Although the concept of “earned value” comes from the field of project management and pre-dates Six Sigma, it is one of many tools that are consistent with and complementary to Six Sigma. In particular, earned value is an effective tool for “fact-based conversations” and is particularly useful as a control mechanism for any sort of larger project.
Many organizations today are undertaking large software deployment projects (building or purchasing) as a result of major initiatives being undertaken by the business. And many of these efforts are running aground on the project management rocks.
When trying to manage dozens of major efforts concurrently, it is not surprising that there are many delays and complications. Projects run late, cost more than planned and sometimes fail to meet operational needs. Many business executives (especially Le Grand Fromage) are frustrated with this state of affairs. They are frustrated by delays per se, but even more by the inability of someone to give them a consistent, understandable picture of the true status of a project.
If you are faced with that problem, good old-fashioned critical path planning combined with earned value analysis may be the solution you need.
First, Sound Critical Path Planning
In order to use earned value analysis you first need a sound critical path method (CPM) plan. Such a plan is produced by beginning with a work breakdown structure (WBS), which is simply a list of all the tasks which must be performed in order to reach the “go live” point. Of course, it is not a trivial task to produce such a list for any relatively large project, but if you cannot make the list, you certainly cannot forecast a reliable completion date either. If you want to be able to give M. Fromage the straight scoop, you will have to know it yourself. (By the way, author and project management expert Rob Thomsett defines an out-of-control project as one which does not have time to plan.)
Assuming you have the fortitude to get past this first step, you really have a chance to avoid getting the big guy cheesed off. (If not, stop reading now and update your resume.) Given the WBS, the next step is to define who will work on each task, how much effort (hours, days, or whatever) each resource will devote, and what the overall duration (calendar time) will be. To make reasonably accurate estimates, you will need to break tasks down into small steps, usually one to two weeks in duration.
Next, you must define the predecessor/successor relationships among the tasks. This will define what can be done in parallel and what must be done in a pre-determined sequence. You will want to enter all of this information into one of the project management tools such as Microsoft Project, Project Workbench or Timeline. Given all of the aforementioned data, these systems will be able to calculate early and late start and finish dates for all tasks, the critical path and the amount of slack available on each task.
Sounds like a lot of work, and we have not even talked about earned value yet. Actually, it is a lot of work. It definitely is not something you want to do unless you really need to know the status of your project as you go along. In a recent project, it took a team of 8 to 10 full-time people working three weeks to get through this process. The team ended up with about 1,500 tasks in the WBS.
The project has a total of about 25 people full-time and a few others who help out now and then. It has been going on for more than four years and nothing has yet been delivered. The project has been recycled three times due to various technical and managerial problems. So, it is not surprising that the vice president in charge really wants to know the status of this one.
Knowing What Is Likely to Be Required
Thanks to the CPM planning process, we now really know what is likely to be required to finally pull this one off. Because we have a detailed plan we can now see, each and every week, what has been completed and what has not. We also will know the impact any delays we encounter will have on the final delivery date. What we may not know, in spite of all of this effort, is the likelihood that our estimates for the tasks not yet started are actually accurate. That is where earned value comes in.
Earned value is simply the ratio of the time (or cost) it took to complete all of the tasks up to a particular point in time, compared to the time (or cost) which was estimated those tasks would take. If the actual time (or cost) is consistently greater than the planned numbers it means that we are consistently under-estimating the effort required. If this is the case, we may wish to revise the remaining estimates upward by a percentage that reflects actual experience to date, rather than be faced with the need to continually take small slips each week. The diagram below (Figure 1) provides an example of one way to depict this information.
Here are what the abbreviations in Figure 1 stand for:
- BCWS – Budgeted Cost of Work Scheduled (planned value)
- BCWP – Budgeted Cost of Work Performed (earned value)
- ACWP – Actual Cost of Work Performed (actual)
- BAC – Budget at Completion
- EAC – Estimate at Completion
- SV – Schedule Variance = BCWP – BCWS
- CV – Cost Variance = BCWP – ACWP (positive is favorable)
- SV% = SV / BCWS
- CV% = CV / BCWP
Earned value shows how the actual rate of development (productivity or capability) compares to the rate that was assumed when estimates were prepared. Many organizations, including the one represented on the chart above, simply do not know historical rates of delivery and must therefore depend on an approach like the one described. Mr. Big may not be delighted by the grim realities, but at least he may finally get a date that is actually achieved.