Final thoughts on: why initiation is so important to IT

I have explained that I see many ways why a failure to initiate projects effectively is undermining the capability of technology to deliver the expected benefits. I now want to concentrate on ways to achieve a good initiation.

Initiation is really just the application of common-sense; unfortunately, any great time spent looking at IT projects will show you that common-sense is anything but common. But accepting that human frailty will always try to snatch defeat from the jaws of victory, I do believe that there are some simple techniques which, if more widely adopted, would enable the expected benefits of IT to be more regularly achieved. I believe that the heart of good initiation requires a simple, robust method, aided by nothing more hi-tech than a roll of brown paper, some blue-tac, and copious numbers of post-it notes.

Having been introduced to Goal Directed Project Management (GDPM) I am convinced that this is the single biggest improvement that we could make to IT. GDPM is the overarching Project Planning methodology that was adopted by Coopers & Lybrand – prior to them ending up as Global Services inside IBM – and I have been reliably informed it is used by NASA to plan all of their projects. Like all great methods it is very simple.

When most people plan a project they open up Microsoft Project, start at the beginning and build a plan bottom-up, from start to finish. Such plans are very daunting, and all too often it is impossible to see the wood for the trees. In GDPM the planning process is done in completely the opposite fashion. You start at the end and go back to the beginning; and you start at the highest level and drop down to detail.

The first thing that GDPM requires of you is to state what “success” is going to be. This is a surprising and refreshing difference to most project planning methods, which take it as read that everyone knows what the project is about and aim to get you started as quickly as possible. The truth is that most projects have only a vague notion of what “success” really is, and very rarely is that criteria effectively articulated. What we know is roughly where we are heading, and we doggedly set off in that general direction. As a result we have no clear success criteria and therefore the “end” is open to debate. In GDPM you state clearly when the end will be; i.e. when a, b, and c occur. As a consequence of doing this exercise on a recent project that I helped to plan, we discovered that in order to achieve success the project actually had to run for 3 or 4 months longer than had been assumed.

Having stated what success is, you then work back in stages identifying the milestones – the changes in state that can be seen by senior management and the end user of the solution that will contribute to that success. This is all done collaboratively with all the stakeholders contributing. This is where the brown paper and post-it notes fit in. You use brown paper on the wall as your plan, and you identify your milestones by writing a clear “milestone” statement, a “when x & y is achieved” statement, and they are stuck on the brown paper in relative time slots from the end to the beginning. Being collaborative, everyone’s ideas are shared, we gain a common understanding of what is involved, and we start to agree a common vocabulary to describe what is required.

The second stage is to then assign roles and responsibilities and define deliverables. For each milestone a deliverable has to be identified which will prove that the milestone has been achieved. The roles and responsibilities will identify who makes decisions, who does the work, who needs to be consulted to ensure the right task is done, and who needs to be informed of the outcome. It is only when all of this shaping of the project has been achieved that the activity plan is put together.

This is a simple yet rigorous process which, as far as I have seen, results in projects that have a far higher chance of success. I think that if more IT vendors advised their customers to use GDPM as an adjunct to their technology, they would have far happier customers.