Recently I took an introductory class in Project Management. Much of the material was silly, or irrelevant, or obvious — but there were also fascinating ideas that struck me as worth remembering, maybe even pondering. Some quotes and fragments from my notes (with ^z asides in italicized parentheses):

• "This is a project management class. We will start on time! We will end on time!"
• Definition: "A project is a temporary endeavor undertaken to create a unique product, service, or result."
• "Risk is not optional — risk is mandatory!"
• There are five elements of a project: time, cost, quality, scope, and risk. (But as I see it, these really boil down to just three dimensions: cost (including time); scope (including quality); and risk.)
• "Hope is not a strategy."
• The official Project Management doctrine is to "deliver only what was agreed upon, and no more". (I disagree, profoundly; this contrasts with the Simply Scheme theme of radical empowerment of individuals to do unexpectedly great things; see ThinkingToolsGoals (9 Apr 1999), ...)
• "Availability is not a skill set." (i.e., don't just assign the next warm body available to a task)
• Productivity drops quickly when working extended periods of overtime. Taking an 8 hour day as the norm, the first week of 10 hour days gives ~95% relative efficiency (i.e., ~9.5 hours of equivalent work per day). The second week efficiency falls to ~85%, the third week it's ~75%, and after 4 weeks it's ~60%. Thus, by the third week less real work is getting done in 10 hours than was normally being done in a regular 8 hour workday. Overtime also increases burnout, turnover, illness, accidents, and errors. It takes ~2 weeks of normal hours to make up for 1 week of 10 hour/day storming.
• Context switching is expensive. Working on a single project, people may be able to achieve ~85% efficiency (the remaining 15% is "overhead"). Assign one person to two projects and each project gets only ~35% effective time (for a total of ~70%). Working three simultaneous projects, a person only can give ~20% to each.
• "Pad is Bad!" Don't use hidden fudge factors in schedule or cost. Instead, put in official engineering reserves, in blocks of a week at a time (for a multi-month task), spread across the whole duration of the project. If necessary (i.e., the organization resists) call these visible reserves "Quality Assurance Reviews" for project planning purposes. (Senior managers won't want to eliminate Quality! ... or at least they'll feel guilty doing so.)

The main lesson I learned from the course? Honesty is key to good project management. The prime directive should be to make things explicit — including assumptions, decisions, and their implications — so as to avoid unpleasant surprises, especially when working on large-scale, complex, collaborative endeavors.

Communication

Proper communication is a vital aspect of project management, yet it's often neglected or just taken for granted.

• Customer/client =X= management – (project specifications)
• Management =X= Team (lead)
• Team lead =X= team members
• Team member =X= team member
• Team =X= customer/client – (project deliverables, testing, ravisions)

... to itemize the major dialogues.

Many project failures can be attributed to communication breakdowns. Even only a few lapses can prove critical, leading to at best time/cost overrun. – Bo Leuf

_{(see also Better Faster Cheaper (29 May 1999), Common Understanding (8 Oct 1999), One Deep (15 Nov 1999), Big Lessons (17 Feb 2001), Project Management Proverbs (2 Jun 2002), Triple Think (25 Jul 2002), Motorcycle Maintenance (6 Jun 2003), ...)}

TopicOrganizations - TopicLife - TopicPersonalHistory - 2005-01-16

Comment 23 Jan 2005^ at 21:00Z

Quality/Project management courses that I've been on often refer to 3 elements which make up the "Quality Triangle": Time, Cost, and Features. The idea is that whenever you push (squeeze) one side of the triangle, you must make alterations to one or both of the other sides to compensate.

For example, if the Project Manager says that a project must be completed within the time estimate you need to affect Cost or Features. To do this you can do one of 2 things:

• Increase cost by employing additional resources or paying overtime
• Cut features to reduce the remaining tasks

- Darren Neimke -

(correlates: YouCanHaveItAll, Ninjas vs. Pirates, TruckNumber, ...)