Does it seem that projects in your organization require frequent firefighting where due dates are challenged and heroics are necessary to get things done? Do projects often go from green to red overnight? Is team morale suffering? Do you believe that your organization could flourish to new heights if this cycle were broken?
Lack of focus and inability to manage uncertainty are two significant causes of project delays, diminished quality, excessive project durations, and low team spirit. By addressing these root causes, critical chain project management (CCPM) techniques improve project speed, quality, on-time performance, and team morale.
Reducing nonproductive multitasking enables high-speed execution, high-quality deliverables, a greater sense of accomplishment, and an increased capacity to think and to innovate for team members. It is significantly more effective to prioritize and execute tasks sequentially rather than juggling many tasks on multiple projects simultaneously. The end result is that even the lower priority tasks finish earlier.
Several leading global organizations, such as Eli Lilly (Lechleiter, 2009) and IBM (Ricketts & Dager, 2010), have adopted CCPM solutions to significantly improve their project management results. Another such organization, Mazda Motors, credits CCPM with a recent, major company turnaround (Ptak, 2013).
Problems Associated with Poor Project Performance
In order to implement effective solutions that improve project performance, it is important to identify the problems and isolate the root causes. When the numbers of problems are many, it is vital to address the critical few that are most impactful to achieving the organizational goal(s). One such problem on projects is that of low team morale. When teams are not operating with passion, energy, enthusiasm, and full engagement, superior project execution is not sustainable and delivery excellence is difficult to achieve.
Another significant problem worth addressing is on-time performance. When stakeholders expect predictable delivery of project results, missing a delivery date can be costly in terms of perceived reliability and customer satisfaction. Improving the frequency of on-time delivery can drastically increase the confidence stakeholders have in an organization's ability to deliver and the potential for a steady stream of additional opportunities.
Excessive project durations are another significant problem that impacts projects and oftentimes leads to decreased return on investment (ROI) and opportunities lost to competitors. This paper explores each of these project problems in greater detail to identify root causes and to uncover effective solutions.
High stress: Low team morale
When project team members are caught in the frenzy of all tasks being top priority, they quickly find themselves in the middle of no-win situations. Any passion, excitement, and enthusiasm with which these project teams started are quickly dissolved into a mindset of simply trying to survive. Project managers begin to fight for team members’ time in an attempt to do what is best for their own projects rather than focusing on what is best for their organization and customers as a whole. Now, the loudest project manager gets the attention. Micromanaging, having more meetings, producing more reports, and adding even more project details appear to be the only techniques that can help. Unfortunately, these tactics deteriorate team morale even further.
When issues strike, project teams begin to focus on what is urgent at the moment and not what is important for the greater good of the organization. When due dates start to slip, the competition for resource time intensifies to unhealthy levels. Soon, the organization becomes one that values heroism to get things done. When heroism is valued, everyone is driven to become the hero. Team members begin to work nights and weekends in an attempt to make up lost time and get ahead. Eventually, everyone is focused on solving crises rather than executing harmoniously according to plans and organizational goals. This situation is devastating to teamwork and collaboration which is vital for delivery excellence. What levels of success can an organization achieve in this state?
Missed delivery dates: Late projects
For many product-based companies, being late to the market can cost millions of dollars in lost revenues. This is especially true when fixed duration patents are involved. Additionally, marketing campaigns that are planning on a specific launch date can face the high costs of modifying their marketing schedules when projects run past their delivery dates. An organization that frequently misses their delivery commitments can also suffer negative impacts to their financial potential and their reputation to customers, shareholders, and other stakeholders.
Conversely, an organization that consistently delivers on time or earlier can create a significant strategic advantage over their competitors. This strategic advantage can catapult the organization into ever-flourishing status of high-growth and high-stability. Steve Pryor, Engineering Director of Propulsion Structures and Systems at Spirit AeroSystems stated in a 2012 PM Network® article, “Any missed commitment erodes our credibility with customers. Even more important is the internal cost for late engineering, which can force high overtime rates, supplier expediting fees, rework and potential quality issues” (Fretty, 2012, p 52).
Excessive project durations: Reduced return on investment (ROI)
Most project managers are familiar with the situation where a project schedule appears to be significantly longer than expected based on experience. It is often the case that organizations begin to spend money when projects start, and they do not begin to make money or save money until projects complete. Therefore, it follows that reducing project durations directly and significantly improves an organization's ability to make a return on their investment. Not only is delivering on-time important, but a company can reap the tremendous rewards of a market bonus, or substantial additional profits, for each day of early delivery beyond what traditional methods allow (D. Daily, personal communication, 1 July 2013). Additionally, having a consistent reputation of delivering on shorter lead times than one's competitors can play a role in helping to create a decisive competitive edge. This competitive advantage can contribute to the ability to extend a mafia offer, an offer that customers cannot refuse, and an offer that competitors cannot or will not match (Lang, 2010).
Causes to the Problems Associated with Poor Project Performance
Addressing these problems themselves would be like merely treating the symptoms of an illness. Until the root causes are understood and treated, the core of the problem can never be effectively eliminated. Investigations into the problems associated with poor project performance begin to uncover root causes such as:
- low-trust environments;
- nonproductive multitasking, which decreases productivity, negatively impacts team morale, and causes missed due dates;
- uncertainty in task estimates that are inadequately addressed;
- student syndrome – waiting for the last possible moment to start a task – means that contingency is wasted and risk is increased;
- Parkinson's Law – allowing work to expand to fill the time allotted – also wastes contingency and increases risk;
- having high amounts of work in progress can cause multitasking and firefighting, which leads to low team morale and extended project durations.
One of the causes of poor project performance is low trust environments. Low trust environments are often created when team members do not feel they have a safe environment to surface issues quickly and transparently without the fear of being blamed. When team members feel they need to hide issues from management, project performance can deteriorate significantly. Issue resolution can take much longer, which can jeopardize delivery commitments, cause project delays, increase stress, and decrease team morale.
In 2009 and early 2010, NASA Dryden Flight Research Center determined workforce morale as one major issue which was blocking their ability to make significant improvements to their project delivery results. NASA decided to implement critical chain project management (Horn, 2012). With regard to workforce morale, managers wanted to address two items. First, they needed to be able to quickly get to the root cause of project issues. Second, they needed to be able to control work in progress in order to create better project outcomes. However, team members viewed the investigation of issues as a punishment and the controlling of work in progress as micromanagement. It is safe to say these perceptions, real or not, were hindering the improvement of trust within the organization. Management recognized the need to change their approach, and not the mindsets of their teams in order to improve project performance (Horn, 2012).
How much more successful could projects be if nonproductive multitasking could be minimized? A McKinsey Quarterly article indicates that multitasking damages productivity, slows people down, hampers creativity, makes employees anxious, and is addictive (Dean & Webb, 2011). Supporting these findings, a recent multitasking study states, “multitasking makes people less productive, less creative, and more likely to get thrown off by distractions” (Realization Technologies, Inc., 2013, ¶9). Multitasking can also be seen as a significant cause of extended project durations as illustrated in Exhibit 1. If a project team member has three project tasks to complete—A, B, and C—it can be seen that each task takes much longer to complete when multitasking. Context switching between tasks increases the completion times even further and causes the project team members to have to stop frequently to remember where they were each time they switch to a different task.
Multitasking also causes additional stress on team members as they try to keep all of the plates spinning simultaneously. The prolonged task durations diminish the sense of accomplishment team members feel when they drive a task to completion. Once the tasks are eventually completed, any sense of accomplishment is replaced by feelings of relief that the stressful multitasking of activities is finally over. The result is decreased morale and a team member who feels more exhausted than excited to have completed their tasks. Additionally, if a problem arises on one of the juggled tasks, delays can propagate among all of the tasks which can impact the on-time performance of multiple projects. Therefore, all three project problems - low team morale, missed due dates, and excessive project durations - can be caused by multitasking.
Uncertainty in task estimates
How much shorter could project schedules be if team members were not fearful of using most likely estimates? If team members are asked to provide a task estimate, are they inclined to provide one that is 50% likely or one that is 95% likely? Most project team members would provide the 95% likely estimate (Walker, 2010). Exhibit 2 illustrates two probability curves. The bell-shaped curve with a long tail represents the probability density of task duration. The “s” shaped curve is the cumulative probability. A project task that is 95% likely is typically two or more times the duration estimate of the 50% likely estimate (Exhibit 2).
According to Budd and Cerveny (2010), project task estimates are frequently taken as commitments. When team members know their estimates will be viewed as time commitments, they will often add excessive hidden padding to the task completion date (Realization Technologies, Inc., 2010). Management, receiving pressure to reduce project durations, will look to decrease task estimates in order to shorten project schedules. Team members, wise to what is going on, will play the game and add even more padding to counter the actions of management. A missed estimate equates to a missed commitment, which can lead to unfavorable evaluations and sometimes demoralizing punishment. It follows that this situation can also be a cause of low team morale.
Student syndrome: Waiting until the last possible moment to start a task
Student syndrome causes a significant waste of well-intended contingency. Most people have felt the effects of student syndrome at some point in their lives. When there is more time to complete a task than is truly necessary, it is typical that team members will wait until the last possible moment to begin (Blackstone, Cox, & Schleier, 2009). Exhibit 3 illustrates the amount of effort applied over time to complete a task. Very little effort is applied in the early stage of the task even though there are no factors preventing full focus (Goldratt, 1997).
The addition of excessive hidden padding in task estimates creates the likelihood of a self-fulfilling prophecy for task completion (Budd & Cerveny, 2010). If a task has an optimistic duration of two days, a most-likely duration of five days, and a pessimistic duration of 12 days (due to the long tail of most tasks), many team members will provide a task estimate in the upper end of the spectrum; one that has 95%+ probability of being achieved. For this example, it is assumed the estimate provided is 12 days. The team member, knowing that the task will only take two days if everything goes smoothly and will only take five days in most cases, is likely to wait until day six (or so) until they become fully engaged. Most of the hidden padding, intended to protect the task from uncertainty, is now wasted before the task even begins. If something unexpected occurs, there is no longer sufficient contingency to prevent the task from being late (Exhibit 3). This late task impacts successor tasks, which can cause an entire project to be delayed. When student syndrome is present in an entire sequence of tasks, the damaging effects get multiplied. A crisis situation is created, increasing stress and damaging team morale even further.
Parkinson's Law: Work expands to fill the time allotted
When tasks have such significant padding built in, it seems possible that some tasks will finish early. That might be a reasonable expectation if it were not for Parkinson's Law. If a team member completes a task early, there is little chance they will declare the task complete. Sometimes, this can be due to the tendency to improve or polish the completed deliverable. Other times, it can be a result of a team member wanting to be viewed as a reliable estimator. Declaring a task complete early can damage this reputation. Moreover, a team member who completes a task early is likely to have his or her task durations cut even more the next time they provide an estimate (Newbold, 2013). If this fear becomes a reality, trust is damaged further, and then let the games begin! The team member will add even more padding next time to make up for the amount that is likely to be cut.
High work in progress in the project pipeline
Another significant cause to the problems associated with poor project performance is having too many projects in the delivery pipeline. Are projects frequently started regardless of an organization's capacity to complete the work? The statement, “Well, it all has to get done!” might even be heard. Organizations can oftentimes be driven to get projects started as early as possible, following the ill-fated logic that the earlier a project starts, the earlier it will finish (Budd & Cerveny, 2010). However, when the capacity of the project delivery pipeline is not taken into consideration, this well-intended tactic can have devastating effects on the system as a whole. Having too many projects in the pipeline can slow the entire system down and drastically exacerbate the causes of poor project performance. What might be a good strategy for activating projects within the project delivery pipeline? Do projects ever get activated simply to keep everyone busy?
Critical Chain Unlocks a Series of New Paradigms
Critical chain project management (CCPM) is the project management application of the theory of constraints (TOC). First introduced by Dr. Eli Goldratt (1997), critical chain has gained significant momentum in the project community as a solution to the problems associated with poor project performance. CCPM takes resource constraints into account for more realistic planning and focused execution. The critical chain is the resource-leveled critical path. CCPM also utilizes pooled contingency to more effectively address the various components of project uncertainty.
The first solution addressed in this discussion, creating environments of high-trust, is not formally a critical chain solution. However, having significant trust within teams and across organizations is important when implementing change (Budd & Cerveny, 2010). The end result of improved morale and quality of work life for team members is more deliberately emphasized in the literature (Kishira, 2009). Of course, most operational managers, VPs, and CxOs appreciate the positive effects of high trust in their organizations regardless of the strategies and tactics utilized.
Additional CCPM solutions are as follows. Pipelining addresses the problems associated with having too much work in progress. Full kitting addresses the problems associated with starting project tasks without the necessary predecessor activities completed. To address wasted contingency, 50/50 task estimates are utilized. Executing with focus, rather than multitasking, addresses elongated task durations, productivity losses due to context switching, and the stress team members feel as they try to keep many plates spinning simultaneously. Lastly, project buffers shorten project durations and protect critical project dates from uncertainty; significantly improving the predictability of project delivery.
In an October 2009 speech, Eli Lilly and Company CEO Dr. John C. Lechleiter, announced the most sweeping changes in company history, indicating three new competencies that are changing drug development at Lilly. One of those competencies is critical chain project management. Dr. Lechleiter stated:
“The COE is implementing a project management methodology called ‘Critical Chain,’ developed by physicist Eli Goldratt. Critical Chain was actually first applied at Lilly in a completely different context by our IT group. They, in turn, helped our research labs launch a pilot program that has proved the power of Critical Chain in drug development. Historically, across our pipeline, we have a 60% success rate in hitting milestones on time. In other words, we miss almost half of our deadlines. In the Critical Chain pilot program, the success rate so far is 100%. That's why we're now applying Critical Chain in force (Lechleiter, 2009, ¶38).”
In a June 2013 keynote address to the Theory of Constraints International Certification Organization (TOCICO) international conference in Bad Nauheim, Germany, Mr. Mitsuo Hitomi, Executive Officer from the Mazda Motor Company spoke of a major company turn-around, and he credited critical chain and TOC (Ptak, 2013). Mr. Hitomi presented remarkable results with an emphasis on employee harmony, improved quality of work life, and positive financial results in 2013 after four fiscal years of operating in the red. Additionally, Mr. Hitomi indicated that critical chain momentum grew and resulted in development project durations being cut in half with all projects delivered with full scope and on time (Ptak, 2013). Echoing the sentiments of Dr. Goldratt himself, Mr. Hitomi stated after the presentation, “Even the sky is not the limit” (Ptak, 2013, ¶6).
Creating environments of transparency and high-trust
Stephen M. R. Covey, author of The Speed of Trust (Covey & Merrill, 2008) states:
“When trust is low, in a company or in a relationship, it places a hidden “tax” on every transaction: every communication, every interaction, every strategy, every decision is taxed, bringing speed down and sending costs up. My experience is that significant distrust doubles the cost of doing business and triples the time it takes to get things done. By contrast, individuals and organizations that have earned and operate with high trust experience the opposite of a tax — a “dividend” that is like a performance multiplier, enabling them to succeed in their communications, interactions, and decisions, and to move with incredible speed. A recent Watson Wyatt study showed that high trust companies outperform low trust companies by nearly 300%! (p 21)”
Patrick Lencioni, author of The Five Dysfunctions of a Team (2002), indicates that absence of trust is a foundational cause for team dysfunction. Yuji Kishira, in his book titled WA: Transformation by Harmony (2009; “WA” means “harmony” in Japanese), emphasizes the superior results observed when organizations are able to optimize high-trust within their environments and create true harmony within teams. In a PM Network® article, Hunsberger (2012) states, “The critical chain approach can provide much-needed relief for overloaded team members on a tight deadline” (p 41). Another PM Network® article states, “Spirit AeroSystems attributes critical chain project management to a 50% reduction in employee overtime.” (Fretty, 2012, p 55)
Overtime and stress can create tension among teams, which can lead teams down the path toward distrust and disharmony. When trust is low, productivity can be drastically impacted. The CCPM approach requires a fair level of trust between teams and management due to the nature of the transparency created. However, once teams begin to adjust to the new paradigms, CCPM behaviors can become a driving force behind improving trust.
Pipelining: Limiting work in progress
Organizations that can successfully focus everyone's attention on the few, critically important goals, rather than getting pulled into the tyranny of the urgent, can create a significant strategic advantage over their competitors. Organizations that are successful with this strategy can be a highly productive, harmonious, collaborative, and innovative entity that operates within a sustainable, high-trust environment.
The critical need to streamline work in progress is to maximize the flow of projects through the delivery pipeline. Local measures, such as how quickly a project gets started, are not valuable to the goals of the organization. Rather, the throughput of the entire organization, global optimization, should be the focus. Most would agree that any system is made up of bottlenecks and non-bottlenecks or constraints and non-constraints. If an organization activates projects just to keep the non-constrained team members busy, critically constrained resources quickly become over-allocated. Critically constrained resources are those who typically have the highest workload and the fewest resource numbers. What impact might this action have on the flow of projects through the delivery pipeline?
Since CCPM is focused on optimizing globally, the critically constrained resources are identified and projects are staggered in a way to keep these resources optimally loaded. Of course, if the critically constrained resources are optimally loaded, the non-constraint resources will be less than optimally loaded; even idle! (Realization Technologies, Inc., 2010) Granted, this flies in the face of conventional wisdom, but keep in mind, local optimization does not lead to global optimization. In which of these are most organizations interested? Asked in a different way, are organizations more concerned with strengthening the weakest link in the chain or the stronger links in the chain? Now that critically constrained resources are optimally loaded and non-constrained resources have capacity (i.e., to think, innovate, help others, train, take on unexpected work, etc.), significantly improved throughput (or flow) within the organization is feasible.
This technique to limit the number of active projects is a key element of critical chain. Concentrating resources on fewer projects at one time allows teams to focus and reduces multitasking. This not only allows projects to be executed faster but creates capacity to undertake more initiatives (Fretty, 2012, p 52).
Making sure project team members have all of the necessary predecessor activities completed (full kit) prior to starting their task ensures they have a greater chance to focus uninterrupted without having to be impacted by potential upstream delays. Also, they are more likely to be able to drive their project tasks to completion with high speed and high quality. The need to start, stop, and start again due to missing predecessors can significantly reduce efficiency and productivity. “[The full kitting] approach requires having all the pieces in place prior to moving a project work package forward. This step helps significantly reduce the number of second actions to complete a single stage.” (Fretty, 2012, p 52)
50/50 task estimates, most likely estimates
Critical chain minimizes student syndrome, waiting until the last possible moment to start a task, and Parkinson's Law, work expanding to fill the time allotted, by eliminating the need for every team member to hoard safety in their task estimates. As discussed previously, when team member's estimates are turned into commitments, project task estimates and entire project schedules can inflate drastically. Keeping in the spirit of optimizing globally rather than locally, it is more important to focus on project delivery performance than task completion performance. When management focus is shifted from on-time task completion to on-time project delivery, team members feel empowered to get aggressive with their task estimates while still accounting for a reasonable amount of uncertainty.
Instead of using the traditional 95% likely pessimistic task estimates, 50/50 estimates, those that have a 50% chance of finishing earlier and a 50% chance of finishing later, are provided during planning. The 50/50 estimates are also, roughly, half of the duration of 95% estimates. Therefore, the project schedule is temporarily cut in half, which is a giant step in the right direction. However, even though the wasting of task contingency has been minimized and team members have a renewed sense of enthusiasm with this safe, aggressive task goal, there is still a need to protect the project commitment from unpredictable events.
Focusing on the task at hand rather than multitasking
With the environment of high-trust and transparency created, the project pipeline optimized for throughput, and project tasks estimated with aggressive and achievable durations, projects are ready for optimized execution. This requires team members to focus on the task at hand and to drive it to completion with high speed and high quality.
Allowing team members to focus on their tasks sequentially rather than multitasking is, counter intuitively, a very effective solution to most organization's need to do more with less. As is illustrated in Exhibit 4, significant improvements can be made by allowing team members to focus on their tasks until completed. Conventional wisdom might have team members work more hours in order to get more work completed faster. Oftentimes, this option is not sustainable and can take a devastating toll on team morale and organizational trust. When organizations begin to realize the significant productivity improvements that can be gained by simply reducing multitasking, a new level of organizational performance can be achieved, creating a decisive competitive edge over the competition.
Tactically, critical chain enables the concept of focus by creating a relay race mindset among project team members rather than utilizing rigid task scheduling and task deadlines. Once a task begins, the team members drive the task to completion as quickly as possible with focus on appropriate quality and without fear of calling out issues that need to be addressed. Some project teams even use relay race batons as a fun way to bring visibility to the team members who are on the critical chain and who are focused on the task at hand. In a low trust environment, being on the critical path might be intimidating and uncomfortable, but in a high-trust environment, being on the critical chain can be a healthy challenge to be conquered.
At a recent conference, Dr. Reddy's Laboratories shared some striking results. In just 12 weeks after taking focused measures to sharply reduce its project team's workload and reduce multitasking, the company's development group completed 83% more projects than it had in the previous 12 weeks without adding any additional resources (Ziv, 2012).
As team members make progress on project tasks, they provide updates as to how much time they feel is left to complete their work. This information is used to communicate to the next team member who will take the baton handoff as soon as it is ready. Also, the remaining time to complete the task is used to determine if any time beyond the original 50/50 estimate will be needed. If the task is running late, due to unforeseen circumstances, a mechanism is needed to prevent an impact to the project delivery. It is not necessary to be concerned with on-time completion of individual tasks. Rather, it is more important to focus on project on-time performance. The mechanism used to protect the project delivery commitment is the project buffer.
Buffering for uncertainty
With the project planned using 50/50 task estimates, a decrease of approximately 50% of the overall project duration is expected. This padding was removed from the individual tasks (Exhibit 5) in order to prevent it from being wasted by behaviors associated with student syndrome and Parkinson's Law. However, there is still a need to have some amount of contingency to protect the project delivery from uncertainty. A portion of the contingency removed from critical chain tasks, approximately 50%, is added back to the project schedule as a pooled contingency, which acts like a shock absorber to protect the project delivery commitment (Exhibit 5). Projects are typically planned to complete 25% faster while protected by a buffer that absorbs project task variations (Fretty, 2012; Leach, 2011).
This pooled contingency, which is used by everyone on the project team, creates a very interesting, healthy dynamic for project success. No longer do team members need to feel alone to protect themselves from unpredictable events (risks). Rather, the entire team is motivated and driven to be successful together. If one team member needs additional time due to unforeseen circumstances, they are able to consume a portion of the project buffer. If another team member completes a task early, they add their time to help replenish the buffer. The use of a centralized and transparent project buffer allows for shorter project durations, improved on-time performance, and healthier team morale. A word of caution, testing is not a buffer.
A buffer consumption chart (Exhibit 6) is an objective, leading indicator of project health and is kept visible for all to see. The number of days of work left on the project compared with the number of days of buffer remaining provides an indication of how likely a team is to complete their project on-time. If the buffer is in the green territory, a team will continue as planned and managers are not required to interrupt team members to search for issues. If the project buffer reaches yellow, buffer recovery plans are made just in case it becomes necessary. If the buffer penetrates the red area, buffer recovery plans are executed. Recovery can include any techniques associated with crashing a project network.
With a buffer consumption chart, all team members, sponsors, and other stakeholders are informed of the health of the project at all times. Additionally, reasons for buffer consumption are documented so that continuous improvements can be made to the project execution and delivery system.
A similar chart can be used to bring transparency to the health of all projects in a program or portfolio. Exhibit 7 displays a buffer consumption chart that can be used in a multi-project environment to display the duration of work remaining versus the amount of buffer remaining for each individual project.
Organizations observe numerous project problems within their environments daily. Some of the most impactful are team morale, missed project commitments, and excessive project durations. It is important to identify root causes and not just treat the symptoms of these problems. Root causes include having too much work in progress, environments of low-trust, multitasking rather than focusing, and wasted contingency at the task level due to student syndrome and Parkinson's Law. Critical chain targets these causes by streamlining the number of projects in the pipeline for maximized throughput and flow, creating an environment of high-trust which allows team members to get aggressive with task estimates and to call out issues without fear of blame, and by creating an effective contingency mechanism, a project buffer used by the team, that is an objective and transparent leading indicator of project health.
The results demonstrated by organizations who have systemically embraced CCPM are impressive. Teams enjoy a sense of accomplishment as they operate with high focus and with harmony. On-time delivery becomes increasingly more predictable and on-time performance can improve from 30% to 60% to over 90%. Project durations can be cut by 25% or more without sacrificing scope, quality, or team members’ lives outside the workplace.