EVM: Project Management with the Lights On!
Project managers are expected to know the progress of their project at all times. Are you meeting expectations? Staying within budget? Staying on schedule? These can be tough questions to answer without the use of Earned Value Management (EVM).
Fortunately, EVM doesn’t require a different approach to project planning and management. Rather, it extracts useful information from planning work that you’re routinely doing.
The Need for EVM
If every project went according to plan, there would be no need for EVM.
EVM serves to illustrate the difference between what was planned and what is actually happening. It’s an early warning system that alerts management to the realities of project performance.
In essence, EVM creates a performance “contract” for a project. It clarifies exactly what was expected to happen before the project was launched, and then measures whether those expectations are coming true during execution.
The problem with most contracts is that sellers (in this case, project managers) don’t want to commit to a fixed price. They want to spend money as necessary in order to get the work done.
On the other hand, the buyers (members of senior management) want to know exactly what their money will be purchasing. They want a firm commitment from the project manager.
EVM helps to solve this conflict by providing a firm estimate of performance (developed in the planning stage) and then generating interim measures of whether that estimate is being realized.
History of EVM: A Hundred Years of Evolution
Earned value is basic cost accounting applied to the one-time events we call projects.
The practices we now call EVM developed out of cost management techniques used in large factories as early as the 1800s. Industrial engineers of the time compared planned vs. actual output, timing, and cost to provide a picture of performance relative to expectations. Modern EVM does exactly the same thing.
Today, EVM has returned to its roots in cost accounting, but has gone high tech regarding the use of computers and statistical analysis. The new EVM is efficient, effective, and easy to use. It is intended for general use by project managers, working on any type of project, where senior management wants to be kept informed of project progress, not just the final outcome.
The Role of EVM: Monitoring Projects
EVM determines how much of a project has been completed at specific points in time, known as milestones (more on those later). Knowing how much has been completed allows senior management to release funds in small increments and see if they are getting value for their money.
EVM allows senior management to monitor progress and to react to poor performance. Using EVM, senior management can tell early in the life of a project whether it is likely to meet its targets. Management can then decide whether to abandon the project early on, before a huge amount of money has been spent. Likewise, if senior management’s expectations are unrealistic, EVM will quickly highlight the problem.
For the project manager, EVM changes the emphasis from performance targets way off in the distance, to targets that are coming up near term. The project manager knows how the project is performing at any given time and whether or not management will be (or should be) pleased.
Measuring the performance of individual tasks rather than the project as a whole gives both the project manager and senior management a steady stream of signals about the health of the project.
Milestones: The Key to EVM
Established during planning, milestones are a fundamental strategic tool used to subdivide a project’s work effort. They’re created with the express purpose of indicating how much work should have been completed as of a given date.
In order to serve as progress measures for controlling projects, milestones must be defined in three dimensions:
- Clearly quantified work: there can be no confusion about whether or not a task has been completed
- Allocated resources: indicated by either time or dollars spent
- Completion date: can be natural (the end of a contract), artificial (every Friday), or based on a need to measure progress as of a certain date/expenditure (as of May 14 the painting will be complete at a cost of $1,500)
Like alarm clocks, milestones tell management that something should have been completed by a certain date and cost. If it has not been done, something is wrong; there has been an exception to the plan.
Milestone timing depends on the reporting needs of the project. However, the shorter the time between milestones, the lower the risk that the project will not deliver the expected value.
EVM Project Planning
EVM is all about planning. EVM starts with your existing project plan and then breaks it down into tasks that have individually planned expectations of scope, budget, and time frame. These are then used to build a resource-loaded schedule that relates expenditures to work and time.
Rewards (payments and continued funding of a project) are related to a cumulative measure of performance based on how well tasks are being completed as compared with planned performance. When those expectations are not met, this is a signal for management to investigate.
The Performance Measurement Baseline (PMB) documents expectations for the project based on a schedule that relates the work (scope) and resources (money) to a specific date range. With a PMB it is possible to compare actual performance to what was planned for specific points in time. The difference between actual and planned is then used to forecast probable completion costs and schedule.
While the structure and style will vary according to your organization’s preferences, there are three key components of a PMB:
WBS (Work Breakdown Structure): Either a WBS or an equivalent analysis of all project work.
Resource-loaded schedule: This is typically a Gantt chart indicating when tasks are to be completed and their associated budget, with milestone markers illustrating control points.
Budget graph: Summarizes the timing and magnitude of expenses. At the beginning of a project, the only line on the graph is PV (planned value). As the project progresses, AC (actual cost) and EV (earned value) are included.
Management by Exception
Management by exception requires that there be a plan; a prediction of how a project should unfold. An exception is any process that is not going according to plan (that is, over budget or behind schedule).
Management by exception uses the guiding principle that managers should concentrate their energies on fixing the most serious problems first. In an EVM project, problems are defined as performance results that are not the same as the planned results.
Management by exception is greatly simplified by EVM because exceptions are so clearly highlighted by routine performance monitoring relative to a highly structured performance baseline. The more severe the discrepancy, the higher is its priority for management attention.
There are four sets of data generated by EVM projects.
BAC (budget at completion): Original completion budget for project
PV (planned value): How much the project was expected to cost at any given time
AC (actual cost): What the accountants say was actually spent
EV (earned value): What the expected cost was for the work that was completed (the planned value for work completed)
Having a visual understanding of the relationship between these data is helpful for interpreting the meaning of EVM statistics. In this figure, the various terms are illustrated with respect to three basic cost lines, planned, actual, and earned.
BAC and PV are forecasted numbers representing planned expectations for the project. BAC does not change. PV is predicted for the entire duration of the project and ends with the BAC. Data for AC and EV are collected through project monitoring.
This is all the information that is required to manage EVM projects. From these data, performance statistics are generated.
There are three variance statistics used in EVM performance measurement, as follows:
Variance at Completion (VAC): the difference between the original planned completion cost (BAC) and the latest prediction of completion cost.
Cost Variance (CV): the difference between the expected cost to complete what has been accomplished and what it actually cost (AC).
Schedule Variance (SV): the difference between, what was planned to be completed as of a given date, and what was actually completed.
There are three index statistics that match up with the three variances, as follows:
To Complete Performance Index (TCPI): an indication of the efficiency rate needed on all remaining work in order to complete the project on budget. A value of more than 1 implies that all remaining tasks will need to come in under budget (on average) in order for the project to meet the BAC.
Cost Performance Index (CPI): shows how well the project is meeting cost targets. A value of 1 indicates that actual costs are the same as planned (AC=EV) for the work that has been completed. A value over 1 indicates that costs are less than planned. A value of less than 1 implies that costs have been higher than planned.
Schedule Performance Index (SPI): indicates whether the project is meeting schedule expectations. A value of 1 indicates that the work completed to date is right on schedule. A value of over 1 indicates the project is ahead of schedule. A value of less than 1 implies the project is behind schedule.
Forecasts use performance indices to predict the future. If the project has been doing really well up to this point, what does that imply about the completion cost and schedule? If tasks are, on average, taking half as long as planned, it suggests that the project overall will be completed in half the time.
While EVM techniques are better able to predict completion cost than schedule, both forecasts remain useful tools.
Forty years of empirical evidence from government acquisition projects has shown that as early as 10 to 15 percent into a project, the CPI provides a reasonable forecast of what a project will achieve. What this means is that cost overruns early in a project will not fix themselves by the end of the project. If anything, the overruns will get worse.
The SPI is a less precise tool for forecasting the future. The principle reason is that tasks affecting the SPI are not necessarily on the critical path for a project, and therefore, they may not affect the completion date. If a large number of tasks not on the critical path are completed ahead of schedule, they will make the SPI look good, even though tasks that are on the critical path may be behind schedule. The SPI provides a high-level overview that, when combined with critical path analysis, allows for an effective prediction of the completion schedule.
The Value of Earned Value
EVM is referred to as “project management with the lights on,” because it shows management where a project is and where it is headed.
Is EVM a magic bullet for perfect projects? No. It will not fix poor planning or poor execution. However, it is a crucial tool for monitoring and reporting performance.
Copyright © Global Knowledge Training LLC. All rights reserved.
This information was drawn from Global Knowledge’s Earned Value Management course developed by Vaddac Consulting in cooperation with Global Knowledge Training LLC. Course director and author Brian Egan is CEO of a manufacturing company (Book Box Company) and a management consultant. He has written three professional development manuals and several white papers on aspects of management science. Since 2000, Brian has been a part-time instructor for Global Knowledge within the Business Training product line.Copyright © Global Knowledge Training LLC. All rights reserved.