This paper demonstrates that project management is as a positive organisational “performance discriminator” and that the project management tool, Earned Value Management (EVM) is, in turn, a positive “project performance discriminator”.
Properly utilised, project management and EVM provide important capabilities which will optimise the Return On Investment (ROI) achieved for investments which are dependent on projects and project performance for delivery.
“Strategic commercial imperatives” are defined, followed by a brief description of the features of EVM focusing on the basic EVM cost and schedule metrics and demonstrating how these metrics provide “early warning” of project performance issues, thus enabling proactive project based corrective actions to be undertaken.
EVMs unique ability to statistically predict final Costs at Compete (CAC) based on project performance to date will be discussed demonstrating how this important EVM “early warning signal” can allow project cost risk to be proactively managed while most of the cost budget remains intact.
Using EVM to proactively monitor and manage project cost risk is an important feature of the methodology for protecting the initial project based investment and positively influencing the ROI ultimately achieved from that investment.
What are “Strategic Commercial Imperatives”?
The overarching strategic imperative for all organisations is organisational survival.
For commercial organisations survival is dependent on achieving revenue and profitability objectives which in turn leads to positive Return on Investment (ROI) which creates additional shareholder value and satisfaction. To achieve profitable outcomes, commercial organisations must win and successfully execute profitable business ventures.
For Government organisations the imperatives are achieving the “capabilities” required to meet Government policy objectives, ideally within the allocated cost budgets and required timeframes.
“Strategic Commercial Imperatives” and Projects
As is evidenced by the continued rapid growth of project management and the membership of project management professional bodies, including the Project Management Institute (PMI®), for many organisations, projects have become the key mechanism for either:
- Winning and executing profitable business; or
- Enabling “capability” to be put in place
which leads to the achievement of corporate objectives; or delivers the capability goals required by Government policy?
Investment definition, assessment, and selection are complex topics which are outside the scope of this paper to address in any detail. It is sufficient to observe that for investments which organisations use projects to deliver, irrespective of whether the benefits to be derived from a project budget allocation have been:
- Formally calculated and assessed using techniques such as Net Present Value (NPV) or Internal Rate of Return (IRR); or
- Informally assessed and selected along the lines of either:
- “It seemed like a good idea at the time”; or
- The initiative was judged to be essential for corporate survival.
The key determinants which will directly influence the ROI ultimately achieved are whether the investment projects deliver:
- Close to the planned completion date
so the benefits and ROI can begin to be derived as soon as possible; and
- Close to the planned budget
to improve the prospect of the investment achieving the projected ROI and benefits identified in the business case.
The third factor essential for successful delivery is whether the project delivers to required technical and functional performance characteristics. Technical Performance Measurement (TPM), while equally important, is outside the scope of this paper. EVM and TPM combined form the field of project management known as Integrated Performance Management (IPM).
Project and Program Management as a Performance Discriminator
Research reported by Velocci undertaken by Pittiglio Rabin Todd & McGrath (PRTM), for Aviation Week & Space Technology, of the United States defense and aerospace corporate sectors cited by Christle (Christle, 2002) concluded, as the title of the article indicates “Effective Program (project) Management Identified as a (performance) Discriminator”:
“Those who win a high percentage of award fees by completing projects on time and on budget take a much more comprehensive business perspective of program management.”
“…effective program management will remain a major differentiator between companies likely to bring a project in on time and on cost most of the time--earning the rewards that go with that level of success -- from less competitive companies who consistently struggle with program execution” (emphasis added) (Velocci, 2001).
Another striking example of the recognition of the benefits of projects and project management to organisations is provided from Japan. This includes the realisation that a partial explanation for the current state of the Japanese economy has been the inability of Japan to derive the cost and delivery benefits of projects and project management in spite of their legendary “quality” related achievements.
Nishi has observed:
“…one criterion commonly used to purchase commodities relied on the hypothesis that corporate brand and history provide confidence in the product quality. However the results did not always meet expectations on both sides when applied to projects.
…We expected that project management (PM) could be the essential tool for grappling with the issues posed by the need to integrate these intricate management systems”. (Nishi, 2002, p1).
The Japanese are in the process of actively addressing these issues, which includes the study and adoption of EVM:
“…the number of PMPs in Japan remarkably increased from 130 to 1,600 over the last four years and influenced other similar types of qualifications and systems relevant to Project Management Professional (PMP®)/PMBOK® Guide. This reflects an important ethos in our Management Renaissance and has simultaneously shown PM to be a key driver not only in the governmental sector but also in the commercial sector” (emphasis added) (Nishi, 2002, p1 and p9).
What is Earned Value Management?
Defining Earned Value Management
The Project Management Institute's definition of EVM is:
“a method for integrating scope, schedule, and resources, and for measuring project performance. It compares the amount of work that was planned with what was actually earned with what was actually spent to determine if cost and schedule performance are as planned.” (PMI, 2000)
The Earned Value Management System (EVMS) methodology is well and comprehensively documented in many public domain sources including Anbari (Anbari, 2003), Christensen (Christensen, 1999), Wilkens (Wilkens, 1999) and text books including Fleming and Koppelman (Fleming & Koppelman, 2000) to which the reader is directed for a complete coverage of the topic.
The term “Earned Value” may, depending upon the context in which it is used, have one of two meanings.
Earned Value, the Management System
Earned Value, the management system has its first formal origin in the “Cost Schedule/Control System Criteria” (CS/CSC or “CS squared”) first published by the United States Air Force in 1967.
Earned Value Management Systems have an extensive history of use by the United States Defense Department (US DoD) for “cost reimbursable” projects (i.e. projects in which the cost risk is borne by the Government) since that time with usage being extended to other US Government agencies (e.g. NASA) and adopted for defence procurements by the governments and defence departments of other countries including the United Kingdom, Canada, Australia, and Sweden.
For a succinct (2 page) and very relevant view of the critical components of EVM (in any environment – government or commercial) from a Swedish perspective see “Earned Value Management in 10 Basic Steps” (Antvik, et al 2001).
Historically EVMs usage within commercial sector organisations has been relatively low, especially where usage has not been mandated by Government regulation and/or contractual obligations.
Many proponents for extension of the usage of EVMS into the commercial sector continue their advocacy. Two illustrative examples amongst many are provided.
Henderson (2002) advocates the use of a “simplified” EVM approach for the direct labour components of commercial sector Information Technology (IT) projects. The Earned Value measure is derived using a percentage of project completion method where the percentage of project completion is obtained from the project schedule.
Fleming and Koppelman, (2001) and who have recently suggested that:
“The Sarbanes-Oxley Act was signed into law July 30, 2002. It would seem to reinforce the contention that there is a fiduciary duty placed on corporate executives to tell the whole truth when reporting the financial condition of their companies. This duty would include an accurate assessment of the true current status and the final required costs to finish all multi-year projects.
Employing a simple form of earned value management can help corporate executives meet this obligation (emphasis added).” (Fleming & Koppelman, 2004b, p28).
Earned Value the Measurement
Earned Value is also a measurement; previously called the Budgeted Cost of Work Performed (BCWP). As the latter term implies, it is the budgeted value of work actually performed.
Wilkens provides a simple, concise and readily understandable explanation for Earned Value, the measurement:
“The simplest way to think of Earned Value is to equate it with physical progress
…Earned Value is also a measure of progress
…there is a direct relationship between Earned Value and percent complete” (Wilkens, 1999).
A simple illustrative example project, utilising a percentage completion method demonstrates the Earned Value measurement concept:
|Budget at Complete (BAC) =||$1,000|
|Percent Complete cumulative assessed as being||50%|
|Earned Value cum = $1,000 x 50% =||$500|
$500 is the “Earned Value” cumulative, in this example expressed in dollars, which represents the physical progress actually achieved to date by this project.
The example highlights that the Earned Value measure is expressed in a common unit of measure. While this may be any meaningful unit of value, is most often expressed as either dollars (or other currency) or hours.
A critical, feature of EVM is variance analysis, which requires a common measurement base to implement.
Exhibit 1 shows the key measures and metrics from an EVMS using the cumulative values of the:
- Planned Values (PV), previously known as the Budgeted Cost of Work Scheduled (BCWS)
The Planned Values cumulative is also known as the Performance Measurement Baseline (PMB);
- Actual Costs (AC), previously known as the Actual Costs of Work Performed (ACWP); and
- Earned Value (EV), previously known as the Budgeted Cost of Work Performed (BCWP)
to produce the familiar EVM “S Curve” Chart and Report.
Basic Earned Value Variances and Performance Indexes
From the Earned Value measure, two basic variances and performance indexes are readily available.
Cost Variance and Cost Performance Index
The EVM Cost Variance (CV) is calculated as:
Cost Variance = Earned Value – Actual Costs
Adding an Actual Cost of $600 to the illustrative project example above, the Cost Variance is:
$500 - $600 = ($100).
This variance is unfavourable as we have only achieved $500 worth of Earned Value (physical progress) for an Actual Cost incurred of $600.
The EVM variances have been consistently constructed so that negative variances are “unfavourable”, positive EVM variances are “favourable” and a variance of zero indicates an “on track” status.
The EVM Cost Performance Index (CPI) is calculated as:
Cost Performance Index = Earned Value / Actual Cost
Continuing with the example project data, the Cost Performance Index is:
$500 / $600 = .83
This is an unfavourable performance index.
The EVM performance indexes have been consistently constructed so that a performance index value less than 1 are “unfavourable”, EVM performance index values greater than 1 are “favourable” and a performance index value of 1 indicates an “on track” status.
Schedule Variance and Schedule Performance Index
The EVM Schedule Variance (SV) is calculated as:
Schedule Variance = Earned Value – Planned Values
Adding a Planned Value of $700 to the illustrative project example above, the Schedule Variance is:
$500 - $700 = ($200)
This is also an unfavourable variance as we have only achieved $500 worth of Earned Value whereas the project had planned to achieve a Planned (and Earned) Value of $700.
The EVM Schedule Performance Index (SPI) is calculated as:
Schedule Performance Index = Earned Value / Planned Values
Continuing with the example project data, the Schedule Performance Index is:
$500 / $700 = .71
This is also an unfavourable performance index.
The more familiar, in a non EVM environment, Spend Variance is calculated as:
Spend Variance = Planned Values - Actual Costs
$700 - $600 = $100.
In a non EVM environment, the positive spend variance may be seen as favourable since the project has expended less costs than planned.
However, the example provided highlights the power of the EVM variances and performance indexes in providing early warning of project performance issues.
Although we have positive Spend Variance, there is both negative cost and schedule performance on this project when compared to the Earned Value or “physical progress” actually achieved. These are clear indicators of adverse project performance issues in need of analysis and corrective action in order to protect the delivery outcomes and projected ROI from the project based investment.
To inject “real world” realism into the discussion, Exhibit 2 provides the reconstructed cost and schedule performance index history and commentary for a small, but seriously troubled commercial IT sector project, called “Example Project #1” to maintain required anonymity. The EVM data from this project data highlights the progressively more severe consequences which can occur from failure to realise that a project is experiencing adverse performance issues and taking early corrective action.
In this case, the non-availability of EVMs “early warning signal” metrics from the organisations project status reporting system and a lack of detailed project performance monitoring by management, possibly due to the small initial budget for this project, were the key contributing factors leading to this seriously adverse situation.
Relating the EVM Schedule Variance to the “Real Schedule”
Review of Exhibit 2 reveals that the EVM Schedule Variance is calculated by reference to the “dollars” (or “value”) Y axis rather than the X “time” axis. A common question arising is how the EVM Schedule Variance relates to the “real” network project schedule.
The first point to be made is that the EVM Schedule Variance is derived from cost accounting data and not the “real schedule”. Consequently the EVM Schedule Variance does not (and cannot) identify the specific work performed, most importantly, critical path versus non critical path work.
In a “quirk of algebra”, because the Earned Value measure is calculated by reference to the Planned Values, the Schedule Variance will always converge and resolve to zero at project completion even if the project has been delayed in duration terms.
As the EVM Schedule Variance does not measure time, authors including Abba (Abba, 2000) and Antvik (Antvik, 2001) have suggested that with the benefit of hindsight, a better name would have been “Accomplishment Variance”. What is really being measured is the volume of work planned versus the volume of work actually accomplished.
The principal strength of the Schedule Variance, confirmed by observation on hundreds of US DoD programs and consistently repeated in my own personal experience of commercial IT sector project recoveries is that the EVM Schedule Variance provides reliable early warning of a project in trouble when it is large, early and unfavourable.
The standard advice is that the EVM schedule metrics should be reviewed in conjunction with the “real project schedule” to ascertain and confirm the real schedule status.
A recent development in EVM has been the definition of a parallel measure to Earned Value for schedule called Earned Schedule by Lipke (2003). From the Earned Schedule measure, analogous time based indicators for schedule performance which parallel the behaviour of the EVM indicators for cost have been developed. The Earned Schedule, Schedule Variance and Schedule Performance Index indicators are identified with a “time” or “(t)” suffix. The Earned Schedule theory has been developed to conceptual parity with EVM, including predictive formulae for project duration which parallel the Independent Estimate Completion (IEAC) formulae for cost which are discussed below. See Henderson (2004, 2003) for details.
Retrospective application of Earned Schedule to a small portfolio of 6 IT projects and subprojects by Henderson, (2003) concluded:
“Lipke's assertion that “The application of Earned Schedule provides a set of schedule indicators, which behave correctly over the entire period of project performance” has, using this sample of projects, been confirmed.”
Research aimed at trying to confirm the empiric validity and validity of the proposed predictive capabilities of Earned Schedule using a US DoD project dataset with large scale programs is, at the time of writing this paper, in progress.
Predictive Uses of Earned Value
EVM also has the currently unique project management feature of providing statistically based predictors of future project cost performance known as Independent Estimates at Completion (IEACs).
IEACs provide another important “early warning signal” of current and/or future project performance issues. This provides the opportunity for management to take proactive corrective action if the early warning signals are heeded and there is appropriate management determination to influence the projected final outcomes.
Exhibit 3 describes the components of the generic IEAC formula:
- Actual Costs cumulative incurred to date (AC)
In investment terms, Actual Cost may also be thought of as irrecoverable “sunk costs”. This also highlights the importance of project “early warning signals” and the benefits of proactive corrective action to enable the amount of nugatory “sunk costs” expended to be minimised.
- The remaining theoretical cost amount the project needs to complete, which is known as the Budgeted Cost of Work Remaining (BCWR).
BCWR is a simple calculation in which the Earned Value (EV) cumulative is subtracted from the Budget At Completion (BAC).
- A Performance Factor.
The performance factors commonly utilised are the:
- Cost Performance Index (CPI);
- Schedule Performance Index (SPI);
- A combination of CPI and SPI.
Where the CPI is utilised as the performance factor, the generic EVM IEAC formula resolves to:
IEAC CPI = BAC / CPI
The US DoD experience and lessons learned were published in the “Earned Value Body of Knowledge” by Fleming and Koppelman (1999). Studies of US DoD project and program performance indicate that:
- IEAC CPI provides the “low-end” or “floor” prediction of the likely project completion cost; and
- IEAC CPI X SPI provides the ‘high-end” or debatably, “most likely” prediction of the likely project completion cost.
IEAC CPI X SPI provides the most pessimistic predictions, especially in circumstances of adverse cost and schedule performance. The product of two numbers less than 1 is less than either number which crudely models the “double” adverse cost impact of a project which is both over budget and behind schedule.
For a comprehensive discussion on the use the IEAC formulae and the various performance factors, the reader is referred to an excellent paper by Christensen (Christensen, 1999) which uses the data available from the cancellation of the US DoD A-12 Avenger stealth aircraft program in 1991 by the then US Secretary for Defense, Mr Dick Cheney as the case study example.
Exhibit 4 provides the remainder of the reconstructed EVM history for “Example Project #1” including the graphing of an IEAC calculated using what is known at the “80/20” rule.
This IEAC performance factor often provides a “mid-range” prediction and is calculated as:
(CPI x .8) + (SPI x .2)
This example also highlights the severe adverse commercial consequences which can occur, even for small IT projects where project performance issues are not proactively identified and effective corrective actions implemented. The consequences in lost ROI and lost opportunity were very significant.
The Key Benefit of EVM
Personal experience and the academic literature highlights that merely comparing planned versus actual costs, especially for IT projects, is not an adequate method for accurately establishing, assessing, monitoring and predicting project performance and outcomes.
This “financial accounting” approach fails to account for the actual “physical progress” (Earned Value) realised to date by the project. In effect, the “physical progress” achieved is implicitly equated to the costs expended. This can be an especially dangerous approach, particularly for IT projects which:
- Are usually very “people intensive” resulting in direct labour costs often being the major component of project “cost risk”; and
- Often have deliverables which lack direct physical visibility (e.g. bespoke software development source code).
In my own experience of IT project recovery situations personally managed there have been remarkably close correlations between historic planned versus actual costs while the projects had been seriously troubled, often for extended periods of time. What later seemed blindingly obvious in retrospect had not been detected in advance over that time as Exhibits 2 and 4 graphically illustrates as one example.
In the common situation of concurrent IT project (process and product) quality failure very large costs can be incurred with little useful or re-useable outputs in spite of what can be the very significant costs and effort expended.
In a further irony, the basic data required for a simplified EVMS (where the Earned Value measure could be derived from “percentage completion” calculations) contained in Exhibits 2 and 4 were required elements of the organisation's project status reporting system. Unfortunately the organisation did not require calculation of the EVM metrics from the data submitted.
EVM as a Project Performance Discriminator
EVM is the only project management methodology currently in existence which integrates project:
- Cost performance;
- Schedule performance; and
- Technical performance or physical progress
into a single measurement and management system.
The EVMS provides integrated objective measures of project performance including the previously described:
- Cost and Schedule Variances and Performance Indexes; and
- IEAC statistical predictors of future outcomes based on historic performance to date.
These metrics assist in confirming “real” project performance to date and provides proactive guidance into priority areas requiring corrective action.
Proactively heeding and acting with the necessary determination upon the messages provided by the EVMS gives management the opportunity to take corrective actions while most of the cost budget is still intact.
A properly implemented EVMS provides the opportunity to avoid or minimise the adverse impact of (undetected) project performance issues on project investments, positively influencing the ROI achieved from those investments.
US DoD Program Performance History with EVM
The US DoD has a long, sometimes controversial and heavily researched history of using EVM for more than 35 years.
Christensen, (1996) citing previous studies noted:
“Only rarely has a large cost overrun on a defense contract resulted in the project's cancellation…. Although the average cost overrun on defense contracts has averaged about 20 percent since the mid-1960s, cost overruns on some nondefense projects of comparable size and complexity have been larger (Drezner, Jarvaise, Hess, Hugh, & Norton, 1993, p. xiv)”.
The much researched and cited cancellation of the A-12 program in 1991 and subsequent inquiry by Beach (1991) provoked of a long period of acquisition reform within the US DoD which included a renewed interest and focus on the use of EVM, particularly the IEAC statistical predictors of future cost performance.
The extraordinary success of the US DoD acquisition reform program in the 1990s is provided by Abba (2003):
“The Department of Defense's success in reforming EVM is apparent in cost growth history tracked by the Office of the Secretary of Defense. In November 1999, the total overrun on more than 100 of the Department's largest, most risky contracts (flexibly priced contracts requiring compliance with EVMS…) was only 5.5%, or $1.2 billion (Christle, 1999). The contracts in aggregate were 66% complete and represented $72.8 billion in target value.”
US Government Initiatives: Program Management, EVM and Investment Protection
Recent developments in the United States Federal Government by the Office of Management and Budget (OMB) via OMB Circular A-11 Part 7 (OMB, 2003) has been the significant expansion of the use of an EVMS for capital acquisition procurements by all US Government agencies, albeit by a continuation of regulatory mandate.
Circular A-11 Part 7 has significantly extended the usage of EVM across the US Federal Government and into much smaller projects and programs compared to the historic usage on very large, very complex DoD acquisition programs.
Section 300.5, Circular A-11 part 7 provides that for “Developmental programs”:
“Agencies must use a performance-based acquisition management system, based on ANSI/EIA Standard 748, (an EVM standard) to obtain timely information regarding the progress of capital investments.…Agencies are expected to achieve, on average, 90% of the cost, schedule and performance goals for major acquisitions.”
Section 300.10, of the Circular makes it clear that the policy intent for introducing the mandatory EVM requirements is to provide a sound basis for investment:
- Selection as part of evaluating business case and capital asset plans;
- Protection via program and budget reviews; and
- When necessary, investment continuation decisions.
Section 300.10, also makes it very clear that the fundamental importance of effective program (project) management in the realisation of ROI for capital investments is also well understood by OMB:
“Assessment of business cases
A business case may score very high on the criteria…but if the program its supports is deemed ineffective there may be no business case that can be made for the investment (emphasis added)”
OMB are also well aware of the “project advocacy” phenomena researched by Christensen (1996):
“EAC‘s are subjective in nature and the contractor and government EAC‘s are often quite optimistic in an attempt to favor investment continuation” (emphasis added). (OMB 2003, Section 300–12).
OMB have also mandated the reporting, in what is known as the “Exhibit 300” report , of the IEAC “prediction formula” calculations using the IEAC CPI “floor” and IEAC CPI X SPI “ceiling” or “most likely” cost predictors so that:
…using the prediction formulas will give the IPT (Integrated Project Team) some proven parameters to structure the discussion. The objective is to provide a realistic EAC for management decisions to continue, restructure or terminate the investment (emphasis added) (OMB 2003, Section 300–12).
The capital procurement issues which OMB are in the process of addressing from a US Federal Government perspective, in part by mandating the extended use of EVM, are issues which are equally relevant to commercial organisations seeking to maximise the ROI from their project based capital investments.
In a world in which projects have become the key mechanism for delivering many investments, capabilities which enable positive project performance, allow those organisations to maximise the ROI achieved from their project based investments.
Organisations, both government and commercial sector, which are able to consistently deliver successful project based investments, thereby optimising ROI, should be able to continue building on the success which comes from successful investments.
EVM is a relatively simple project management tool which integrates project cost, schedule and technical performance into a single project measurement and management system from which objective measures of project performance to date and predictions of future performance are readily derived.
Properly implemented, with appropriate management support and a willingness to heed the messages from the EVMS and take corrective action when required with the necessary determination, an EVMS can be a powerful project performance discriminator.
In my own project management practice, I use a simplified version of EVM whether I am obliged to or not because of the positive benefits it provides me in being able to objectively status, track and monitor overall project performance.
EVMs “early warning signals” of project performance issues allow me to position for and take corrective action. In non EM environments, this usually occurs before others even realise there is an issue. EVM is a positive project performance discriminator.