Task managers assign resources to current tasks in order of priority. If resources are not enough to handle even the red tasks (tasks that have crossed the threshold of acceptable buffer consumption), overtime and other such decisions are implemented.

Preparing Tasks

After taking care of current tasks on their plate, task managers turn their attention to upcoming tasks. They ensure that all necessary preparations, such as getting approvals, drawings, materials, etc., are made so that tasks can be done without interruption as soon as the work of the preceding task is complete and available.

Organizations find it useful to formalize the responsibilities of front-line managers around the aforementioned aspects of task management.

Reminder: Do not pressure resources to meet planning estimates! Otherwise, you will soon be back to Square One.

Step 5: Implement Surrounding Processes

After reducing WIP and establishing task management, the benefits of synchronization will already be evident. Projects will be completing faster, firefighting and multitasking will be substantially less, and managers will feel more in control. However, the following surrounding processes are also needed to complete the picture.

Project Control

This is typically a formal weekly process to respond to uncertainties such as scope changes, technical problems, etc., that cannot be combated through routine task management. If the rate of buffer consumption (percentage buffer consumed versus percentage work completed in the longest leg) is too high, then project managers know which legs of the project are in the "red." They can then develop and execute recovery plans for those legs. Recovery plans can consist of run-of-the-mill items like scope adjustments and overtime as well as unique, even brilliant, solutions for specific situations.

Pipeline Control

While project managers can keep the buffers within their individual projects in control, it works only when just some projects are "red." If most projects are running behind schedule, there is probably a more systemic or global issue at play that is affecting all projects in the pipeline. This is where senior managers step in and make global decisions like putting some projects temporarily on hold, reprioritizing projects, or authorizing across-the-board overtime.

For example, about a year after the initial implementation at an aircraft maintenance and repair depot, the number of red projects jumped from 30 to 70 percent. The underlying reason was a sudden increase in the sheet-metal work required on incoming aircraft. During the three months it took to ramp up sheet-metal capacity, the number of active aircraft in the sheet-metal department was reduced from four to three and the maximum allowed overtime was authorized. As a result, the duration of the sheet-metal phase came down from 65 days to 47 days; projects trended back to "green" and the required Throughput rate was achieved.

Pipeline Planning

During WIP reduction (Step 2), Execution is monitored to verify that the initially selected drum is still valid. If the original drum is getting starved for projects, then the real drum could be an earlier phase or an upstream resource; similarly, if queues build up downstream of the original drum, then the real drum could be a later phase or a downstream resource. The drum can be changed if such conditions persist.

In the event the drum changes, managers formally meet to reset project priorities and possibly revise due-date commitments. Then decisions about project priorities can be made routinely as new projects are undertaken or as business conditions change.

While actual decisions are made by managers in charge of the project operations (in consultation with other affected functions such as manufacturing, sales, and marketing), a dedicated "Master Scheduler" or "Pipeline Analyst" is typically required to provide analytical support.

Capacity Management

The loop is closed with a capacity management process that identifies and mitigates resource shortages. The required information comes from an aggregate database of project plans, which shows the total resource "load to capacity," as well as buffer analysis, which identifies the resources that drive high-buffer consumption.

An important point is that the capacity of resources that are recovering buffers should be maintained or even increased (at least temporarily), even if it shows up as excess capacity in the "load to capacity" view.

In IT and engineering projects, for example, subject matter specialists do not have many explicit tasks in the project plans. A "load-to-capacity" view will show them as being 20 to 30 percent utilized. However, these specialists are vital for recovering buffers. Keeping their planned workload at 20 to 30 percent is a good practice that ensures both project delivery and pipeline Throughput.

Step 6: Identify Opportunities for Continuous Improvement (POOGI)

Having put the Three Rules of Execution Management into practice, the holy grail of projects-how to prioritize improvement efforts-can now be pursued. Since almost every process in projects can be improved, it is essential to pinpoint and focus on those improvements that will have the biggest impact on global performance.

As is known, the most harm to lead times and Throughput is done by practices and resources that have the most impact on project buffers. Therefore, the logical way to prioritize improvement efforts is: Record reasons for delay in task completions.

During buffer consumption calculations, identify the tasks that are affecting project buffers the most, and classify the corresponding reasons for delays.

Do a Pareto analysis of the reasons for delays across all projects and address the top reasons.

Organizations that have focused and prioritized their improvement efforts in this manner have achieved even shorter cycle times and much higher Throughput than they achieved during the initial implementation.

Step 7: (When Applicable) Use Superior Delivery as a Competitive Advantage to Win More Business

When most competitors don't deliver their projects on time, and late delivery has a big effect on their clients, reliable due-date performance can give companies a competitive edge. Some companies in engineer-to-order manufacturing, after stabilizing Execution, have been able to win more clients by coupling their offers with large penalties for delays.

While completing projects early is not always relevant for the client, in some cases it is critical. For example, the U.S. Air Force captive MRO facilities improved their service and value by offering faster turnaround on aircraft that were in high demand.

Similarly, a supplier of equipment for power plants was on the critical path of projects to set up those plants. This supplier was able to increase its win rates without offering price concessions by promising and delivering shorter lead times.

Lessons Learned

Following are some of the key lessons18 drawn and shared by hundreds of managers who have implemented Critical Chain.

Performance Gains Come from Managing Differently, Not Better Planning and Visibility

While good plans are essential and the rate of buffer consumption status is an effective way to monitor projects, increasing the rate of execution requires changing the way execution is managed.

Projects can be planned better even without Critical Chain, and the rate of buffer consumption provides similar information about project status as a comparison of actual time-lines against the baseline. If only better planning and visibility are what is required, they can very well be achieved with traditional methods.

However, unless WIP is reduced, task-level measurements are abandoned, projects are planned with shorter cycle times, and buffer-based priorities are followed, execution priorities will not be synchronized and projects will not be done faster. Nor will Throughput be increased.

Implement All of the Three Rules

Experience over the years has shown that the Three Rules of Critical Chain must all go together (see Fig. 4-2). Not implementing any one of them only shows up as lack of results or resistance to change. For example, organizations doing multiple projects with shared resources might be tempted to implement Critical Chain one project at a time. They ignore the pipelining rule. In a shared resource environment when WIP is not lowered, conflicts for resources continue. Priorities cannot be followed, buffers are consumed, and commitments are missed. Very quickly, faith in Critical Chain is lost.

Many times organizations aim just to gain control without increasing speed and Throughput. They compromise the buffering rule (for example, cycle times are not cut, but buffers are added). When cycle times are not cut, pipelining is compromised because long cycle times mean high WIP. When WIP is not lowered, Buffer Management cannot be done. The entire system falls apart.

Some managers compromise Buffer Management because they feel this is micromanagement. However, without people working to a single priority system and without timely interventions, buffers are wasted. This creates a sense that shorter cycle times were unrealistic. Eventually the organization reverts to its old ways (high WIP, safeties embedded inside individual tasks, and ad hoc priorities in Execution).

FIGURE 4-2 Why implement all the Three Rules.

Top Managers Must Play an Active Role

Mere sponsorship by top managers is not enough. Even though the top managers' role is typically to set policies and make planning-time decisions (project execution is delegated to middle and front-line managers), in successful implementations the top managers take on a more active role for the first 6 to 12 months.

The first reason is that middle managers and front-line managers encounter policy obstacles that they do not even know can be removed. Only senior managers can identify and eliminate those policy obstacles. For example, middle managers frequently assume that project starts cannot be staggered because clients will not buy-in; however, when the matter is brought up to top management, they are often willing to explain personally to their clients the benefits of pipelining projects. The CEO of one medium-sized manufacturer of industrial equipment even undertook a tour of customers around the world to explain pipelining and get their buy-in.

Second, managing buffers takes time to become a habit. It is only human to revert to old ways as soon as there is a minor hiccup. Close oversight by top management is necessary until managing buffers becomes second nature ("constantly peering over the shoulders" as an engineering manager from one company put it). The leadership in a U.S. Air Force Logistics Center went on daily rounds and for three months personally got involved in resolving issues.

Finally, outsiders can teach concepts. However, how to manage differently is better "taught" by top managers. For example, officers of senior rank in military organizations and "C" level executives of multibillion dollar companies have personally taught and coached their middle and front-line managers in the principles and practices of Buffer Management.

Actively Manage the Buffers

Buffer reports provide an accurate status of Execution. However, merely communicating status is not where the advantage of buffer reports is. The power of Buffer Management comes into play only when used by managers to respond actively to uncertainties. Here is how buffers are managed at various levels in an organization: Task managers-In contrast with traditional project management, the advantage of Critical Chain in execution is at task level because that is where the work is done. All organizations implementing Critical Chain, ranging from tens to thousands of people working on projects (whether they do research, engineering, or manufacturing projects), have realized the importance of task management. Talking about its implementation at a fashion garments supplier in Australia, the responsible person observed: "It is quite simple. You update your tasks, follow priorities, and get the work done." According to the engineering director of a home appliances company, "Setting processes and guidelines for Task Management is the key." Another successful adopter from a submarine maintenance facility put it as, "The supervisors look at their task list and allocate resources based on priority. It is that straightforward."