Friday, July 30, 2004

Project Management Methodology

Due to the interviews I am now doing, and the fact that we eschewed formal management certifications at my previous workplace (due to a perceived low ROI), I have been looking for ways to describe how we managed projects to recruitment consultants - who are usually just looking for the right buzzwords (and quite rightly too, after all they can't be experts at everything). Unfortunately, I can't just say I'm PMP or PRINCE2 certified or that we based everything on PMBOK*.

What we did do was spend the last five years hammering out our proprietary methodology, based upon our collective Price Waterhouse Urwick and Andersen Consulting/Accenture experiences.

“Analysing this showed us that in many cases a task that had a 'generous' estimate given to it was not completed under budget, but in fact took up the whole of the time allocated. ”
In the last three years, we used Earned Value Management (EVM) methods to control our projects, which increased our ability to predict and control schedule and budget variances. We also played with Extreme Programming (XP) and other agile processes, but never approached them seriously.

We noticed that despite EVM's precision, we still did not always get the desired result. Projects still went over budget with alarming frequency (although we rarely went over the scheduled delivery date). Analysing this showed us that in many cases a task that had a 'generous' estimate given to it was not completed under budget, but in fact took up the whole of the time allocated.

Further analysis showed that this was a combination of people fitting other unplanned (and often nonchargeable) tasks into the time required for their project tasks, and people simply coasting or 'spinning their wheels' unnecessarily (e.g. over-engineering the solution because they had more time than required for the original design).

“Allowing buffer at the task level simply encourages people to use that buffer unnecessarily.”
The answer we were looking for, and had started to move towards, was Critical Chain Project Management, which removes buffer from individual tasks and allocates it to the project as a whole, thus concentrating resources on the task at hand.

Of course we had yet to perfect this method, but we had already recognised that:

  • People work better when they are not multi-tasking, and that the less time (not effort) we scheduled a project for, the less overruns we had. This also reduces issues with people working for multiple project managers, because they are more likely to be on one project at a time.

  • Project managers need to communicate better in order to ensure that people working across programs are available for the critical tasks whenever they might be able to start them - this required flexibility and honest appraisals of task criticality (sometimes the criticality to our company of that task/project would outweigh the criticality of another task that was more critical but for a less critical project).

  • Allowing buffer at the task level simply encourages people to use that buffer unnecessarily. It is much better to slim down the estimates they work to, and put buffer at the overall project level.

  • Buffering the scheduled delivery date will ensure that it is met, even if Murphy's Law hits the project.

    Where we still needed to improve was in the way we calculated the size of and tracked the use of buffers. However, it is interesting to see how our proprietary methodology mirrored recent changes in PMBOK (i.e. the introduction of EVM in the 2000 edition), and even included ideas from new methods such as Critical Chain Project Management without any of us attaining formal project management certifications.

    * You could say we did follow PMBOK, because the term is defined inclusively:
    “The Project Management Body Of Knowledge (PMBOK) is an inclusive term that describes the sum of knowledge within the profession of project management. As with other professions such as law, medicine and accounting, the body of knowledge rests with the practitioners and academics that apply and advance it.”
    PMBOK Guide - 2000 Edition
  • No comments:

    Post a comment