By Colin Ellis
Sept. 2, 2016
Plenty has been written about the recent Census technology debacle – much of it has focused on the security of Australian systems and our ability to protect ourselves and our personal information from outside hackers.
Mainstream and business articles analysing the subject tended to focus on the inadequacy of performance testing, limited capacity, how it’s almost impossible to prevent denial of service (DDOS) attacks and the sums of money paid to the suppliers to get the job done.
However, once again, the person/department sponsoring the IT project and the person/department managing the IT project get off scot-free and face little in the way of scrutiny.
By now, our confidence is now so low in those who have the accountability and responsibility to get IT projects done in the right way, that we expect them to do an adequate job at best and look for other (usually outside/external) excuses when major projects fail.
Indeed, there are many people in the project management profession, including the industry bodies that are meant to represent it, who will go to great lengths to draw up lists of reasons why IT projects fail.
Some of the excuses we hear include:
- Scope creep
- Poor resource management
- Lack of governance
- Lack of visibility of status
- Poor project selection process
- Lack of prioritisation
- Poor risk management
- Lack of senior management support
- Lack of strategic alignment
- Inadequate testing
- Not learning lessons from the past
IT project managers offer up every one of these excuses every time a project fails.
But let’s be crystal clear about this. There are actually only two reasons for project failure: poor project sponsorship and poor project management.
Peter Shergold said as much in his report on the continuing failure of IT projects in the Australian Public Service in 2015.
Peter Achterstraat said it in his white paper on failed NSW projects in 2013, and the Victorian Ombudsman said so in their report into the major Victorian IT project failures in 2011. Yet the malady lingers on.
IT projects will continue to fail because IT project managers and IT project sponsors lack the basics skills to deliver them:
The plans that we build for IT projects, just aren’t good enough. In the case of the Census project, all of the partners (internal and external) should have been working together – and sometimes that means in the same room – to build the best, most comprehensive plan that they can. One that evolves over time and ensured that adequate security measures were taken and there was enough time for testing. This doesn’t seem to have happened, given the level of blame that we’re seeing.