Recently, there have been news reports of several write-offs at large IT household names. It seems we are trapped in a vicious cycle of failure, which is nothing new.
According to industry figures, the history books say that, on average, you should expect your IT programme to run 60% over budget, 60% over time and deliver only 60% of what you expected in the first place.
Additionally, the need to compete, together with requests from regulators and customers, means there is a higher demand for change at a faster pace. We could be complicating matters further by adding new capabilities for web, mobile and social media, but without fully removing the legacy platforms. In addition, the IT supply chain is longer and less controllable.
How failed IT impacts on businesses
Failed IT programmes have a significant impact, as a company's dependence on IT means that a failed programme equates to more than mere financial loss – it is a matter of business continuity.
One in 20 IT projects allegedly turn out so badly, they threaten the survival of the very business they are trying to serve. Social media ensures that failures can no longer be kept private, but are potentially exploited to the detriment of the company.
The point is, you can’t afford to let IT projects fail. And if you wait until you suspect a problem has arisen, it is probably too late to fix it. The causes of most failing programmes can often be found in the earliest stages – inadequate planning, bad decisions, or excessive zeal. Projects rarely come off the rails – they just weren’t on them to begin with.
For downloads on project management:
Test your programmes early on
So, how do you spot the duds at an early stage? The answer usually lies in being realistic about the ambition, the alignment of the stakeholders, or the understanding of the terrain. You need to ask the CEO's question: “Is my programme going to deliver, and what do I have to do to ensure that it does?”
Make sure you ask this question early on, and keep on asking it, and do not be afraid to sacrifice early progress for a level of confidence in the trajectory. That confidence may come from convincing the stakeholders of a programme's success, trying and proving the architecture, testing the development approach or stress-testing the benefits.
So, do your checks early on, then your project will get onto those rails from the outset. Don’t stop checking that you are still on them, as it will be a whole lot cheaper in the long run. If you wait until you suspect things are not going to plan, you risk having to just limit the damage.
Chris Saunders (pictured) is an IT expert at PA Consulting Group. For more information, visit www.paconsulting.com/it-delivery-assurance
This was first published in September 2014