Why project control works so well for some projects and fails so miserably for others?
Controlling the performance of a project in progress is key to the success or failure of a project. It requires a set of metrics to measure the current progress and to forecast the future expected project behavior such that the project manager is able to timely detect project problems and take corrective actions to bring the project back on track. In order to be able to take high quality corrective actions, the project manager should set action thresholds that act as triggers to take these actions, which can be done in two ways:
A bottom-up control approach using schedule risk analysis
A top-down control approach using earned value management
Project control
In a bottom-up control approach, the lowest levels of the work breakdown structure (WBS) serve as static predictors of the impact changes might have on the project objective. More precisely, sensitivity information for each activity obtained from a schedule risk analysis (SRA) enables the project manager to distinguish between sensitive and insensitive activities. Action thresholds can then be set at a certain sensitivity value such that the activities with a higher sensitivity value than the action threshold value are said to be risky and are assumed to have a huge potential impact on the project objective. This limited set of activities requires an intensive control during project progress and might be subject to corrective actions in case these activities might endanger the project.
A top-down project control approach requires a dynamic knowledge of the overall project performance at a current moment in time, which is provided by earned value management (EVM) indicators such as the schedule and cost performance indices. These indicators serve as early warning signals and must trigger corrective actions when they drop below a predefined acceptable threshold. In this case, the project manager must drill down to the lower levels of the work breakdown structure to find the set of activities that are in trouble and that need to be put back on the right track.
Control efficiency
The action thresholds in both control approaches define the points of focus and attention of the project during project progress and serve as triggers to take corrective actions to bring project in trouble back on the right path. Consequently, in order to efficiently control projects in progress, the project manager must have access to reliable data provided by SRA or EVM metrics. This should enable the manager to focus on the activities in trouble in order to take high quality corrective actions that bring the project performance back to normal. Therefore, setting action thresholds involves a trade-off between the intensity of control at the lowest WBS levels and the potential impact actions might have on the overall project objective (highest WBS level).
Intensity of control: The value of the action thresholds defines the intensity of control and consequently the effort a project manager has to make during project progress. A low control intensity runs the risk of masking problems in certain activities in progress which will only be detected much later in the project life. A higher control intensity reduces the risk that problems in activities will not be detected but results in a higher effort for the project manager.
Bottom-up control: Low (high) thresholds result in a bigger (smaller) set of activities that are assumed to be sensitive and hence result in a higher (lower) intensity of control.
Top-down control: Low (high) thresholds mean that the probability to drill down decreases (increases) and hence this results in a lower (higher) intensity of control.
Impact of actions: When actions are taken in order to bring activities in progress back to their expected behavior, the project manager aims at bringing the project performance back on track as stipulated in the baseline schedule.
Bottom-up control: Corrective actions taken on very sensitive activities probably might have a bigger beneficial impact on the project objective than actions taken on activities which are assumed to have a lower impact. Selecting the right activities during a schedule risk analysis is therefore crucial in a bottom-up control approach.
Top-down control: Selecting the activities at the right moment is crucial to detect the activities in progress that really cause the project problems. Therefore, reliable performance data during project progress to drill down to the project activities at the crucial moment in the life of the project are essential for the quality of the corrective action decision making process.
The top-down approach using the schedule performance index SPI shows a relatively low efficiency. This can be mainly contributed to the low reliability of the SPI at the final stages of the project.
The top-down approach using the new schedule performance index SPI(t) is much more efficient, and this efficiency increases as the project becomes more and more serial.
The bottom-up approach using schedule risk analysis (SRA) has a high efficiency when projects contain a lot of activities in parallel.
The combined approach where both activity sensitivity information as well as project based performance indicators are used performs best and leads to a significant increase in the control efficiency.
Bottom-up project control: Setting action thresholds using schedule risk analysis
The interest in the sensitivity of project activities from both academics and practitioners lies in the need to focus a project manager’s attention on those activities that influence the (time and cost) performance of the project. When management has a certain indication of the potential impact of the various activities on the project objective, a better management focus and a more accurate response during project control should positively contribute to the overall performance of the project.
Schedule Risk Analysis (SRA) is a simple yet effective technique to connect the risk information of project activities to the baseline schedule, in order to provide sensitivity information of individual project activities to assess the potential impact of uncertainty on the final project duration and cost. A traditional schedule risk analysis requires four steps to report activity sensitivity measures that evaluate each activity’s time and cost estimate on a scale of risk.
These sensitivity measures can be used by the project manager to distinguish between risky and non-risky activities in order to focus on those activities that might have an impact on the overall project objective during the project progress.
Sensitivity measures – Risk classification for project activities
The output of a schedule risk analysis is a set of measures that define a degree of activity criticality and sensitivity. The values of the activity sensitivity measures give an indication of the potential impact of changes in the baseline activity durations and costs on the overall project objective (project duration or total cost). The values for the sensitivity measures for each activity need to be interpreted by the project manager such that they can be used to distinguish between risky and non-risky activities and to set action thresholds that act as a mechanism to take corrective actions in case the project in progress runs into trouble. Well known sensitivity measures used in a schedule risk analysis are the criticality index, the significance index, the schedule sensitivity index and the cruciality index, In the remainder of this article, a fictitious sensitivity measure is used that expresses the sensitivity of each activity on a scale between 0% (insensitive) to 100% (highly sensitive).
Setting action thresholds – Interpretation of the measures to restrict a project manager’s focus.
Action thresholds should be set by the project manager using the activity sensitivity information in order to put his/her focus on the most risky activities of the project. A predefined action threshold defines the intensity of control measured by the portion of project activities that require high attention. Consequently, knowledge of activity sensitivity information is crucial to steer a project manager’s attention towards a subset of the project activities that have a high expected impact on the overall project performance and objectives (time and cost). These highly sensitive activities are subject to more intensive control, while other less sensitive activities require less or no attention during project progress. In the illustrative figure 1, the set of activities is split into three classes, based on the values of the sensitivity measures obtained from a schedule risk analysis (the horizontal bars), as follows:
If “sensitivity measure < 30%” then “safe zone”: These activities are considered to be insensitive to the project objective and hence they do not require special attention from the project manager once the project is in progress.
If “30% ≤ sensitivity measure < 60%” then “watch out zone”: These activities are considered to have a low to moderate potential impact on the project objective. They do not require special actions, but a higher degree of attention during project progress could prevent future problems.
If “sensitivity measure ≥ 60%” then “dangerous zone”: These project activities are highly sensitive and are assumed to have a high potential impact on the project objective. These activities must be under constant attention of the project manager during their progress and require immediate corrective actions in case these activities show deviations from their baseline values.
Bottom-up project control – A project control approach.
A work breakdown structure (WBS) of the project which splits the project objective into smaller pieces. The bottom-up control approach on the sets of activities and the impact of actions on the overall project objective. Since the action threshold(s) define(s) the intensity of control at the lowest level of the WBS (activity level) in order to control and manage the overall project objective at the highest WBS level, this approach is referred to as bottom-up project control. Since the set of project activities is split into various classes where each class has its own intensity of control, this approach aims at reducing the project manager’s effort such that he/she can focus on only the risky activities that might have an impact on the project objective.
Research studies have shown that the bottom-up activity based control approach using sensitivity information of activities obtained through a standard schedule risk analysis is particularly useful when projects contain a lot of parallel activities.
Top-down project control: Setting action thresholds using earned value management
Measuring and controlling the performance of a project in progress is key to the success or failure of a project. It requires a set of metrics to measure the current progress and to forecast the future expected project behavior such that the project manager is able to timely detect project problems and take corrective actions to bring the project back on track.
Earned Value Management (EVM) provides such performance measures at periodic time periods to compare the initial baseline schedule with the current project performance. It is based on three key metrics and simple formulas and provides the project manager with a set of measures that act as an early warning signal to detect project problems in an early stage. In this article, the following three topics will be briefly explained:
Earned value management: Controlling the performance of projects in progress
Action threshold(s): Interpretation of the performance measures as early warnings for action
Top-down project control: A project control approach
Earned value management
Earned value management is a project control technique to measure the performance of projects in progress using three key metrics. The key metrics are used to measure the (time and cost) performance at the current moment in time where the project manager is right in the middle of the project progress and still has room to take corrective actions in case the project is in trouble. The current project performance can be evaluated by the following performance measures: two versions of the schedule performance index (SPI or SPI(t)) to measure the current time performance and the cost performance index (CPI) to measure the current cost performance. In the remainder of this article, the performance of a project in progress over time will be measured on fictitious project data with a duration of 11 weeks using a performance measure that reports the value 1 if the project is on schedule (SPI or SPI(t)) or on cost (CPI).
Setting action thresholds
The EVM performance measures that are periodically reported to the project manager must act as early warning signals to timely detect project problems and to trigger actions in case the project performance runs out of control. Consequently, they provide the project manager with useful information of the overall project performance at a certain moment in time, but they do not provide detailed information about the performance of the individual project activities. This action threshold defines the intensity of control. When the project performance measure drops below the action threshold at a certain moment, it is assumed that the current performance is too far away from the initial expectations which might put the overall project success in danger. Consequently, the project performance must act as a trigger to take corrective actions in order to bring the project performance back on track.
Action threshold setting to distinguish between risky and non-risky activities
Top-down project control
A work breakdown structure (WBS) of the project which splits the project objective into smaller pieces. The top-down control approach where project based performance measures give an indication of the current performance that act as a trigger to drill down to the lowest WBS levels to find out where the problems really are.
Research studies have shown that the top-down project based control approach using schedule and cost performance metrics obtained by using earned value management is particularly useful when projects contain a lot of serial activities.