How would you react to the presentation of a method for measuring project performance, forecasting costs and deadlines, identifying problems and making the right decisions? The response would undoubtedly be excellent and would consign many of your current practices to oblivion!
Born at the end of the 60s in the U.S. Department of Defence, EVM (Earned Value Management) has been delivering on these promises for several decades now, but its use remains discreet in France.
How can we put this method into practice? What pitfalls should we be wary of? How can we avoid them? And what are the benefits?
All these questions deserve answers. We invite you to study the subject, using as examples first a nuclear power plant construction project, then a portfolio of industrial projects. We hope that a summary of the lessons learned will enable you to move from theory to practice!
Survival Glossary (*)
(*) For a more detailed description of the method, please refer to the end of this article.
The case studied is that of a project to produce execution studies (plans and models), for the construction of a nuclear power plant, by a group of companies.
The buildings to be constructed are made up of storeys, and the work sequences are well established. The project is therefore repetitive and industrial. This organization is reflected in the specifications. The budget for each task (BAC task) is known from the outset. This organization is also reflected in the project structures, with coherent WBS (Work Breakdown Structure) and CBS (Cost Breakdown Structure).
The contract requires the implementation of a process for managing developments from start of the project. It also stipulates that the financial management of changes must comply with the budgetary rules in force on the project. These provisions facilitate PV (Planned Value) maintenance.
The project team established common rules for measuring physical progress right from the start of the project. Particular care was taken to ensure that these rules were deployed and applied uniformly and continuously by all those involved throughout the life of the project.
From a planning point of view, the following data has been maintained in Primavera:
Costs were re-injected into Excel, enabling the following data to be maintained:
This information was concatenated monthly in Excel to establish the EVM indicators. These indicators were injected into Power BI for better visualization.
Thanks to a rigorous EVM approach, the project team and the customer had a shared view of progress throughout the life of the project. The precision of the indicators positioned discussions on the real situation of the project, its successes and risks, and made it easier to take steering actions, at the right moment, for maximum efficiency. The ability to zoom in on progress at project level or activity level enabled us to fine-tune diagnoses and action plans. Ultimately, schedule (SV’) and cost (CV) variances remained under control.
The case studied is that of a portfolio of engineering projects in the industrial sector, containing several hundred projects.
The main pitfall of this method stems directly from its founding principle, i.e. to implement an annualized EVM focused on the current year’s performance and not on the project’s terminal performance. It is thus possible for a project initially planned for five years to start its seventh year of existence with the new (and seventh!) annual PV as its performance measurement point. It’s as if at the start of each new year, the project’s performance measurement can be restarted from scratch. This mode of operation, based on a succession of annual performances, distances teams from the notion of overall project performance. While there are safeguards in the form of annual drift ceilings for the SPI and CPI indicators, what about the project that manages to stay just below the authorized threshold year after year? It’s doubtful that its overall performance will be there.
The first benefit of annualized EVM is that it provides project managers with a simple, turnkey method of measuring project progress, and objectifying this measurement in relation to expenditure.
A second benefit is the reliability of budget forecasting for the portfolio’s financiers, despite the very large number of projects. The availability of monthly performance reports not only ensures accurate budget marking, but also enables a pooled support team to be activated to help the most at-risk projects implement appropriate security plans.
We might add that an annualized EVM is less demanding in terms of planning than a classic EVM. This may be worth considering if the project portfolio is regularly confronted with a lack of planning maturity, whatever the cause: complex or innovative nature of the projects, planning skills of the teams, etc.
What can we learn from this long presentation?
The essential prerequisite for EVM is the existence of coherent, compatible project structures, in particular the WBS and CBS.
Earned value management is based on the following indicators, whose names, calculation methods and, above all, concrete meaning must be remembered!
Note 1: The BAC is not set in stone. It is regularly updated by integrating the changes decided on for the project. This is obviously crucial for industrial projects that are subject to numerous changes.
Note 2: 0-100, 0-50-100, equivalent units, intermediate milestones, expert judgement, etc. There are a wide variety of rules for calculating physical progress. It’s up to the project team to set the rules as early as possible… and stick to them.
Note 3: Expenditure is not earned value: EV is not AC. If this were the case, there would be no need for project managers!
Note 4: In the figure above, EAC is greater than BAC. When a project cannot finance its cost overruns, it dips into its management reserves (PMR).
Note 5: There is a fundamental difference between the three EAC calculation formulas given above. The first is based on the notion of remaining work, which implies a complete review of the remaining expenditure, which is very tedious but generally more coherent if carried out rigorously. The next two, on the other hand, are based on an extrapolation of the project’s current situation, which generally leads to a less precise result, but one that is immediate and easy to deploy.