[7] "Where does the additional throughput come from?"

"Where does the additional throughput come from?"

That was the question asked by my most recent client. It took some doing, but finally he understood, and when he did, he was a bit surprised.

"You are too Socratic," the client cautioned me. "You should have told me sooner," he declared once he really understood.

Of course, I had told him sooner, several times in fact. But he and his people hadn't believed me. They needed to develop the whole solution for themselves, before they believed it. They needed to experience the damage to real productivity. They needed to understand the underlying causes. And they needed to experience the solution, over a three-day-long simulation of a multiproject operation.

The client, we'll call him Mike to protect his privacy and my business, was referring to the doubling of the rate at which projects are completed by development organizations once their multiproject operations are in order, a doubling that takes place without the need to hire additional resources.

Doubling the number of projects completed per year sounds too good to be true, unfortunately. But it's fact. Confluence, for example, increased its project completion rate from six per year to more than eleven per year, without hiring a single additional developer. Bill Baron[1] increased his organization's project completion rate from five in 1998 to sixteen in 1999, again without hiring a single additional developer. These represent astronomical increases in the real efficiency of product development organizations. But they are not only facts, they are typical.

So where does the additional throughput come from? It comes from the elimination of a cause of tremendous waste, multitasking. Multitasking is the culprit that today destroys multiproject operations throughout the world, and it does its damage in a very insidious manner, by creating the illusion of productivity.

Every resource who contributes to the projects of his/her employer can work in one of two ways. The resource can work each task in a focused manner until it is completed, or the resource can multitask. Multitasking gives individuals and their managers the impression that they are incredibly busy. In fact, each time that somebody multitasks, he/she is destroying the organization's real productivity. By multitasking, a person is choosing to use some of his/her available capacity not to generate throughput but to create additional work in process, wip. It's that simple.

The schedule damage that multitasking inflicts on projects is monstrous. Multitasking acts as a multiplier of project duration. If the average number of tasks available to developers is two, then the average project duration for their organization is two times greater than it need be.

Most developers multitask with three or four active tasks at one time. This means that the number of active projects for most organizations is three to four times greater than it should be. It means, also, that the average duration of the projects of their organizations is three to four times longer than it need be. Take a look at the scatter plot in the previous article, and you'll see how much faster projects can be once multitasking is eliminated. Along with the shorter duration, of course, there is the increase in throughput that takes place as the full capacity of the organization becomes focused on completing projects rather than being misdirected to creating unnecessary and damaging work in process.

"But isn't the organization's full capacity recovered anyway, once all those projects in process are completed?" asked my client when he realized what was happening to the projects in his 23 plants?

The full capacity of the organization would be recovered, if everyone concerned could afford to wait for all those active projects to be completed. But that's not reality. Reality is that many projects take too long and are killed by management. When this happens, all the capacity that was spent on those projects is wasted. All the throughput that could have been created with that capacity is also wasted. This is why there's a real and very significant increase in throughput available to organizations that streamline their multiproject operations. The greater speed means that fewer or no projects are killed. It means that a greater number of projects are completed fast enough to make a real impact. In the case of homeland security, it means that a greater number of defense systems become available each year, with the same resources and with the same costs.

Every Reader’s Responsibility

If you agree with all that I’ve discussed in this blog, then it’s up to you to help make Secretary Ridge and other decision-makers in government aware of all that the government can do, to improve our ability to provide for the national defense. To that end, I urge you to share this document either directly with government officials or with individuals who have the ability to gain the attention of government officials.

[i] William Baron, Management Roundtable, Executive Summit with Dr. E. M. Goldratt, April, 2000. Baron’s business developed new leading edge optical cables for telecom industry. For Premise Cable products - 100% of projects completed on-time (Compared to 40% before); 8 projects completed before 50/50 points and New Product Introduction (NPI) lead times reduced 50%. For outside cables - NPI lead times reduced 50% and tripled development capacity (16 projects vs. 5 before, with no staff increase).

[So that others might also subscribe, please share the following link with friends and colleagues.Subscribe!]

[6] Tactic: Focus Resource Capacity

To achieve the company's dramatic improvement in performance, the leadership of Confluence made several changes. One such change involved the company's process for allocating resource capacity to projects. In the past, the company's approach was similar to that of the defense contractors. Each project was allocated a perceived fair-share of resource capacity. As a result of the leadership team's new paradigm, each project is now allocated all the resource capacity that it can use effectively. Further, given the company's prioritization strategy, each project actually gets all the resource capacity that it can use effectively, at the right time.

The result of this sustained focus on the right project at the right time is illustrated by Figure 1, below. That figure shows a scatter plot of project size versus project duration. The horizontal axis denotes project duration, expressed in business days. The vertical axis denotes project size, expressed as the number of staff-days needed to complete each project. Data points at the same height denote projects of equal size. The blue data points indicate the projects completed during the years 2000 and 2001, prior to the company's adoption of TOC Multiproject Management (not merely the critical chain model for individual projects). The white data points indicate the projects started and completed during the years 2002 and 2003, after the company's adoption of TOC Multiproject Management (not merely the critical chain model for individual projects).

The data in Figure 1 show an unquestionable and dramatic reduction in project duration. However, Confluence did not achieve faster projects simply by planning smaller projects. The summary statistics of the two data sets indicate that the average size of projects was unchanged throughout the interval covered by the data. Mean project size was 92 staff-days, before the change in the company's management process. It was 96 staff-days after the change in the management process, essentially the same. Therefore, the huge reduction in project duration was the direct result of the highly focused effort that each project now enjoyed. The new management process, of course, permitted that highly focused effort to take place.


Figure 1: The leadership team of Confluence achieved a dramatic reduction in project duration, by focusing resource capacity rather than diluting capacity on too many active projects.

Notice the two data points at the very top of the figure. The blue data point in the upper-right of the figure indicates a project that required 534 staff-days of effort and lasted 328 business days. The white data point directly across indicates a project that required 537 staff-days of effort. The latter project was completed in only 85 business days. This represents an astounding 74% reduction in project duration. The intriguing trend line suggested by the data for larger projects (white dots), which indicates that project duration is a but a week function of project size, is the result of the company's strategy of focusing as much capacity as possible on each project.

Now, the question remains. What can the Department of Homeland Security and other government offices do, to move the nation's defense contractors in this direction? In addition to eliminating the two obstacles discussed in my earlier articles, the answer is quite a bit. Once a specific contractor location has been targeted for improvement, and the projects at that location have been prioritized and sequenced, the government can achieve the highly desirable focus of capacity with a most powerful tool already in its possession, the budget. By simply shifting budget dollars from the lower priority projects to the highest priority project, the government can cause a corresponding shift in the contractor's allocation of resources. Should any government office object to such a shift in budget dollars, one need only ask why capacity that can be applied effectively to the highest priority project (from a homeland security perspective) should be spent on a lower priority project. The answer, of course, is that it should not.

Ideally, the contractor's executive team would cooperate with the government's effort to focus resource capacity. In fact, an executive team that put the nation's security ahead of its own short-term profits would champion the change effort rather than have it forced upon them by the government. However, while the cooperation of the contractors' leadership teams might be desirable, it is not essential. Much of the current dilution of capacity has been caused (and continues to be caused) by the government's existing policies and practices. The contractors are merely doing their best to optimize financial performance within the financial eco-system that the government has created.

By now some readers may be asking an important question: does this approach really cause the completion of more projects in the same time and with the same resources? Look for the answer to this question in the next article.

=========================================
Getting To Speed - a web-based presentation.

1) Discover how much more you can achieve with the resources currently on your payroll;

2) Learn which few but highly effective changes to make;

3) Understand the link between your policies and the performance of your product development enterprise;

4) Learn which new policies to deploy, for greatly enhanced logistical and financial performance.

Access to this webcast is limited. Click here for information.
=========================================

[So that others might also subscribe, please share the following link with friends and colleagues._Subscribe!]

[5] A Practicable Holistic Strategy

[So that others might also subscribe, please share the following link with friends and colleagues.Subscribe!]

Holistic Contractor Management

Currently, defense contractors treat all government-funded development projects as if they were of equal importance to the national defense and of equal urgency. As a result, defense contractors allocate to each active project a perceived fair-share of development capacity. Worse, with each new project that any government agency decides to fund, defense contractors dilute the level of capacity available to earlier projects, so that they might staff the new project with its own fair-share of capacity. They do this, because they are unable to bring onto their payrolls a sufficiently large number of new workers fast enough to satisfy the unrealistic start-dates that win them the contracts in the first place.

Given the current threat to the security of the United States, this fair-share approach to the allocation of resources is itself a threat to homeland security. It delays the few projects that are made truly urgent by the threat of terrorism and related military operations, while favoring projects that should be given a lower level of priority today.

The ideal situation would be to have all government-funded development projects prioritized and sequenced, 1 through n, across all services. This would enable the optimum deployment of all development resources, and it would ensure that the most urgent of all active projects would be completed at maximum speed. Indeed, a limited version of this is exactly what happens when the generals in the Pentagon decide that ONE project, perhaps needed to fulfill one urgent mission, is given overwhelming priority at every defense contractor in the nation. However, the magnitude of such a massive prioritization task and the considerable political interests preclude such an ideal situation from ever existing. There is an alternative that is practicable and nearly as effective.

In most cases, development projects are performed largely by resources at a single contractor location. Therefore, to enable a much higher yet sustainable level of speed, it is sufficient to prioritize and sequence the projects at each contractor location. This smaller prioritization task greatly reduces the number of government offices that have to cooperate with each other at any one time; it minimizes the number of political interests that have to be satisfied.

Does this more practical approach eliminate one of the two obstacles to speed? It does. Recall. The first obstacle is that the executive teams of defense contractors lack the authority with which to prioritize and sequence the government’s projects. If the government prioritizes and sequences its own projects, the executive teams then simply subordinate their organizations’ operations to the government’s prioritization decisions. However, the second obstacle remains.

As mentioned earlier, the second obstacle is created by the additive profit model of the defense contractors. Given their additive profit model, if defense contractors do subordinate development operations to the government’s prioritization decisions, their resource utilization level and (more importantly for the contractors) profitability will be reduced.

The additive profit model is defined by the government’s own policies, measurements, and practices, which determine the detailed features of the various development contracts. To eliminate this obstacle, contracts must be designed so as to reward the rapid, complete, and effective delivery of the tangible items specified by each contract. In other words, the contracts must be performance-based and not cost-based, as they are now. All contractor payments must be triggered solely by the satisfactory delivery of tangible items, so that the focus of the contractors' executive teams shifts, from keeping people busy on anything to delivering all that has been promised to the government as fast as possible.

These two changes can be achieved only by the government. Further, given the current threat to homeland security, the responsibility for creating these changes falls squarely on the shoulders of the Department of Homeland Security and Secretary Ridge (or his replacement). It is the responsibility of all the other government services to cooperate fully with Secretary Ridge, should he decide to take these vitally important steps, so that the national security of the United States might be guaranteed to the maximum degree possible.


A Strategy for Transition

Given the current, significant threat of terrorism, making these and related changes is of utmost urgency. Accordingly, the Department of Homeland Security should begin by first prioritizing and sequencing its own technology development projects. Then, the Department should strive first to achieve the prioritization of the projects at the contractor location where its own most urgent project is being performed. The Department’s own prioritization list should guide the sequence in which other contractor locations adopt this more effective operational model. For each of these locations, the Department should solicit the cooperation of all other government offices whose projects are being performed at that same location. Those offices, in turn, should be mindful of their own responsibility to subordinate themselves to the national security of the United States.

Without question, the transition will require a considerable degree of re-education, both, in the government and within each contractor location, beginning with the highest levels of management in each. The re-education of decisions-makers, managers, and ultimately resources should precede all other steps, within the Department and at each contractor location.


Summary

Secretary Ridge observes correctly that the current rate of technology development is the one factor that limits our ability to collect defense-related intelligence and to secure the national defense of the United States. However, the current pace of technology development is made unnecessarily lethargic by management methods that are, themselves, forced by the government and employed by defense contractors. A new, more effective management model exists, which approximately doubles the speed with which technology development projects are completed. All that remains if for Secretary Ridge to understand the new model and to deploy it rapidly and effectively.

The new management model will be the focus of future articles on this blog.


Every Reader’s Responsibility

If you agree with all that I’ve discussed in this document, then it’s up to you to help make Secretary Ridge and other decision-makers in government aware of all that the government can do, to improve our ability to provide for the national defense. To that end, I urge you to share this document either directly with government officials or with individuals who have the ability to gain the attention of government officials.

[So that others might also subscribe, please share the following link with friends and colleagues.Subscribe!]

[4] Two Obstacles

There exist two obstacles, which prevent all defense contractors from adopting the highly effective management practices used successfully by Confluence and the others mentioned in the previous article. The first obstacle is insurmountable by any executive officer of any defense contractor. This obstacle is the complete lack of authority with which to prioritize and sequence the government programs performed within a contractor facility, particularly if those programs are funded by different government offices. The authority required for this cross-program prioritization simply does not rest with any defense contractor. It rests solely with government representatives, who are the customers of the defense contractors.

The second obstacle also appears to be within reach of the executive teams of defense contractors at first. But, in fact, it too is created by the government and can be overcome by willing executive teams only for the briefest of intervals and only very occasionally. To gain a necessary appreciation of this second obstacle, we need to understand the primary profit model of the defense industry.

The primary profit model of the technology development operations of defense contractors today may be described as an additive model. That is, the profit generated by each defense contractor is directly proportional to the number of contractor personnel that charge their time to government contracts. I refer to these employees as profit-contributing employees, as opposed to profit-draining employees. A contractor’s profit for any interval increases linearly with the number of profit-contributing employees, and it decreases rapidly as the number of profit-draining employees rises. Profit-draining employees are those with no billable project work. Through no fault of their own, these employees are unable to charge some significant fraction of their available time to government contracts; their salaries and benefits are subtracted entirely from the value of the defense contractor that employs them.

This additive profit model strongly motivates defense contractors to maintain very high levels of workforce utilization, even when maintaining high utilization levels requires occasionally that the employment of some workers be terminated. To maintain the high utilization rates, the financial officers of virtually all defense contractors monitor religiously the number of “overhead hours” accumulated by every employee. Employees who are unfortunate enough to accumulate any significant number of overhead hours become identified as profit-draining employees, and they often find themselves at risk of termination.

The additive profit model is one of two phenomena that are at odds with each other. The second phenomenon, which is involved in a conflict of gargantuan proportions with the additive profit model, is the nature of work flow within organizations that perform projects. Specifically, the speed (or lack of speed) with which projects flow through any system of resources is attributable largely to the interactions among those resources. Since the delivered output of defense contractors is directly proportional to the rate of project completion (speed), the output delivered to government customers (systems vitally important for homeland defense and intelligence acquisition) is largely the result not of the additive contributions of contractor employees but of the interactions among those employees.

These two phenomena, the interaction-based speed of output generation and the additive profit model, define the overriding conflict within which the nation's defense contractors exist today. The conflict is illustrated in Figure 1, below. As that figure shows, there are four operating possibilities open to the executive teams of defense contractors. These possibilities are represented by the four quadrants. Quadrant 1 is defined by high utilization levels and low speed; quadrant 2 is defined by high utilization levels and high speed; quadrant 3 is defined by low utilization levels and high speed; and quadrant 4 is defined by low utilization levels and low speed.

Figure 1: The conflict between speed and the utilization-level of available resources defines the governing parameter space within which defense contractors exist today.

Of the four possible quadrants, only quadrant 1 represents a sustainable environment for defense contractors, for two reasons. First, quadrants 3 and 4, which are defined by lower levels of resource utilization, are in direct conflict with the additive profit model of defense contractors. Given the additive profit model, defense contractors that are forced to exist within either of these two quadrants for extended periods lose money and risk bankruptcy. This leaves quadrants 1 and 2 as the sole remaining possibilities.

Second, the existence of defense contractors within quadrant 2 is precluded by the laws of physics. Specifically, queuing theory dictates that as the utilization level of any system approaches 100%, the queue of work before the system of resources increases without bound, and consequently the speed with which work flows through the system of resources approaches zero. Figure 2 illustrates this relationship between speed and utilization level. Therefore, the only logical solution to this widespread conflict, for any defense contractor today, is to operate profitably within quadrant 1, as illustrated further down by Figure 3.

Figure 2: Queuing theory dictates that maximum speed and maximum utilization are mutually exclusive conditions. Consequently, the executive teams of defense contractors are forced to choose between maximizing speed and maximizing utilization level. The additive profit-model, which is created by the current design of defense contracts, guarantees that the contractors always choose to maximize the utilization level of their workforces, since this also maximizes profitability.

Figure 3: The additive profit model guarantees bankruptcy for defense contractors that try to exist for any extended period within quadrant 3 or 4.

Queuing theory guarantees that no defense contractor, indeed, no product development organization, can exist within quadrant 2. Therefore, quadrant 1 provides the only possibility of sustained profitability for defense contractors. Indeed, today virtually all defense contractors strive to maintain maximum utilization of their workforces.

One powerful direction for improvement is for the government to adopt a more holistic strategy toward contractor management.
However, any solution sought in this direction must include two necessary (though insufficient) components. First, the government must define the cross-program prioritization that the defense contractors need for the more effective management of their development resources. Ideally, the cross-program prioritization defined by the government would be consistent with the government’s own strategy for national defense and homeland security.

Second, the government must resolve the conflict that today drives defense contractors to operate under the conditions defined by quadrant 1: maximum workforce utilization at the expense of speed. By designing new contracts in such a way as to significantly reward speed, perhaps while also penalizing grossly deficient schedule performance, the government can consistently motivate the executive teams of the defense contractors to shift their operations from quadrant 1 to quadrant 3, as illustrated in Figure 4, below.

Figure 4: Defense contractors are not inclined to move their operations from quadrant 1 to quadrant 3, given their ongoing conflict between speed of output generation and profitability. Their conflict must be resolved first by the government, in such a way as to make it more profitable for the defense contractors to operate in quadrant 3 than in quadrant 1.

The requirements for sustainable and effective multi-project management do not exist anywhere within the defense sector today. Further, without these requirements, any attempt to improve the multi-project logistics or even the single-project logistics of defense programs is doomed from the start.

A strategy with which the government can overcome the two obstacles and begin creating the requirements for effective multiproject management is the focus of the next article.

[3] The Opportunity

Experience demonstrates that the rate, at which projects are completed today by product development organizations, is no more than 50% of what that rate could be. Evidence that supports this hypothesis is provided by Figure 1, below, which presents the performance data supplied by Confluence, a software development company located in Pittsburgh.

The vertical axis of the figure below denotes project duration, measured in business days. The horizontal axis of the figure shows simply the calendar for the years 2000 through 2003. Confluence reports that during the year 2000 and during most of 2001, before that company’s implementation of the Theory of Constraints Multi-Project Management Method, the average duration of the company’s projects was 140 business days. After the company’s implementation of the Theory of Constraints Multi-Project Management Method, the average duration of the company’s projects was reduced to 43 days. This indicates a tremendous increase in the project throughput provided by the company’s resources. The nature and scope of the company’s projects did not change throughout this period. In addition, Confluence maintained a constant workforce throughout the entire period indicated in Figure 1. Indeed, today Kirk Botula, the company’s Chief Operating Officer, reports that during the implementation period his company actually lost a few of its developers.




Confluence is but one of a number of organizations that report noteworthy gains in speed within their product development operations or similar multi-project operations (such as IT), as a result of streamlining the logistics (flow of work) during project execution. These organizations include Lucent Technologies Optical Cable Systems[i], Hampton Conservatories Ltd[ii], LSI Logic’s Design Technology Development Group[iii], Pharmacia Drug Development and Clinical Trials Division[iv], Seagate Technologies – Hard Disk Drives Division[v], and the United States Air Force[vi].

The reports from Confluence and the numerous others suggest a significant opportunity for greatly increasing the speed of project execution, for the programs funded by the government. It is my hypothesis that these successes can be duplicated within the defense sector, provided the right changes are made first by the related government organizations and subsequently by the contractor organizations. Therefore, it remains for us to determine how the specific steps, with which Confluence changed its management methods and operations, can be duplicated in the defense sector.

Acting on the advice of the Product Development Institute, the leadership team of Confluence took the following actions:

1) The leadership team created (and uses to this day) a process with which all projects were prioritized and sequenced.

2) The leadership team required the use of an effective project planning process (Robust Project Planning) with which the logistics of individual projects could be optimized up front.

3) The leadership team eliminated the earlier measurements of individual performance and project -schedule performance, replacing these instead with real-time estimates of the risk that the company’s projects might overrun their commitment dates. This was a more advanced version of the practice known as buffer management.

Any one of these steps, by itself, might have provided some minor improvement. It is my opinion that the revolutionary performance increase demonstrated by Confluence is attributable largely to an interaction among all three steps. To be sure, the full implementation involved a number of additional changes at all levels of the organization. But these three steps were the senior management actions that caused the day-to-day decisions and behaviors of middle managers and of resources to change and consequently to improve significantly the schedule performance of all projects.

This brings us to the question at hand. While all the other minor changes undertaken by the personnel of Confluence can be implemented readily by the corresponding personnel of any defense contractor, can these three enabling steps, which were undertaken by the leadership team of Confluence, also be applied to the product development operations of defense contractors? At first, the answer would appear to be “yes.” However, there exist two significant obstacles to achieving this objective. These obstacles are the subject of the next article.

Footnotes:

[i] William Baron, Management Roundtable, Executive Summit with Dr. E. M. Goldratt, April, 2000. Baron’s business developed new leading edge optical cables for telecom industry. For Premise Cable products - 100% of projects completed on-time (Compared to 40% before); 8 projects completed before 50/50 points and New Product Introduction (NPI) lead times reduced 50%. For outside cables - NPI lead times reduced 50% and tripled development capacity (16 projects vs. 5 before, with no staff increase).
[ii] Michael Dinham, TOC Reference Bank (toc-goldratt.com), 2002. Hampton Conservatories Ltd. designs, manufactures and installs high end custom hard wood conservatories for the commercial and domestic markets. Long lead times and poor due date performance were blocking the company’s growth. Improved the due date performance from 60% to 90% while reducing overall project lead times by over 50%. Throughput increased nearly 100%, while projects in progress dropped from 10 to 3. Project completion dates are now highly predictable and quality is also improved significantly.

[iii] Management Round Table Symposium hosted by IBM Global Services, March 2002. LSI Logic’s Design Technology Development Group has a few hundred engineers, producing technologies for designing chips. These technologies have to be available on promised dates so that the company can meet customer requirements. Over 90% of projects in the library development group now finish within 2 weeks of planned dates. Throughput has also increased by more than 10%, despite reduction in total number of resources due to reorganization.

[iv] Management Round Table Symposium hosted by IBM Global Services, March 2002. Pharmacia implemented the Critical Chain method within the clinical supplies area, a small but vital link in the drug development process. In a very short time, lead time was reduced over 60%, due-date delivery improved to over 90% and throughput increased. Package rate was increased from 20 per month to 50 per month (a throughput increase of over150%).

[v] TOC World Conference, 2000. Seagate Technologies – Hard Disk Drives Division, with the guidance of the team from AGI, the Core Team working on the Cheetah X-15 disk drive project beat their target date by nearly five weeks! Five weeks in their industry is extremely valuable. A 1997 internal case study done at Seagate showed that when the company lagged its competition by just one quarter, Seagate missed $500 million in incremental revenue opportunity and about $200 million in gross margin. The Cheetah X-15 was the first 15,000 rpm disk drive to ever hit the market. All of Seagate’s competitors dropped out due to technological challenges or being late to market.

[vi] Bob Casey, David K. Christ, Jim Pitstick, TOC World Conference 2001. “Scrap and rework on the F-22 fighter program has been remarkably lower than any previous fighter program. The F-22 is meeting and exceeding its testing and affordability goals.” These results were presented by Bob Casey (F-22 Weapons System Integration and Test, Lockheed Martin Aero), David K. Christ (F-22 Manufacturing Plans, The Boeing Company), and Jim Pitstick (Industrial Engineer, F-22 Special Program Office) at TOC World® 2001.

[2] The Effect of the Three Policies

In the previous article I discussed the three policies that constitute business as usual for defense contractors. The policies are:

1) The most important measurement for all managers and all resources is the number of overhead hours accumulated by individuals, groups, and departments. This measurement is to be minimized at all times, or jobs are lost.

2) The number of active projects is determined not by the capacity of the organization but by the number of successful proposals, which almost always exceeds the capacity of the organization.

3) All active projects must demonstrate progress continuously.

Now I want to describe how these policies do their damage.

From the perspective of the Department of Homeland Security, the best situation would exist if the development projects performed by defense contractors were completed as fast as possible, always. This would deliver the greatest number of defense systems to the Department, in the earliest practicable time.

In order for maximum speed to be achieved and sustained by a multiproject operation, like those of the defense contractors, two conditions must exist. First, every developer must know which of several tasks in the developer's in-box should be completed first. After identifying the most urgent task among all the tasks in his/her in-box, every developer must perform that task at a full level of effort and without unnecessary interruptions. The policies listed above prevent these two necessary conditions for speed from being achieved.

The first policy forces resources to stay busy at all times, so that their "overhead" measurement can be maintained at or very close to zero. Keeping busy is a very easy thing to do, if one stops trying to perform the right task and simply takes on any available task. Keeping busy is even easier to do when the second policy is at work, since the second policy guarantees that there is always too much work in process within the organization. Thus, the first two policies create the stage for the complete destruction of task-level prioritization and speed of project execution.

The destruction of task-level prioritization and speed is guaranteed with the third policy: All active projects must demonstrate progress continuously. The extreme dilution of resource capacity and widespread multitasking are the only ways in which this third policy can be met by developers. The extreme dilution of resource capacity guarantees that every project sees only a small fraction of the capacity that it might otherwise utilize effectively. Thus, the duration of each project is significantly extended, unnecessarily. The widespread multitasking simply takes this dilution solution to the totally absurd extreme where even the capacity of a single individual is shared by multiple tasks and multiple projects. The effect created by these three policies, therefore, is to decrease the speed of project execution dramatically.

That speed is reduced dramatically is beyond question at this point. I have seen enough multiproject organizations, suffering the ill-effects of the three policies, to conclude that the policies and the damage that they inflict are essentially global. They are at work in every defense contractor, worldwide. And they are at work in every product development organization as well, worldwide.

The policies exist for specific reasons. These reasons are different for companies in the commercial sector than for companies in the defense industry. In the next article I will discuss why the policies exist in the defense industry. More importantly, I will discuss what the government can do to eliminate these policies and to increase by a factor of two the speed with which technology development projects are completed.

[1] The Problem

In a c-span interview Tom Ridge observed that the limiting factor for homeland defense at this time was the pace at which technology could be developed. The implication was that Ridge and his newly formed department could do a more effective job of defending the US, if the pace of technology development could be accelerated. I take Mr. Ridge at his word. If the pace of technology development is the limiting factor for more effective homeland security, then it makes good sense that we do everything possible or at least practical, to speed up the pace with which homeland defense technololy is developed.

I know how to double the pace with which technology development takes place.

"Why write a blog?" one might ask. I write this blog, because I have no means of getting my knowledge into the brains of Tom Ridge and other decisions-makers in the Department of Homeland Security. My hope is that some of the readers of this blog have the means with which to alert Mr. Ridge et. al. of the possibilities. Once alerted, if Mr. Ridge is true to his word, I would expect him to take action and implement the changes that I discuss here.

What is the problem? The problem is that current pace of project completion is lethargic at best, for all defense contractors. If the nation's defense against terrorism depends in large part on our ability to complete defense acquisition projects, as Mr. Ridge states, then we're are at risk, and we will remain at risk unnecessarily for years to come.

Consider this. Whenever the decision-maker of a company feels the squeeze of the proverbial vice upon his more spherical organs, because some project is late, he (certainly not she if spherical organs are at risk) always responds by making the troublesome project priority-one for everyone whose paycheck requires his signature.

The effect of the step-increase in the priority of the troublesome project is that the project experiences two outcomes. First, the project is given all the resources that it can utilize effectively. Second, for each of the resources assigned to the project, newly assigned or not, the tasks of that project become of the highest priority. The result is that the project enjoys an outstanding level of focus; it virtually flies through the organization; and it is completed in record time.

So why can't all projects be completed just as fast? The answer is simple. After the pain-generating project is done, the decision-maker and his entire organization go back to business as usual. For defense contractors, business as usual means that the following polices and measurements are in full effect:

1) The most important measurement for all managers and all resources is the number of overhead hours accumulated by individuals, groups, and departments. This measurement is to be minimized at all times, or jobs are lost.

2) The number of active projects is determined not by the capacity of the organization but by the number of successful proposals, which almost always exceeds the capacity of the organization.

3) All active projects must demonstrate progress continuously.

So that we can understand the full depth of the problem, I will discuss in the next three articles how these policies limit the schedule performance of defense contractors today. In subsequent articles, I will discuss the policies imposed upon defense contractors by the government, which cause the defense contractors to live or die by the three policies above.