tolerance’ being added to projects or not, this ‘typical project’ measure does allow us to distinguish between those projects that are failing in a significant sense and those that just miss the mark.
This report went on to identify that project failure was greater the higher the budget, the longer the project; the larger the project to that normally tackled by organisation, and the greater the project challenge. Project challenge was ‘partitioned into complexity (measured as intensity of interaction with other systems and business processes), internal volatility (a measure of change within the organisation based on restarts, target changes, project manager changes and client changes), and contextual uncertainty (measured in terms of uncertainty and change to requirements)’. It is interesting to note in passing that the report also stated that an authoritarian style of project management leads to project failure.
What we are able to conclude from this analysis is that change initiatives, with their greater project challenge, are more likely to fail than infrastructure projects
So, although there is considerable uncertainty about the precise meaning of these figures and how they apply to change initiatives that have a high intangible component (cultural or ways of working change), it appears reasonable to conclude that at the very best 45 per cent of projects fail (i.e. nine out of every 20). For change initiatives, where the task is likely to be more complex and uncertain, the rate will probably be closer to the 84 per cent failure rate. This research supports the general estimate that 80 per cent of change initiatives (those projects involving a large degree of change management) fail.
What these figures imply is that within the 16% of successful projects there are people who know how to run change projects. However, if project success was random it would be reasonable to expect a 30% success rate. We, however, are achieving less than that. Of course, project success is not random, but the figures suggest that with all our growing expertise in project management, project success is still more a matter of luck than skill.
I would contend that something more is required to ensure a greater probability of success than just improving project management skills. I believe I have found that answer in the Onemind® approach to change.