Buffers one or all?

(This POST was submitted by Michael McGinty an expert consultant on Critical Thinking and Project Management.)

Eliyahu Goldratt is credited as the author of the Theory of Constraints – a powerful concept which created something of a revolution in process improvement.  He applied these insights to project management in his book Critical Chain.  Here at BPI, we share some of his key insights with project leaders when they attend our one-day workshop in Systematic Project Management.

Goldratt teaches us that one of the best ways to ensure projects don’t violate their deadlines is to re-think our treatment of buffers.  What people typically do, unconsciously, is they build in a buffer for every action step in their project.  For example, let’s say that based on our previous experience, we know that a particular task usually takes about 3 hours to complete.  However, when we create our project plan, most of us will allocate more time to complete that task – you know, just to cover our bases.  So, the task that usually only takes us three hours to complete might instead be scheduled to take five hours.  And we do this for every action step in our plan, which stretches out the project timeline.  We unconsciously set our project deadlines based on a ‘worst-case scenario’ with these frozen buffers.

Goldratt teaches there’s a way to guard against this but without putting your plan at risk.  He says to schedule the average time it takes for each action step, and then build your buffer into the back end of the project.  So, a task that typically takes three hours is scheduled for exactly three hours.  The additional two-hour buffer is built into the tail-end of the project.

You might ask, “Why is that any better?  You’re still allotting five hours for that step!”  That’s true, but the location of the buffer is crucial.  When buffers are built into each step, the time is not available to help other steps.  And, if something actually does get done on time, other people/resources are not ready because they didn’t plan for it to get done “early.”  By scheduling the average time for each action step and collecting the buffers in one place, other people can be prepared for a quick hand-off to the next task plus they can dip into the common project buffer when needed – that’s what it’s there for!

In this way, Goldratt’s techniques are helping companies complete projects in what seems like ‘record-time’, 50% less time or better is not unusual!  But for those of us in-the-know, it’s just ‘average time’!