A project plan is simply an estimate of the effort required to meet the project objectives, layered over a best guess timescale. For complex projects, work breakdown structures can be used to create smaller work packages that will simplify the estimating process. However, progress is unlikely to go according to plan for all but the simplest projects.
The prudent project manager will want to measure progress early on in the project’s life. Are we where we should be after the first month? (schedule variance) and by how much might we have overspent? (cost variance). Early warning of poor project performance gives time for corrective action to be taken.
But how best to measure the project’s progress? Too often a simple comparison of hours spent to date with those planned is used. Yet without an objective measure of what’s actually been achieved, assessment of the cost and schedule performance is likely to be deeply flawed. If there is little confidence in the measure of progress to date, there will be even less when revising estimates for the remaining work.
What’s missing is an objective measure of what’s been achieved for the effort expended, which may be very different from what’s planned:
Earned value measurement
This chart shows the essential ingredients for measuring progress:
- The agreed project plan. This is often referred to as the project baseline, because it provides the yardstick for measuring progress. It, too, must be updated as changes in project scope are agreed, or realization dawns that the original estimate was too optimistic.
- The amount of effort expended, which is usually collected through timesheets
- The valued of the work done, or the Earned Value. This needs to be objectively and consistently measured.
It is relatively easy to calculate and agree on the project baseline values (from the signed off project plan), and the actual effort to date (from timesheet bookings). Estimating the value of what’s been achieved can be much more contentious. Each task will have a work value from when the estimate was agreed. The risk is that the project manager will over state what’s been achieved – ‘we’ve started the task and were now 90% complete’ is not uncommon, but a number of best practice guidelines are here to help.
- Keep each task duration as short as possible. If you have a 2 year project with many tasks, each of which is no more than 2 months long, then only permit the task value to be claimed when it is physically completed. This avoids the need for any subjective percent complete assessment at the reporting date.
- For longer tasks, insert a number of value milestones that can only be claimed with they have been achieved. E.g. a system implementation task may consist of acceptance testing followed by roll out training over a period of months. You can make successful acceptance testing an interim milestone valued at, say, 25% of the task value, which can be claimed part way through the task.
- Where a task consists of a number of repetitive sequences, e.g. producing 15 widgets, then the number of widgets produced by the reporting date can be used to calculate the task’s earned value.
For the project to be on track, each curve should follow the baseline. Schedule concerns will arise when the earned value line is lower than the baseline, and cost concerns when the amount of effort expended is higher than the earned value. These are what the second chart shows, so this project looks to be in serious trouble – but by how much?
Quantifying the project performance to date
With earned value measurement added, project progress can be quantified by calculating Schedule and Cost variances.
- Cost variance = Earned value – Actual work to date
- Schedule variance = Earned value – Baseline value to date
In both cases, negative values should cause concern. Whilst these will help to answer the question of how well is the project doing to date, it is useful to convert these to performance indices, when re-estimating the remaining work.
Estimating to completion
The variances can easily be changed into performance indices.
- Schedule performance index (SPI) = Earned value / Project baseline
- Cost performance index (CPI) = Earned value / Actual effort
Looking at the second chart example again, the results are:
- SPI = 22/34.5 = 64%
- CPI = 22/30 = 73%
The target is 100%, so both are way below where they should be. Unless the project manager can convince his bosses that the project is over the worst, there is a high probability that the eventual cost will be at least 25% over budget, with a delay of 30-40% to the completion date highly likely.
Transform your project progress reviews
This example shows why early warning of poor project performance is so important. Only if detected in the early stages will there be time for corrective action to pull the project back on track. Earned value makes project performance measurement much more objective and gives project sponsors the ammunition to challenge over optimistic forecasts to completion. It is easy to adopt and should be used much more widely than it is.
In order to understand how the benefits of introducing earned value are multiplied in a multi-project environment you can visit this article here.