Abstract
Contrary to popular opinion, Earned Value is not all about numbers and formulas; they are just the means to an end. An Earned Value Management System (EVMS) is about integrating the scope, schedule and cost components of a project into a single baseline that can then be used to report status, define problems, take corrective action, and forecast final results. At its heart an EVMS is a control system, containing the essential elements of any such device; baseline, measure, compare, and correct. When properly executed it helps insure that planning, controlling, data collection, analysis, and reporting are done according to best practices. This paper will lay out the key concepts of an EVMS that will be applied during the hands-on simulation.
Introduction
Earned Value Management (EVM) has its roots in manufacturing. In a fixed, stable environment it is possible to set manufacturing baselines that are derived from creating the same product multiple times. Extremely mature and accurate manufacturing control systems can be developed when millions or even billions of the same deliverables are produced. Applying a similar system in a project environment presents special challenges.
▪ Projects are temporary, unique ventures. Manufacturing makes the same thing over and over, offering huge opportunities for data collection, correction and lessons learned.
▪ Projects often involve “knowledge work” to create processes or results, not just products. Manufacturing generally creates tangible, measurable products.
Since all projects create a unique deliverable, there is limited opportunity to benefit from other projects’ products. The project manager must strive to identify previous similar projects and adapt what they can from baselines and performance history for EVM.
Knowledge work can be difficult to divide into discrete elements that can be budgeted and scheduled. Any project, even pure R&D efforts, can be decomposed, if the project manager considers the steps involved in the work, and the deliverables that will be generated.
Control System
Earned Value Management is best thought of as a control system for projects. Any control system, no matter how simple (e.g., a thermostat) or complex (an EVMS), consists of four components.
▪ Baseline
▪ Measure
▪ Compare
▪ Correct
An EVMS is a closed-loop control system, where feedback is continuously provided to keep performance near the baseline. Open-loop systems are simpler because they involve only a single output. Common examples of open-loop systems include alarm clocks and toaster ovens.
Baseline
A planned state must be set as the target for the control system.
▪ A room thermostat has a baseline equal to the desired temperature.
▪ The Performance Measurement Baseline (PMB) is an integration of the project’s plan documents that detail the project’s scope, schedule and cost.
Measure
Control systems measure and report the current state of the system.
▪ Thermostats measure a room’s temperature and report it as a number.
▪ An EVMS measures the amount of scope accomplished at any point in time as a dollar value.
▪ An EVMS also measures the actual cost to create the scope.
Compare
The actual state of the system as measured must be compared to the baseline and the variance (any difference between what actually occurs and what is planned) determined.
▪ Thermostats compare the set number to the actual temperature in the room.
▪ Earned Value Management formulas compare the value of the work completed to the value planned, and to the costs incurred.
Note that all of the EVM formulas serve simply as the compare mechanism for the project. Just as compare is only one component of a control system, formulas and numbers are only one component of an EVMS.
Correct
If the variance is outside allowable limits (control tolerances) the system takes corrective action.
▪ A thermostat adjusts the heating or cooling to bring the room temperature to the baseline.
▪ The project team must first determine the cause of the variance, and then decide on and take the appropriate corrective action to bring the project performance back to the baseline.
Variances
Rarely is there an exact match between the planned baseline and the measured actual state. Variance is a normal and expected state in any control system, including an EVMS.
▪ A project with a positive or negative variance is not necessarily cause for alarm, depending on the size of the variance.
▪ An exact match of performance to plan can be expected only a statistically insignificant amount of the time.
▪ In fact, the truthfulness of a project manager who reports absolutely no variance can (and should) be questioned.
Tolerances
To function, a control system must set a limit (control tolerances or thresholds) on the allowable variances in the system.
▪ Control tolerances tell the system or project when it is necessary to respond to a variance.
▪ To effectively use a control system, tolerances both above and below the baseline must be set.
▪ The goal of a project is to perform to the baseline within the allowable tolerances.
▪ The project should deliver the planned scope, no less, and just as importantly, no more.
▪ A successful project delivers the defined scope as close as possible to the planned date and for the planned budget.
▪ A positive project variance is just as bad (or possibly worse) than a negative variance.
Setting Tolerances
Setting the right tolerances is critical to the efficiency of any control system, be it a thermostat or an EVMS.
Tolerances too tight cause the system to “thrash” or over-control.
▪ Heating and/or cooling cycles on/off constantly.
▪ Unnecessary corrections are made by the project manager.
Tolerances too loose allow too much “swing” or under-control.
▪ The room gets too hot or cold before the system reacts.
▪ The project gets too far off plan to be easily corrected.
Implications for Projects
▪ If too tight a tolerance is set, the system may report a lot of “false positives,” conditions that are flagged as problems, but are in fact not. For example, if the range is set to +/- 2% or even 5%, factors such as the timeliness, accuracy, or natural fluctuations of the data can cause a project team to address unnecessary issues.
▪ If too loose a tolerance is set, the system may not report true problem conditions early enough. For example, if the range is set to +/- 15% or 20%, performance may deteriorate so far before it is reported that the situation becomes unfixable.
▪ Setting the right tolerances will allow the project manager to manage by exception, that is, only analyze real problems and take corrective actions when truly necessary.
Reporting Project Performance
When management asks “How is the project doing?” they primarily want to hear about performance in two areas:
▪ Budget
▪ Schedule
Implied, but not usually stated is “for the amount of work (scope) accomplished”. Using earned value formulas explicitly factors in scope for both measures.
▪ Unlike standard percent of budget, or time-consumed metrics, EVM does not ignore (or assume) the physical work accomplished.
▪ Without this relationship to scope, numbers do little to reveal the true status of a project.
▪ The amount of scope accomplished is the only data element that is present in almost every EVM calculation.
What “On” Means
▪ On schedule or on budget does not mean an exact match to plan.
▪ It means that measured performance is within the established control tolerances for the baseline.
▪ An organization, program, or failing that, a project team needs to define what “on” means.
▪ When a project budget or schedule is reported as “on,” it is generally considered in control.
EVM Reporting
▪ Terms like over, ahead, on, behind, and under have broad meanings.
▪ Earned Value Management provides objective numeric measures of variances and efficiency.
▪ EVM can help forecast probable project outcomes.
Conclusion
Earned Value Management should not be viewed as calculating a set of numbers, or a contractual reporting requirement. Simply put, EVM is a control system for projects. Although it takes effort and discipline to implement properly, Earned Value Management equates to fundamental best practices in project management.
