I had been reflecting on a comment Geoff Rankins made at a previous PRINCE2 User Group Meeting regarding most IT projects not delivering the stated benefits. I have concluded that I agree. Most of the business cases for these projects are put up by the IT departments and vastly overstate the benefits to get those approved. Some even go to the extent of hiding the cost in operational budgets, so they don’t have to put up a business case in the first place! The consequence is those benefits can never really be measured or met. IT is not the core business of most organisations. I don’t think many of us in the IT profession have quite grasped it. It exists to provide an answer to a business question. Therefore, IT should not be putting up business cases. Their role should be to be the subject matter experts that advise the business in how they can enable capabilities.
From my perspective, delivering IT projects, there are two significant gaps in PRINCE2. One the text readily acknowledges is the lack of guidance around specialist areas like IT. It is interesting given the genesis of PRINCE2 was to control capital it projects in the UK. The second is in transitioning the outputs into the business. Of the cabinet office products, MSP has the best guidance around transition, but what if your project is not part of a programme? Many it projects aren’t.
Now let’s consider the environment in which the projects are usually delivered. Rarely will a project be delivered in a Greenfield environment. Most organizations today will either have an infrastructure or existing capability of some manner. They have a framework for delivering to their business. More and more the best practice framework used is ITIL.
The process that wraps around service design, transition and operation in ITIL is the continuous service improvement process. This is responsible for measurement of existing services to identify areas of improvement to deliver maximum value. This is the crucial link to connect a service management framework like ITIL to a project management framework like PRINCE2. The organization has already put a value to having this service available to them. Tying this improvement process to business cases ensures what makes sense technically also makes sense strategically.
In a real life example … a software that is used by an organization was about to become unsupported. It was flagged as something needing upgrade. But the organization was thinking about changing business processes that would no longer require that software. The IT team had moved the cost of this upgrade into operational budget, so they wouldn’t have to go through a business case for it! What was the consequence? They upgraded the system, delivered within budget and time, but discontinued it in less than 4 months! Was this a successful project?
Paradigm is shifting in the way IT delivers to the business. Cloud is becoming a realistic option for many services, applications and even infrastructure delivery. The advantage of this approach is a known risk pattern and the level of elasticity it gives you. The fact that you can scale to demand and not have to design for peaks, are good ingredients for success. Custom development therefore should be the exception, rather than the norm. It is somewhat ironic coming from someone that started his career as a software developer.
There will be times when you will have to build custom systems. Either due to legislation, or simply the problem is unique. In my experience, software development is part science … we’re building machine instructions, part art … designing user interfaces for example … and part voodoo. The problem the customer explains, the BA understands, the developer codes … are often coloured by their interpretation of the problem. It is the role of the project manager to ensure the language isn’t muddled. This is where I have found agile practices very useful.
Agile methods allow for quick feedback cycles – starting early and often. The biggest benefit I have found is stopping unnecessary features being developed – all developers are guilty of that, trust me. It also ensures what is delivered does not come as a surprise to the customer. Agile requires delivery of working software at the end of each iteration. Lining iteration boundaries coincide with management stage boundaries will ensure if it is ever determined that the project is unlikely to deliver the desired benefits and is cancelled; the customer is left with the latest working product.
The model I strived to establish in projects is Continuous Service Improvement forming the basis for startup activities, using the service design principles during initiation as you define you product description, use agile practices to build the product and use transition processes to deploy to the organization and finally as part of closure to establish the operational practices so a new cycle of service improvement activities can begin.
I want to end by sharing a statistic from Keith Ellis of IAG consulting … only 1 in 3 software projects make it to completion in the United States. That is the good bit. Of the ones that do, 85% are late and over budget. Traditional project centric thinking has a lot to answer for in IT project failures. We must tailor PRINCE2 in context of the delivery environment and project challenges. Frameworks like ITIL; practices like agile can help us achieve that.