Critical Chain Project Management
Length: 1119 words (3.2 double-spaced pages)
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Almost all projects fail in terms of timings and costs, but this is not because we don’t plan carefully enough, but rather because we plan to fail. Main reasons for wrong planning are:
1. We plan assuming our project will not change its scope, but it always does. You will never find a project where the scope never changed during the project phase. In the best case, the change was minor, but usually changes are significant enough to impact your project. The question is: if we know the plan will change, why do we plan assuming it will not?
2. Tasks end after the deadline or in the best case the last day: A project is made of tasks, and the probability of tasks ending before the deadline is zero, which means that the probability of a project ending by or before the deadline is also zero. This is due to 2 main factors:
a. Student’s syndrome: as students do, if I know I have enough time I’ll start my tasks after the project’s start of that tasks (i.e. if I have 10 days to perform a task but I know I can do it in 5 days, I’ll probably start day 3, 4 or 5, which means I’ll eliminate the safety stock from my tasks and may, in the best case, finish that task the last day (if nothing happens)
b. The last 20% is the slowest: when you look at how people advance in their project they usually get very fast to the 80% completion of a task but take much more time for the last mile. When people start early and get fast to the “80%”, they tend to relax or if they know that they may finish earlier they spend more time than needed improving details that are not really needed or just wait until the last day to end their tasks.
Additionally people don’t tend to finish earlier than expected because when timings for tasks are negotiated, managers push for less time and employees to have a Safety Stock as they don’t want to be punished for being late. If it happens that they finish a task earlier than planned they would be giving the message to their managers that they could have finished earlier and they won’t receive extra-time in their next negotiation”. Although one might think “this won’t happen to me and my project”, one should just look at the facts.
When estimating the duration for a task, if statistically right, 50% of the times tasks would end before the deadline and 50% after it. Does this really happen? I don’t think so and probably you neither…. Understanding a possible solution requires changing the way we plan buffers, but also to acknowledge that having safety stocks is a good thing. As with forecasts, the one thing we know for sure is that our estimate will be wrong, but hopefully we will plan within a median. Having a buffer in a project is just like keeping inventories to protect forecast error. As explained, the problem is that we hide the safety stock within a task and that we consume it rather than “buffer” it.
CCPM: The Critical Chain Project Management (CCPM) suggests managing safety stocks in a different way.
1. Task duration need to be estimated using a 50/50 approach, without Safety Stock.
2. Tasks needs to be analyzed and understand if they fall within the critical path or not.
The Critical Path is the path that sets the pace of the project (i.e. if A and A’ can start
the same day, A takes 2 days for completion and A’ 5 days, and if task B depends on A
and A’, then A’ is part of the Critical Path as it constraints the speed of the project. A
would then be a non-critical task)
3. For tasks within the Critical Path the accumulated buffer size that would have been
used for each individual task is put at the end of the project as “Project Buffer”. One
simple rule is to use 50% of the total project leadtime as Project Buffer.
4. Non-Critical Tasks are also managed in a different way. First, these tasks are started as
late as possible (ALAP) versus the traditional as soon as possible (ASAP). Although this
might not seem smart, the logic behind is that knowing that projects change their scope,
it doesn’t make sense to invest time and resources to perform a tasks before it is really
needed. It is much smarter to use resources exactly when they are required, to avoid
duplicate or unnecessary work from changes in scope. Second, a special “Feeder
Buffer” is added to non-critical tasks to ensure that delays in non-critical tasks don’t
delay the start of a critical one (which would turn a non-critical task into critical and
delay the whole project)
Managing a project with this approach reduces lead times and costs and increases probability
of finishing a project on time. Of course, the way these projects need to be managed differ from
the traditional way as for example communication becomes critical (i.e. tasks don’t have a fixed
start date, but resources must be ready)
Some important changes in Project Management using the CCPM approach are related to
Project Tracking. Under this methodology, project tracking is
not done based on % of task completeness, but rather on
buffer usage/penetration. This methodology acknowledges that
tasks can finish earlier or after the planned date, which means
that finishing late is not bad if we have enough unused buffer
(or Safety Stock). This means we need to track buffer usage. If
buffer penetration is within a green zone, no actions are
needed. If they are within a yellow zone we need to plan for
buffer recovery and if we are in the right zone we need to
implement the recovery.
It’s the same with inventories. Using the Safety Stock is not
bad. Safety Stocks are part of the plan, but if inventories are in
the red zone we need to ensure the recovery to avoid service
Conclusions and references
Usually most project finish late and cost more than planned. Not because we don’t spend
enough time planning the project, but rather because we assume projects don’t change during
its life and because our human side plays against us.
Luckily there is a way of managing projects that considers these issues and tackles them.
For additional information I suggest the following readings:
Goldratt – The Goal (develops the concept of Theory of Constraints)
Goldratt – Critical Chain (uses the concept of TOC for Project Management)
Leach – Lean Project Management: Eight Principles for Success (a Project Management Book)
By Stefan Toma