the Monday Musings column from that consultant bloke

Almost anything that you change can be a project, so from simple things within the office to nationwide rollouts I have seen a lot. Some I have been on the receiving end for, in others I have been part of the roll out team and for some I have been the sponsor. Not all have gone well, so let us have a look at why.

Small projects are only easier because the scope and scale is smaller. They are much more controllable because you can immediately see if anything is not going to work as planned and you can quickly fix it, but they can still fall victim to flaws in design or being poorly thought out, through a lack of preparation or either failing to follow the plan or not have a plan at all..

As you move up the scale the need to be sure about what the objective is, how you will recognise success, to plan, test that planning, revise as necessary and test it again becomes more vital. Almost every project that I have not had full control of has run into the fundamental problem of running late simply because it wasn’t thought through properly in the first place and that grows exponentially as the project gets larger.

Major projects over-run in most cases because the first thing that gets inked in is the Go-Live Date (GLD). Then they start on the design and the planning and, whilst it is usually obvious in the latter stage that there is no way that you can get it done by the GLD no-one will change that date to something more realistic. We plough blindly on and fail.

But working too a silly end date is not the only problem in a really big project. The other key killer is that things change over the life of the project so that what you end up with is either not quite what you need by the time that you finish, or that you finish even later because you change the specification along the way. This has been classically demonstrated by so many public sector projects over the years, many of them being abandoned before completion.

At the concept stage of a project you need to have an understanding of what you are trying to achieve and by when. If you can be clear about that you should also be able to build in some risk factors around things that you know might be prone to change so that you can design some change controls. you also need to understand what can practically be achieved in what time; that will allow you to plan resources and to get a realistic schedule worked out. The schedule will have to have contingency because there will be things that foil even the best plan and then you can start to see when your GLD will fall. If that is beyond the desired GLD then you can look at reducing the scope, but you should avoid looking at speeding the implementation up because that will almost always prove to result in a failure.

A problem for many projects is that the people that design them often fail to understand the impact that their plans will have. In theory they will work, but a lack of understanding of real life on the front line. Churchill once said something to the effect of you should stop admiring a beautiful strategy and occasionally consider the results. I have seen many projects that look great on paper that fail miserably because the planners had no concept of anything beyond their ideas as translated onto spreadsheet and project management planning software.

One of my old bosses was a man who liked the latter style of project management. He went on to plan a national distribution network change centered on a new facility near the M1/M6 interchange. On the first night there was a queue almost a mile long on the motorways and again on the second night. The Police were not happy and changes were forced upon us as the magnificent plan sank without trace and a more practical one was implemented by the operational team who had the hands on experience to make things work.

One crucial aspect of planning any project is to make sure that it can actually deliver the desired results and you can only do that by having people on board who understand how it can be made to work. And then you have to test the plan. And test it again as many times as you need to to be sure that it will work.