I am often brought in to run a project some time after it has been started. The typical situation is that an initiative has been started by a senior executive. The executive has usually made promises based on the success of initiative: when it will be finished, what will be achieved and so forth. A vague set of tasks or objectives has been agreed across this executive’s management team. The managers individually get on with their own tasks: but there is no plan or only a very high level one, there is no co-ordination of activity, and the managers pursue their individual elements at their own pace. Worse, each manager has interpreted the original instruction and aims in their own way.
The situation can be summarised as this: a project manager having to take over a project in that is being run by no one. (I do not mean the situation in which a project that has been set up and run by another project manager. That is a different issue).
I have developed a way of handling these sorts of situations. I do not however claim this is perfect. Like many aspects of project management I have developed it by trial and error, and I know there may be many people out there who have faced similar challenges. So I am writing this article as much as possible to generate debate as I am to claim to have some great wisdom in this domain!
How I handle these situations
The temptation facing an unstructured piece of work can be to run in all guns blazing and try to enforce structure quickly on to what is an unstructured piece of work. The most obvious way of doing this is to stop the project completely. Then follow the normal project management steps: writing a project brief, understanding objectives and scope, developing a plan and so on. In some cases, this is required, but generally I try to avoid doing this. I avoid it for 2 reasons:
- It stops what progress is being made. Poor progress is better than none!
- Rightly or wrongly it is often politically unacceptable to be seen to stop the project.
Therefore I need to build a project management structure around the work that is ongoing, rather than starting from scratch.
The mindset I adopt is to think in terms of a pair of lines. One line represents what the project should be doing; the other line represents what the project is doing. My aim is to make these two lines converge, so at some point in the future this uncoordinated initiative, without ever stopping, evolves into the disciplined project needed.
To achieve this I go through the following steps:
- Objectives, scoping and expectation setting
- Short term control
- Tactical control
- Strategic control
Objectives, scoping and expectation setting
The problem with projects of this nature is that it is often unclear what the project is trying to deliver and what has happened so far. I want to get to a position in which I understand the objectives and scope of the project very clearly, and the project team and sponsor have realistic expectations about what I can deliver.
I do this by interviewing a range of people involved in the project. This can be anything between 6 and 20 people. These are typically hour long meetings when we discuss where we are, what the outcomes are, problems and issues and their personal perspectives on the project. I run as many of these as I need, stopping when I start to hear broadly the same information from the people I interview.
I try to run these purely as information collection. I make no changes to the project and I try to avoid setting too many expectations. I explain that I will bring structure and direction to the project, but only when I understand it enough. I stress I am not a miracle worker. Any good project manager can add value to an unstructured and complex piece of work with a better planning and discipline. But you cannot convert the impossible into the possible.
One of the interviews has to be with the executive running the initiative. What is critically important is to understand how flexible the goals are. There are various aspects to the goals – I’m most interested in time and outcomes expected. Usually, the delivery date cannot slip, because it has been promised. But what has been promised at this date is less fixed. It is common to end up with a two phase project:
- Phase 1: delivering something for the originally agreed date. This is the political deliverable.
- Phase 2: delivering what is really required at some later date. This is the real deliverable.
The outcome from these interviews is my understanding of the project and the context. I develop a simple statement of objectives and the scope. I do not worry if this is not 100% perfect. It is a baseline for everyone to comment on, before I produce an agreed set of objectives and scope definition.
Short term control
Having developed a scope I now start to bring some control to the project. The main problem with an in flight project you do not have the luxury of planning before anyone works. You have to plan in parallel with delivering the project. My aims in the first 2-4 weeks are to make sure that:
- Whatever activity that is being undertaken is at least moving the project towards the desired objectives. I am not at this stage concerned whether it is doing it in the best way, or right order, but I want to stop anything that is directly contrary to achieving the objectives or is outside of scope.
- Get the people working on the project used to working with a project manager and to using basic disciplines and reporting.
- This is also the time at which I want to start warning the sponsor of the project if I think the goals are achievable. I generally do not know for certain yet, but I will at least start to have a feeling. If my feeling is not positive, now is the time to start managing expectations more assertively.
In the next month of so I want to bring the project to a level of tactical control. In this phase my concerns are to:
- Build a better and fuller plan. The most important thing is that the plan is realistic.
- Set realistic planning timescales and expectations. Develop an unrealistic plan and you will soon be adrift from it and confidence in you as a project manager will decline.
- Assess for myself the key risks facing the project. What can go wrong and how is it best dealt with?
- Start get the team into the habit of directing their activity and measuring progress around the plan
- Make the team understand that the project management process adds value.
One way to get the team working as a project team is to give them some quick wins. Usually in an uncoordinated initiative there are some issues which have been floating around which are not complex to resolve, they just require bringing some people together and agreeing a common way forward. I try to find a few of these, even if they are not priority concerns. Resolving them builds faith in the project management process.
Over the following 3 months I want to take strategic control. My aim at the end of this phase is to have a fully functioning project that appears to anyone outside as if it has always been working as a project.
I want everyone working as a fully disciplined project team. One of the best levers to do this is the simplest. Every week progress is reported against the plan. People are asked to explain variances from the plan. This simple, traditional project management approach tends to develop a much greater focus on the initiative as a project. This is most effective if the sponsor actively participates in some of these sessions, at least to start off with.
In parallel with this I start involving the team in the risk management processes. I want to move the team beyond thinking about today’s work to looking forward to what is coming up and what may get in the way.
The other thing I want to bring in is better project communications. It is usually a sign of a well controlled project when there is a disciplined, consistent and co-ordinated set of communications to all project stakeholders.
This is a very high level view of how I approach this situation. I don’t think there is a universal answer
as it depends on the context and the state of the project. Typically, when I take over a project which no one has been running before, I adopt a gradualist approach to intervention as a project manager. Occasionally, I have to step in straightaway and stop everything, but usually this is politically counterproductive.
The timescales I have given here are indicative and based around a large programme which will run for 12-18 months at least. For a smaller project, strategic control has to come much more quickly.