Monday, February 9, 2009

Sacred PM Practices -- Define Project Schedule

In preparing for battle I have always found that plans are useless, but planning is indispensable” is a well known quote from General Eisenhower. While I would not propose managing a project is the same as ridding the world of tyranny, his sentiment strongly suggests that we should pay careful attention to planning. If a project manager is to be successful, he or she should take great care in planning. Eisenhower didn’t mention the use of any single planning approach like the PERT Estimate, or Simpson’s Rule, but surely a planning methodology is necessary to build a project schedule.



  • None of the projects reported a formality (e.g., Global Efficiency Factor, Productivity Adjustment Percentage, PERT Estimate) in scheduling

  • Project managers for 64% of the projects had to make all interim milestones fit a final milestone

This, by the way, was the most shocking discovery of my research project… not a single project used a formal estimating approach to define its project schedule. Others have found project management methodologies to be unnecessary for some types of projects (see We Don’t Need No PM Methodology by Pawel Brodzinski).

I tend to use one of two planning approaches. For unfamiliar types of projects, I will work with my technical leads using something like PERT Estimate. For types of projects where I am familiar with the activities to be performed, I just lean on my own judgment (a.k.a. relative comparison). It is quicker.

What about you? How much do you depend on formal methods of project planning?


  1. Can you tell a bit more about the research which brought results you present? What kind of people/companies were involved? How big was a sample?

    I'd say that to some point results are something I can feel in my guts, although I'm surprised there's virtually none of examples of formal estimating techniques.

  2. I introduced the study in a blog posting on Jan 25, 2009, but it seems sample size didn’t make the final edit.

    I decided to study 25 projects initially, and then increase the sample size to 50-200. After about the 12th project, I kept on hearing the same observations. Given the high effort of scheduling and conducting the interviews, I haven’t added to the sample size yet. I hesitated to publish any study with such a small sample size, but I learned from the findings and thought others might benefit from it too.

    The types of organizations where the projects were executed can be described by the following:
    • Federal Government – 7
    • Manufacturing – 6
    • Insurance – 3
    • Publishing – 3
    • Telecom – 2
    • Travel – 1
    • Consulting – 1
    • Healthcare – 1
    • Software – 1

    The types of projects in this initial sample are categorized as the following.
    • Customization of software product – 12
    • Software implementation & business changes – 6
    • Software development – 4
    • Data Warehouse & Knowledge Management – 1
    • Incremental delivery of business data – 1
    • Redesign of corporate website – 1

    I hope this helps.

  3. Thanks for additional details. It brings more insight to your analysis.