Project Management Central

Please login or join to subscribe to this thread

Critical Chain: Have you used this method in your projects?

The Critical Chain is a path-breaking theory advanced by the Theory of Constraints (TOC) author Eliyahu Goldratt. It offers breakthrough project performance by cutting down on both buffers and avoiding the "Student Syndrome".
Have you read the book? If so, have you used some of these concepts in your projects?
If so, what was your experience?
Please share your knowledge for the benefit of our PM community here.
Sort By:
Page: 1 2 next>

Critical Chain is nothing new for people who were trained in operation research which is the basement for TOC. I have the opportunity to use Critical Chain and I have the opportunity to be trained for Mr. Goldratt. And I am involved on implementing Critical Chain. The best advise I can give to people who like to use it is: Critical Chain demands a cultural change in all stakeholders mainly in project manager. Read and understand Critical Chain and you will understand why the cultural change.
1 reply by Karthik Ramamurthy
Feb 11, 2019 9:45 AM
Karthik Ramamurthy
Wow, Sergio Luis Conte!
That must have indeed been an amazing, once-in-a-lifetime opportunity, being trained in TOC by the grat man himself!
I completely agree that achieving success in applying Critical Chain requires a pradigm shift in the way Project Managers think about Project Management.

I have read Mr. Goldratt''s seminal work. What he makes explicit in the critical chain methodology is what most experienced project managers know: those who estimate lie.

Most people who are asked to provide estimate are usually not accountable for the accuracy of their estimate. (I worked with one company who actually gave a superior performance rating to those people whose estimates were alwways withing /- 10%).
As long as individual contributors are assessed for being under budget and on schedule, these conservative estimates will continue.

As a critical chain project manager, you understand that this is part of the game. You then expose the play by cutting the estimate back by 50%, ay, and allocating the cut amount to the buffer.

Of course, you also need to make sure the resource is dedicated to the task. (Part of the lie is because people get pulled in multiple directions.).

Hold your estimators accountable but make sure they are fully allocated to their activities.
1 reply by Karthik Ramamurthy
Feb 11, 2019 9:54 AM
Karthik Ramamurthy
Excellent points, Stephane!
Thanks a million for adding great value to this discussion by sharing knowledge from real experience.
It does make sense to cut from estimates and add to the buffer. However, at the same time, it takes a lot of people skills to convince resources that the job can be done in half the time.
And that, despite excellent performance, if something unforseen happens to delay the activity, you will be there to protect your resources.
Also, even whe the resource is dedicated to the task, we need to make every effort to make sure that when a task is scheduled to start, we have almost everything required to complete it.

One challenge I faced in implementing buffers is when contract language for the project specified that float belonged to the customer - and the customer''s interpretation of buffers was that it was an attempt to hide float. If you have to share detailed schedules with the customer, it''s essential that they understand what buffers are for.

Cultural change is really the greatest challenge a PM is to face when adopting Critical Chain approach. Not only to get the overestimated cuts, dimensioning buffers and making them understood to customers and teams but to get work package owners to report correctly remaining work durations and to get rid of the planned start dates to really start work as predecessors are concluded.
1 reply by Karthik Ramamurthy
Feb 11, 2019 10:04 AM
Karthik Ramamurthy
@Ricardo do_Rego_Barros: Great points which add real value to our discussion on Critical Chain!
Cultural change also becomes very important for other key stakeholders such as vedors.
After all, in several projects, vendors may have to complete significant work.
If they are not convinced, they will not accept terms at reduced timelines!
In some cases, incentive clauses may have to be put in to convince them. If the incentives are far lesser than the gains to be had through shortened timelines, they easily pay for themselves!

Am still trying to see how using this technique will improve the process. Take an estimate of 8d to complete an activity or 18mo to complete a project. Half goes to the buffer. Come day 4 or month 9 and if the estimate was within a definitive range to start, it will still consumes the buffer. At the end, the actual was 8d or 18mo /- the definitive range of the estimate.
1 reply by Karthik Ramamurthy
Feb 11, 2019 10:12 AM
Karthik Ramamurthy
Great question, Luke!
I see that there are already some excellent responses!
Please do keep contributing for the benefit of our vibrant community here!


Critical chain does not make much sense on a single activity. It also assumes that estimators tend to spread the work over greater periods than necessary. "That will take me a week", rarely translates into 40 hours of effort. (That''s why I never ask for estimates in days.)

Where critical chain really shines is when you take all the contingency out of the activities on a given path, for example the critical path, and put it at the end of the path. From a schedule point of view, you have not changed the projected end date for that path.
But by forcing the resources to do things quicker, you reduce the chances of the path being late. (Resources really should be dedicated to the task to finish as soon as possible.)

When you read the book, you will see that Goldratt also prones making sure that the key resources are ready and available to start when their task is expected to start.

Thanks Stephane. I read his book and appreciate what was discussed in terms of problems the approach was trying to solve. It didn''t seem like critical chain will manage the risks and issues discussed any better: culture of padded estimates, lack of dedicated or over allocated resources and bottlenecks.

Taking half an estimate and calling it buffer/contingency doesn’t seem to solve the problem even using hours. Seems like it boils down to estimate accuracy how much a resource is allocated to a project. Most of what I heard in the book is that people pad estimates but CC somehow resolves that. If the estimate is the same, the allocation is the same and the people are the same. Seems the actual time to complete will be the same regardless if someone talks half of it and calls if buffer or not. That’s the part I’m trying to figure out in terms of how the outcome would be different. Wouldn’t the feeder and/or the project buffer just get fully consumed?

You seemed concerned that people will use up the buffer regardless of what you do. Most people''s estimates are given under normal, unoptimized conditions of work: interruptions, secondary projects, ...

You really have to play both sides. As a PM, your job is to make it possible for the work to be done in less than the original estimate. As a project team member, it is your job to focus on the activity at hand and produce it as quickly as possible.

If you don''t make it work from both end, you will definitely wind up with the same problem. CC is not a panacea and, as Ricardo pointed out, it does require a culture change. For everyone.

Agreed, culture is critical to building a healthy project environment.

I''m not aligned with the thinking that the job of a PM is to deliver in less than the estimate - it''s to deliver to agreed upon expectations. If it''s to deliver to less than an estimate, then no wonder folks add extra contingency to their estimates. I''m ok with an entire buffer being consumed, as long as the work consumes its fair share - since half the estimate goes to contingency.

Good point, Luke.
I also think PM must deliver upon agreed upon expectations, and I understand I meet "agreed upon" when the Project results are equal to or better than set agreements.
As for the buffers, if all goes fine they should be consumed up to their limits, but I trust that it is only possible when cultural changes are such in the team that start dates are not considered as in standard planning (like - "Mr. PM, I''ll start my task in the planned date as I said"), preventing anticipations & delays compensation, which help buffers management. Activity end dates must actually dictate the start (immediately) of next successor tasks. Although seemingly easy, this is this actually the true challenge in using Critical Chain.
Page: 1 2 next>  

Please login or join to reply

Content ID:

We are ready for any unforeseen event that may or may not occur.

- Dan Quayle