Be Fast or Be Gone: Racing the Clock with Critical Chain Project Management by Andreas SchererLast annotated on November 17, 2014
every company I’ve ever known has had timing problems. Almost everyone is always running later than they want to be.
“During the planning phase we look for two things. First, tasks or sequences of tasks that can be worked in parallel instead of in series. That’s a huge source of speed. Second, we try to find the true touch time for individual tasks—that’s the hands-on time that it takes to get something done. Once we have that we’re able to come up with much more compact project schedules.
“We also model the resources required to get the work done. This lets us find the longest chain of tasks that takes task and resource constraints into account. That is the Critical Chain. Then we work to shorten it. If your team has identified the shortest possible Critical Chain, it has found the fastest way to get a project done. We combine that with a week-to-week update process that turns that plan into reality. That’s how we saved thirty-five percent of our cycle time.
“Every week we identify those specific tasks that could delay the whole project schedule. We call them “Critical Chain” tasks. By definition these tasks drive the timeline. We ask any person working on one of these Critical Chain tasks to work on it with as much focus as possible.
The best part is that the clear focus and priorities that come from the schedules can help dramatically reduce multitasking, which means everyone can be faster and more efficient. It’s almost like a relay race for knowledge workers.”
“You’re always going to have stuff that comes up, problems you didn’t anticipate. What we do is take some of that time we took from the individual tasks and give it back to the overall schedule. We call it a ‘buffer.’ It’s safety time that protects the whole project, not just one person or department. When you make safety time explicit like that, and when you share it out through the whole project, you can manage it and use it where it’s needed. Critical Chain schedules provide fantastic on-time performance.”
If anything, we found at Versa that quality increases with this approach. We want people to focus on their essential tasks. We want them to get it right the first time. Doing things that way avoids expensive rework.
The principles are pretty fundamental. It’s really just about improving the speed and consistency with which projects get executed.
give me two weeks. Let me get a close look at your processes, interview your team, and create an initial project schedule. I call it a network build. I would need the team to be all together in a room for about three days. We’ll look into every corner of your processes. Then I can tell you what’s possible here at Altus and what’s not.
“But when you’re in charge it drives you crazy. Instead of giving you the bad news right when they get it and when there might be time to address the problem by maybe borrowing some people from another project, your crew waits around and everything gets later and later and more and more expensive.
You put a little extra time in so that the committees have something to cut. When they cut that amount, you’re just fine. When they cut a little more, you have to hustle or play a little Schedule Chicken. But sometimes you get lucky and they cut less. And then you win the game right there. And everybody knows they’re playing and nobody talks about it,
To allow for all the time we lose from the multitasking and switching of priorities we just add more padding to our templates.
“And if they don’t cut you to the bone and you end up with a little extra time, I bet nothing ever comes in early, because when something looks like it will get done early there’s always something else to distract from it that might be late. Some people call it Student Syndrome, because students always seem to want to wait until the last minute and then scramble to put their stuff together.”
The irony in this world was that most people met their commitments individually but a significant number of projects were late.
I have a pretty good picture of how you run your project timelines. You have these milestones for any project. They’re negotiated dates—very safe, very conservative, because it counts against people if they don’t make them, regardless of why. So, everyone makes sure that there is plenty of protection for their individual tasks. There’s so much padding no one needs to push, and no one’s ever early.”
If someone in a project gets a task done early, they probably don’t say a word about it, because they don’t want to be expected to deliver that fast all the time. No one wants to see their protection disappear. Also, since nobody’s expecting anyone’s work to finish early, there is no advantage in the current system even if part of a job does get finished faster.
Basically, if something good happens and you are ahead of time, then you don’t have a way to take advantage of this early finish. And if something bad happens, if things take longer and a milestone is pushed out, it ripples through the entire project and you’re late.
What do you need from me?” “Just two things. I need your endorsement. Let’s bring the team together and tell them you’re trying something new. You’re thinking about piloting a new approach to Project Management. Then I need your support in challenging the templates.”
Sometimes, there are tasks that are worked in series when there is no real reason for it. If we find those, we can go ahead and ‘break’ the link between them. If we do that whenever possible, we can save a lot of time.
We need to figure out what we call the ‘touch time’ for each task. What I mean by touch time is the actual time it takes to get a task done if you do only that task.
Instead of putting in our personal buffers on each task, we will create a project buffer as a shared protection for the whole team to use.
The thing to get about it right now is that when you’re using Critical Chain, no one is going to punish you if you don’t hit your task estimate exactly. All you have to do is to work with as much focus as possible on tasks that are crucial to the project.”
Then we can use this software to compute what is called the Critical Chain. This is the longest chain through a project network that takes into account all task and all resource limitations.
We need to know which tasks you need to do in what order. We need to know what resources they need and what their touch times are.
we can come up with a sequence of key tasks that gets us from A to B. We will also have identified the true touch time for those key tasks. As a result we’ll have a strong plan that keeps our focus where it needs to be as we go into project execution.
“Then, we’ll meet on a weekly basis. We’ll discuss which tasks are on the Critical Chain. Those tasks will need to be worked with high priority. Everyone in this room will be given permission to focus on Critical Chain tasks until they are done. That way the work can be handed to the next person who is waiting for it. We’ll minimize multitasking.
“Then, last, there’s the project manager. It’s my job to help overcome obstacles, run interference so that you can do your job, and screen the project network. I’ll constantly look for any risks to tasks, in particular those that are on the Critical Chain. Now, the PM will need your help identifying these risks.Formun Üstü
“Say I have ten books to read, each with two hundred pages. If I read twenty pages per day and I multitask, I’ll have read twenty pages of each book after ten days of reading. And I probably won’t have a clue what’s happening in any of them. After twenty days, I’ll be forty pages into each book, and I’ll finish them all somewhere between days ninety-one and one hundred. And since I don’t have a perfect memory, in all likelihood, I will have to go back to remind myself of what I’ve read. Going back over material I’ve already read is what I call switching costs. Switching costs in the real world can be even higher that. They can easily make up twenty to thirty percent of the whole task.“But, if I read in a focused way, I will have read all of my first book on day ten. I will finish my second book on day twenty. At the end of day ninety I will have read nine out of ten books and be ready to start my tenth. I won’t have any redundant reading to do. I won’t have any switching costs.
“So, the first takeaway is: Reading in a focused way, I will have read nine out of ten books on day ninety. By multitasking I won’t have completed any of the books by then. It’s all still work in progress. That means I cannot fully take advantage of the content of any of the books, nor can I pass any of the books to someone else. Now, the switching costs are a big deal. The more books I read in parallel, the harder it will be to keep track of where I am. If there were a hundred books on my night stand, I would get nothing done. I would have forgotten what was in the first twenty pages of the first book by the time I finished the first twenty pages of the hundredth book. I would have to read it all over again.”Formun Üstü
“The typical project buffer is about 50% of the Critical Chain length. So, for Supragrel, we would add three months to the project schedule.
This methodology relies on three simple principles:
One: Aggressive but Realistic Schedules—Create the best possible schedule to get a project from where we are to where we want to go, taking into account the inevitable bumps in the road.
Two: Running the Relay Race—Apply focus on key tasks that are drivers of the overall timeline. Get those tasks done quickly and move the results of the work quickly to the next person. Have the mindset of a relay racer.
Three: Proactive Risk Management—Identify and address as early as possible all key risks that might impact the project and its timeline negatively.
“We’re going to make it very easy to track how our race is going. Every week we’ll update how far along each task is. Based on the input of each team member, we’ll be able to calculate something called a fever chart. A fever chart shows us how much work we have done, how much we have left to do, as well as how much of our buffer we’ve used up and how much is left. This is what a typical fever chart might look like. On the x-axis it shows us the percent of the Critical Chain we have completed. On the y-axis it shows how much buffer time we have used up. The colors show how much trouble we’re in: the more buffer is used, the hotter the fever. Green means the project is on track for an early finish. Yellow means the project is on track for an on-time finish, but the team needs to watch out for additional buffer consumption. Red means: Watch out; if the team doesn’t find ways to gain back some of the time, it will be late. Being in the red does not mean that everything is lost. It just means that the team needs to execute a buffer recovery plan.”
in this project as in every other project he’d worked on, some tasks that weren’t formally on the Critical Chain were very close to it. Identifying those and keeping an eye on them was essential.
7:30 A.M. Those meetings, scheduled so that all the team members could hit the work week up-to-date and ready to run their lap of the relay, allowed the team to work through the most immediate deadlines and to clarify who was doing what. Most importantly, Mike always used the meeting to alert team members who were about to receive input from other members who were ready to finish and pass on their Critical Chain tasks. Managing these handoffs was crucial for speed. With fair warning, people were able to reschedule personal and office commitments so they could focus on their Critical Chain task when they got the baton.
People who had the baton on their desks received special treatment. They were exempt from going to meetings that were not essential to their Critical Chain task. They weren’t expected to return emails or phone calls right away.
The kind of focused work, creative thinking, and good humor that spurred the use of a real baton was becoming a hallmark of Mike’s team.
As a result, people in R&D were multitasking like crazy because they didn’t want to be seen as uncooperative. No one at Altus Labs liked to say no.
“My team has some very talented and experienced project managers. I’d say about ten percent of the group are rock stars. Everybody wants to have them leading their projects. They get the job done. Then, I have about ten percent in my group that I’d rather not have. Eighty percent are somewhere between okay and good. They’re solid but not exceptional. What I need is more rock stars. We keep buying new software, but I think we’re chasing the wrong problem. We don’t need new tech. We need better processes, and we need people who can execute them consistently.”
We need revenues to continue our research. The drugs that make it to market don’t just need to recoup the investment we make in that particular drug. They also need to recoup the investment in all the other drug projects that never will be a product.”
He said that the entire industry has a cold but that Altus Labs has pneumonia.
Altus Labs is obese, perhaps morbidly obese. We are not talking about XXL obese. We are talking being overweight by two hundred pounds. Altus can barely walk. It is on a diet of fast food, soda, and chocolate chip cookies. This needs to change. We need to put Altus Labs on a health plan to get it back on track. Altus Labs needs to learn how to eat healthy. Altus Labs need to exercise. There won’t be rest for anyone at Altus Labs. We are going to have to walk faster—even run—because otherwise we are not going to make it.”
you can’t even keep your commitments half the time. If we were an airline, nobody would fly with us.”
When it comes down to it, the milestones and due dates you are using to manage your work are nothing but negotiated timelines. For management, nothing can be fast enough, so you try to get to aggressive commitments. For the people doing the work, it’s important to hit commitment dates because that’s how they make their annual bonuses. So they try to get the least aggressive commitments.
second major issue has to do with prioritization, or really with the lack of it. You have introduced a vast number of new projects into your R&D pipeline
“Here is the problem. Humans are not good at multitasking. If I asked you to drive from here to Philadelphia at sixty-five miles per hour while having a meaningful phone conversation with one of the Wall Street analysts, you would probably do one of those things pretty poorly. If I asked you to do these things and also write an email to your staff about their annual goals and objectives, you would probably either give the wrong guidance to your organization, mishandle an important call with a key analyst, or crash your car. However, you routinely expect your scientists to juggle many projects. They have no chance to stay focused on critical tasks, because they really don’t know what is critical.”
“But if we don’t have milestones, how can we hold people accountable for their timelines? How can we make sure that they do their job?”
We ask them to hit those, when possible, but to know that if they estimated incorrectly, we’ve got some buffer built in to take care of them. By doing that, we encourage people to finish early, and we can take advantage of that. Since the whole project team owns the buffer, they work together to conserve it. That way we have more accountability on the team level, in addition to the increased visibility on the management level.
“Our teams need to have the ability to prioritize the work on their project. If a task is on the Critical Chain, our teams need to have permission to do anything they can to get that task done. If that means someone needs to skip a non-essential meeting, then let it happen. If a team member is working on multiple projects, she must be allowed to focus on Critical Chain tasks over work that is not yet on the Critical Chain. If someone must choose between two critical tasks, they need to be able to tell which belongs to the higher priority project.
“I doubt that we’re going to be able to get all of our drugs perfectly prioritized. How can we possibly prioritize correctly given all that uncertainty in this industry?”
“You’re right. There will never be a perfect prioritization scheme. But right now you have none. Each of the fifteen thousand employees in R&D can develop his or her own priorities, and I guarantee you they won’t look the same. As a result, you have a huge amount of multitasking. “What you need is the relay race mentality
We will have the portfolio ranked and we will have a mechanism to communicate our priorities.
I need the top ten PMs in the company to learn how to teach, learn how to run projects, and learn how to go through team planning. It will take three to six months to get this initial group ramped up. They’ll be trained on the job, while we are putting Critical Chain plans in place for our key portfolio projects. I think we’ll need about a year to convert the entire project portfolio.
We need a combination of software and process that lets us support the Critical Chain methodology and the ‘relay race’ for the whole company.
“Does your EPM system come with a process to take an organization from nothing to a full rollout of a consistent project management process? Does it have training and certification for project teams and project managers to create an organization that can consistently execute good project management? Does it allow you to communicate buffers and fever charts and identify the Critical Chain of a project? Have any of their customers reported double-digit cycle-time improvements in their lines and a ninety-five percent or higher on-time delivery rate?”
make no mistake, the Critical Chain software is only the catalyst. This is about changing your corporate culture. If we want to successfully turn into a culture of diligent planning and stellar project management, then we need a Senior Leadership Team that is engaged in the process. We need your critical thinking, your endorsement, and your ability to lead change.”
we need to put a steering team in place that probably consists mostly of your direct reports.
With the prioritization scheme set, and with the state of the pipeline out there for everyone in management to see, it was time to get serious about fixing the things that had allowed the problems at Altus to get this far.
“The consultants will show you how to lead a team through an effective network build. It’s your job to learn how to help teams build tight networks. They’ll also show you how to move a team into using Critical Chain behaviors. Once we’re done with the twenty-six Eagles, the consultants are out of here. So use every minute of their time. You need to learn the process, and you need to learn the software. This is a huge opportunity, but it’s not one you can afford to screw up.”
“The consultants are going to be able to help you a lot with buy-in. They have the tools to prepare you for in-depth questions, and they have solid process expertise to lead teams through the planning and execution cycle. But you’re right to be concerned. This is a change in the way we conduct business here. Some people will get right in line with this, some will take more time, others might even resist.
So, all the Eagle teams will have an executive sponsor. That means you can expect that Graham, Stephen, or one of their direct reports will show up for kickoff sessions. They will personally explain to the team why this approach is so critical to the company. That ought to help, but if we can’t overcome buy-in issues quickly, please let me know about them right away.”
Critical Chain wasn’t one more improvement initiative at Altus Labs anymore. It was “the way we do it from here on out.”
We did a bottleneck analysis based on the existing Critical Chain schedules. It seems we have strong evidence that Clinical is driving the timeline for most projects.”
That’s how you create capacity. Kill a few projects. And pretty soon, we’ll find out where we’re getting held up and we’ll know where to put all those extra people you’ve just found.”
We are going to be able to reduce the front end of our clinical trial process significantly. This means we will be able to start the enrollment process faster for hundreds of studies every year. This is huge. Turns out that with the new process clinical operations is on the Critical Chain only forty-two percent of the time. Manufacturing and other functions show up more on the Critical Chain now, too. Mike and I were talking, and we think it’s time to call the other functional VPs. This kind of targeted analysis would make sense for their areas too. What happened with clinical operations is not a one-off thing, and it’s not limited to that particular function. It looks like we’ll be able to repeat that kind of improvement on any process from here on out.”
knew that you want Critical Chain to be a priority and that you care about their work situation as well. Ultimately, success will breed success. Let people hear what we’re getting out of this. Let them see what they’ll get out of it. Make it known.”
Implementation PhasesA Critical Chain implementation is done in two major phases: The Pilot Phase and the Rollout Phase. Each of these phases contains a Learn phase and a Confirm phase. The key objective of the Learn phase is to understand the causes and effects of suboptimal project execution. In the Confirm phase, the team uses the Critical Chain Project Management processes to overcome these deficiencies.
Phase I: The PilotThe main objective of the Pilot is to learn how the principles of Critical Chain and the Critical Chain Project Management processes and tools apply in the particular organization. The Pilot also shows the efficacy of the overall approach.
Phase II: The Rolloututilizes the “Learn and Confirm” approach at an organizational level. In the Learn phase the implementation team completes their understanding of all key objectives of the executive stakeholders. They develop a cause and effect analysis that is based on the Pilot and on additional findings made during the interviews for the Rollout. The burning platform of the Rollout is based on strategic issues that are relevant to the entire organization. Typically, it revolves around a combination of time-to-market, on-time delivery rate, or quality of execution.
The Confirm phase of a Rollout includes a number of key elements. There is a senior-level steering team overseeing the implementation. Its role is to establish the vision of the rollout as well as monitoring its ongoing progress. There is a communication plan that ensures that people within the organization understand the importance of the rollout and its objectives, and are kept informed about ongoing developments. This can also include communication with external stakeholders (e.g., shareholders, scientific community, etc.).
During a rollout, it is critical for an organization to move its entire portfolio to the Critical Chain project management methodology. That helps minimize the disconnects that can come from different ways of working and helps maintain momentum toward adopting the new approach. In addition, as project schedules are developed for the entire portfolio, the organization and its internal experts gain important knowledge about the approach, ensuring that the organization can independently execute the Critical Chain methodology. The certification approach ensures that projects are planned and managed on a weekly basis in a consistent way. Project teams are following a consistent process and report status based on a common set of indicators.
We strongly advise our clients to capture on an ongoing basis key metrics such as on-time delivery rates and cycle times. As the organization increasingly understands what processes are holding up its work, improvement initiatives—so-called “enabling projects”—are launched to further improve the operation. Ultimately, the Critical Chain approach becomes the lynchpin of the new relay race approach that brings not just the immediate benefits experienced by Altus Labs, but more and more benefits into the future.