The Waterfall development cycle has been the de facto methodology behind many software development life cycles for years. Over time, some companies have been able to make it work reasonably efficiently, others have moved to iterative development methodologies. Using these means, you can write software quickly, but what does this mean for the business?
Business has initiatives, not projects, and it prioritises these, develops business cases for them, and adds them to the development system 'to do' list. The problem comes in when priorities change, and initiatives that are no longer urgent remain at the top of the development 'to do' stack. Business constantly needs to review its initiatives and communicate its changing priorities to development in order to ensure resources are being applied where they are most important to the business. Demand management, in other words, says Ziaan Hattingh, MD of IndigoCube.
Once things get off the stack into development, they seem to develop a right to exist and are never killed, even if the business' priorities change and they become redundant. Eighteen months later, the project comes out of production and business says, 'but we cancelled that?'.
We need to constantly reassess to determine if we continue with a project; are the resources allocated sufficient given its priority? Organisations don't manage the interface between what the business needs from a strategic level and what development is delivering. Projects don't have a right to life. Sometimes they should be killed, or mothballed, or scaled down, so resources can be allocated to more urgent things.
Solving the communication problem requires closer collaboration between business and development, and even within development, where the business analysts can be working on version 1.3, the developers on 1.2 and the testers on 1.1. That way, business gets maximum strategic benefit from scarce resources, and developers don't end up working on projects for 18 months only to see them retired at go live, due to no fault on their part.
So, how many projects have you killed today?
Share