Concept: In Earned Schedule, there are two distinct views of schedule performance: efficiency and adherence. The Schedule Performance Index for time (SPIt) measures the efficiency with which the schedule is being executed. It looks at the volume of work done. By contrast, the P-Factor and Schedule Adherence Index (SAI) look at the sequence in which work is done.
A close examination of P-Factor and SAI reveals that they also present two distinct views—this time, of schedule adherence. P-Factor measures how much the work done as of the ES time matches the planned sequence of delivery. SAI estimates how much rework is expected on work done ahead of time.
P-Factor can be measured directly through the results as of the ES time: the tasks completed as scheduled and the tasks that have not yet contributed the EV they were planned to contribute (the Impeded or Constrained tasks).
Tasks that potentially require rework can be identified directly as the ones earning value before they were planned to do so. But, how much Rework will actually be required is not directly accessible at the same time, as it is work that will be done in the future. The amount of rework that will be required is, instead, estimated using a mathematical model.
For a complete picture of schedule adherence, therefore, both the P-Factor and SAI are required.
Practice:The Schedule Performance Index for time (SPIt) measures the volume of work completed, where work is expressed as the value that is earned, i.e., the Earned Value (EV). The SPIt is calculated by the formula:
…where ES is the amount of schedule earned and AT is the Actual Time. The amount of schedule earned is the time at which the value currently earned should have been earned. Figure 1 is the customary representation of ES. (For details on interpreting Figure 1 below, click here.)
Figure 1
While SPIt focuses on how much work is done, P-Factor and SAI focus not only on what work is done but also when it is completed. That is, they express how well or poorly execution adheres to the plan. Figure 2 is often used to illustrate schedule adherence [1].
In Figure 2, the work Output (O/P) is represented by the bars. Each bar can hold a certain amount of value. As value is earned, the bar fills up (i.e., the shaded portion). At the Actual Time (AT), the value currently earned (EV) maps back to the time when it should have been earned. That time is the point on the Planned Value (PV) curve intersected by the horizontal dotted line. On Figure 2, this
is represented by the red line at Period 3.
The terminus of the value curves is the Budget at Completion (BAC). The terminus of the time line is the Planned Duration (PD). As mentioned earlier, the Earned Schedule time (ES) is the point on the timeline when the value that has been earned should have been earned.
Figure 2
Schedule Adherence (SA) measures how well the shaded portions cover the plan-bars.The P-Factor focuses on one set of the tasks in Figure 2, and SAI focuses on a separate set. That’s why the two metrics should be used together.
Why think that P-Factor and SAI focus on disjoint sets of tasks?
P-Factor measures tasks that not only have earned value as of AT but also have planned value as of ES. By contrast, SAI, or more precisely, its Rework component, measures tasks that not only have earned value as of AT but also have earned value after ES.
Figure 3 below illustrates the difference with invented data that fit the picture in Figure 2. [2]
Figure 3
Given that ES is exactly the end of Period 3 and the Actual Time is exactly the end of Period 4, the totals for each task are listed in Figure 3 under the headings PV, PV@ES, and EV@AT. The totals are followed by the difference between the EV at the Actual Time and the PV at the ES Time. The next column indicates how much EV is aligned with the scheduled delivery of value as of the ES Time. The final column indicates if the tasks are Impeded/Constrained, require Rework, or are on schedule (shaded).
Tasks 1 and 3 have EV@AT = PV@ES and a difference of 0. Such tasks are on schedule, and their value is included in the amount that adheres to the schedule.
Tasks 2, 4, and 6 have EV@AT < PV@ES and a negative difference. Such tasks are Impeded or Constrained and have fallen behind schedule. That has two effects. First, their incomplete value contributes to the ES Time falling earlier than the Actual Time, i.e., incomplete value opens a gap between the two times. Second, although incomplete, the portion of the task that has been done as of the Actual Time is still in sequence. That portion is added to the amount adhering to the schedule.
Finally,Tasks 5, 7, and 8 have EV@AT >PV@ES and a positive difference. Such tasks are being done prematurely and are therefore likely to incur Rework. The value delivered from those tasks falls in the gap between the ES Time and the Actual Time and is excluded from the total that adheres to the schedule. In the P-Factor calculation, these amounts are subtracted from the EV at the Actual Time.
As shown in Figure 3, the total amount of value planned as of the ES Time is 40. Most of that value has been earned in accordance with the schedule, even if some of it is incomplete or constrained. There is, however, value that has been earned prematurely and is not aligned with the schedule. It is identified by positive differences between EV@AT and PV@ES. In the example, the total is 7, and it is excluded from the aligned value.
Thus, 33 of the 40 units planned to be done as of the ES Time were delivered exactly as sequenced, and that makes the P-Factor equal 33/40 or 0.825.
The calculation can be expressed succinctly by the following formula:
…where EV denotes the value earned, PV denotes the value planned, and, most important, the subscripts “i” and “j” denote the target tasks. The target tasks for the PV is the set of tasks scheduled (S) for delivery (in full or in part) as of the ES Time. The target tasks for EV is a subset (“i”) of the “j” tasks, specifically, just the “j” tasks that have delivered value aligned (A) with the PVj as of the ES Time.
So, the formula reads as: the P-Factor equals the total value actually earned in alignment with the schedule divided by the total value that should have been earned as scheduled. In both cases, the value is as of the ES Time.
By contrast, the Rework that is likely to be incurred by premature delivery has not yet occurred and cannot be measured directly. Instead, it is estimated.
The estimate uses the P-Factor (or, more precisely, its complement, 1 – P) to identify the amount of earned value that is at risk, namely:
…where the total earned value as of the Actual Time, (i.e., EVi,) is multiplied by the complement of P. Given that P is all and only the earned value as of ES that matches the planned value as of ES, the complement of P is all the other earned value—that is, all the value earned after ES and as of AT.
Keep in mind, however, that only a portion of that earned value will ultimately result in rework. Furthermore, the fraction requiring rework declines throughout the project, partly because the project’s knowledge base grows and partly because there are fewer things that can go wrong.
Thus, the amount of Rework (R) is calculated as a fraction, call it f(r), of the earned value at risk, yielding the equation:
The fraction is modeled by the imposing formula [3]:
The effect of this formula is to reduce the amount of Rework likely to occur over the timeline of the project. This is desirable because as domain knowledge grows and number of work items shrinks, the probability of Rework falls.
Finally, R is normalized to the remaining work on the project. Doing so offsets the inevitable rise of P-Factor to a value of 1.0 at the end of a project. [4] Thus, SAI accurately gauges schedule adherence throughout the project lifecycle.
Conclusion
P-Factor is measured directly by the amount of value that is earned strictly in accordance with the schedule, i.e., value that is both planned and earned as of the ES time. SAI, or, more precisely, its progenitor, R, is estimated based on the amount of value that is likely to incur rework because it is completed out of sequence. Such value is earned after ES but as of AT. The two sets of value-bearing tasks are distinct and separate, i.e., as claimed at the start, they are disjoint.
For a complete picture of schedule adherence, therefore, we need measures of both sets of tasks--the P-Factor and the SAI.
Notes:
[1] See Lipke (2009).
[2] The table is from Lipke (2009) with the addition of the “Aligned EV” column.
[3] For an explanation of the formula, click here.
[4] The rise is inevitable because the value earned ultimately aligns with the value planned. Normalization offsets the rise by adjusting the amount of rework to reflect the shrinking pool of value to be earned.
References:
Lipke, W. (2009). Schedule Adherence…A Useful Measure for Project Management.The Measurable News, Issue 3.
Lipke, W. (2012). Schedule Adherence and Rework. CrossTalk, November-December.
Lipke, W. (2011b) Schedule Adherence and Rework. PM World Today, July.
Lipke, W. (2011a) Schedule Adherence and Rework. The Measurable News, Issue 1 (corrected version). |
