DYNAMIC CONTROL

Traditional project management in static environments is about building a detailed plan and then controlling the project according to the plan. In the dynamic environment, a plan rapidly becomes out of sync at a speed that makes it impractical to maintain. Static planning approaches need to be complemented with dynamic control approaches.

Static Project Control Using the Plan

Static project management, as defined by the various bodies of knowledge, is focused on what is known as behavior control as a way of directing and regulating actions from above (Williams, 2004). Project plans, policies, guidelines, and methodologies are developed as control tools in the belief that they will lead to the desired results. Adherence to these controls is monitored, and deviations are corrected. This works best if a well understood and stable plan can be created, so its effectiveness is dependent on what is described as “task programmability” often not available in dynamic environments (Kirsch, 1996). Behavior control can be less helpful when such controls are difficult to create. Consider how implausible it would be to write instructions for the creation of a unique work of art, or groundbreaking research. Inappropriate controls can result in unexpected and counterproductive behavior. Even if the work is plannable, the manager may lack the knowledge or experience to develop the right controls. In order to achieve the measured objectives of the control (plan), workers with limited time may bypass other less measurable, but more important, objectives. If the control is flawed—even if the employee can see it is flawed—it may be difficult to correct if their performance is measured by adherence to the plan.

Furthermore, excessive behavior control can reduce staff productivity (Bonner, Ruekert, & Walker, 2002). Enforcement of excessive control can offend workers, thereby affecting morale or restricting creativity (Bonner, Ruekert, & Walker, 2002). Burdening workers with onerous processes and few incentives could discourage adaptation to a dynamic environment. Lastly, the cost of surveillance can simply be greater than the benefits of adherence (Sugden, 2001; Sachs & Meditz, 1979, p. 1081; Ashton, Johnson, & Cook, 1990; Williams, 2004).

Fortunately, there are a number of other control approaches available that are not discussed in the various bodies of knowledge.

Input and Output Control

Another form of control is output or outcome control (Snell, 1992). The desired outcome and benefits are defined more clearly than usual, but the path required to get there (e.g., project plan) is allowed less detail, especially in the early stages, enabling the project team to make use of their experience and professional judgment to adapt to a dynamic environment in order to optimize the result. In a dynamic environment of rapid change and many unknowns, tight behavior controls might be inaccurate and counterproductive, so a combination of looser behavior controls and tighter output control can serve well. For instance, while there is no way to define the precise steps required for a scientist to produce groundbreaking research, produce it they do. For scientists the goals are often well defined (e.g., the need to cure a disease), and rewards are available to guide and motivate the researcher to the desired outcome (e.g., peer recognition, public good, or Nobel Prize). A real estate agent is well motivated by the prospect of a sale. A leader can motivate a team by announcing a crisis (e.g., consider the motivation behind the Manhattan Project).

One of the dangers of output control is that mistakes are harder to prevent early, as they may not be discovered until the output is produced and measured. Another problem is that sometimes outputs can be difficult to measure. An example might be trying to measure improved morale. However, for project management in dynamic environments, a clear statement of the desired goals, benefits, and vision, with clear incentives for motivation and milestone deliverables for progress, recognition will often be an important complement to dynamic planning.

Another method of control to complement the plan is input control (Snell, 1992; Eisenhardt, 1985). In dynamic environments, where defining behavior or measuring output can be difficult, the manager can select and induct staff, and train staff to have compatible skills, experience, values, motivation, and compliance with traditions (Ouchi, 1979; George & Jones, 2002; Eisenhardt, 1985). Snell (1992) described input control as relating to “the knowledge, skills, abilities, values and motives” of employees (p. 297). So while a university cannot develop a detailed procedure that guarantees groundbreaking research, and may not even be able to predict exactly which research results might be achievable, a university can, for instance, select academics with a track record of achievement. The “science of sales” may be elusive, but an agency can have success selecting sales professionals with a successful sales history. An advertising company can provide training and induction and then allow staff freedom to achieve.

According to George and Jones (2002, p. 549), although it would be viable to create detailed procedures for simple roles like a warehouse picker, it would be much harder for a very complex and subtle role as the foreman's position. A better approach might be to select foremen for the job who have previously demonstrated a high-level commitment to the organization's objectives. Input control minimizes “divergence of preference,” thereby enhancing the ability of employees to work together (Eisenhardt, 1985). The same can be applied to project work. Rather than attempting to control staff with a detailed and ultimately inaccurate and unwieldy project plan, it can be better to invest more time in selecting staff who have experience with the work and demonstrate a commitment to achieving the company's objectives.

Figure 4.1 provides guidance on the balance between input, process, and output control. It might be wise for the project manager to adjust the mix of controls according to the viability of each one. For instance, academia has evolved to have a mix of input control (selecting academics with a track record), lower levels of process control (to give them freedom), and higher levels of output control, in the form of self-satisfaction, and recognition for publications and discoveries. Careful selection of controls may involve an acceptance that managers have less control than they would like, but that it is better to optimize than focus on a single unrealistic approach.

Diagnostic, Belief, Interactive, and Boundary

Robert Simons (1995a) wrote about the difficulty of controlling work in organizations that demand flexibility, innovation, and creativity. He described four types of management control that are particularly relevant for rapidly changing environments. He likened diagnostic controls to the instruments in an airplane cockpit, enabling the pilot to monitor critical performance indicators, and make timely adjustment of the controls (Simons, 1995b). Diagnostic controls summarize activities to give a snapshot indicator of progress toward goals, and correct deviations to keep performance within limits. As with behavior controls, care must be taken in deciding what to measure and present, and they only work when there is a known goal and measurable performance.

Simons described belief systems as another control approach where companies articulate aspirational values that can be used in decision making in unplannable situations. Belief systems are typically value-laden and inspirational, including principles such as open communications or the pursuit of value. Beliefs systems are an important part of the mix of controls to lean on in the absence of a crystal ball.

Interactive controls involve data that are interpreted and discussed face-to-face in a meeting of experts. This interactive forum is also used to debate and revise assumptions, and action plans with a focus on the final vision. Interactive control systems specifically track the most significant uncertainties, such as changes in technology, government regulations, customer tastes, or industry competition (Simons, 1995a).

Finally, Simons (1995a) describes boundary control systems that allow innovation within set limits; they might include codes of conduct, workplace health and safety regulations, gender equity, and anti-racism regulations. If a belief system guides with a carrot, boundary controls guide with a stick. Like road rules for drivers, boundaries allow freedom to explore within clear limits. In static environments, it makes sense to have tight boundaries, while in dynamic environments, excessive rules will restrict adaptability. Boundary controls are useful for projects with many unknowns as a way of allowing staff flexibility of behavior within reasonable boundaries.

Analogy – Driving to a Destination

To illustrate the different control approaches, consider which controls are used to drive to a friend's house for the first time to visit. Behavior control would be a detailed map showing exactly which streets to drive down. For most situations a map would be fine, but in your dynamic environment you discover traffic jams, road works, and that your friend has gone out. You really just want to catch up, so you negotiate a new plan to meet at a café (output control). You start with the benefit of having learned to drive, and deciding to use the car (input control). Being careful to follow the road rules (boundary control), and monitoring your speed and location (diagnostic controls), you drive, touching base with your friend on the phone occasionally to receive directions (interactive control).

Analogy – Delivering a Package Destination

Consider which controls you might use to deliver a package to a friend who works in a shop. Behavior control: You could draw a detailed map with the route and give that to someone to follow to deliver the package. Input control: You could seek someone who has been to the shop before and ask him to deliver the package. Output control: You could offer someone $100 to just find the shop and deliver the package. Interactive control: You could give someone directions over the phone as he drives to the shop. Boundary control: You tell someone the name of the shop and the shopping center it's in, and let him roam around inside to find it.

Examples of Dynamic Control

One of the research participants, a military campaign commander, gave an example of input control, describing how he preempted battle with intensified research and training. The planning engineer for road tunnel construction projects related an example of output control describing how staff were rewarded with a bonus when the project was ahead of schedule.

The project manager for a new energy storage research project described how staff-performance measurement was ineffective because there was no way to formulate a stable plan to compare with work behavior. Instead, she measured performance against milestone achievement, which gave the project team freedom to be creative and apply expertise within those milestone goals. To complement this, she motivated staff with an employee option plan (output control), where the team was granted options linked to a future liquidity event. She hoped this provided motivation for staff to apply themselves in the way they see best fits this goal, as they are subject matter experts beyond what our managers can be.

One of the project manager focus groups recommended giving the team objectives and constraints (output and boundary control) and taking the time to discover what motivates your team members to get the best performance out of them. Examples given of output control were typically in the form of recognition and skill extension; however, one military campaign commander reported how he took advantage of arguably the most powerful form of output control, that of survival, to motivate soldiers to come up with the right tactics. In training there was a greater emphasis on on-the-spot problem solving, in order to deal with unpredictable elements that occur during campaigns, rather than waiting for orders. They provide clear success indicators to measure goal achievement as follows:

In the orders they specifically say what constitutes success; for example, at the end of this operation I will have destroyed 30% of the armored force, so everyone is clear whether it's been successful or not, and work out alternate methods to achieve that.

All of the practitioner focus groups confirmed results from the in-depth interviews regarding the need to lower dependence on process control, in favor of other control approaches in dynamic environments.

Summary for Dynamic Control

In dynamic environments a larger variety of control techniques should be employed to reduce reliance on a detailed plan, since detailed predictions of the future are likely to be somewhat inaccurate. The plan should be expressed at a higher level, with more emphasis on desired benefits, and a series of efforts to explore realization of those benefits. More reliance is placed on input controls (staff and project selection, induction, training, etc.) and output controls (goals, rewards, visions). Behavior control in dynamic environments is more about creating and adjusting stage gates in a series of exploration efforts. Diagnostic controls should report on the results of each stage for the relative cost-benefit. Interactive controls facilitate discussions that monitor key changes in the environment, and review the validity of the original goals, and progress toward them. Decision making on the fly is facilitated by fixed boundary controls in the form of minimal rules and processes, and through belief systems that articulate stable guiding aspirational principles. For the purposes of this book, I will call these alternate controls guideline controls.