Most project managers have heard of Earned Value Management. Fewer can explain why it was invented, what problem it actually solves, or why it produces insights that no other project tracking method can replicate. This post starts from first principles.

The Problem EVM Was Built to Solve

Before earned value, a project manager had two numbers: how much they planned to spend by a given date, and how much they actually spent. If planned spending was $10 million and actual spending was $9 million, the instinctive read was favorable — the project was under budget. But that comparison tells you nothing about how much work was actually accomplished. A project team that has completed 60% of the work while spending 90% of the budget is in serious trouble, even if today's actual cost happens to be below the cumulative spending curve.

This was the fundamental problem the U.S. Department of Defense identified in the 1960s when cost overruns on major defense acquisition programs were endemic. Projects looked financially healthy until they didn't — and by the time the gap between reported progress and actual accomplishment became visible, it was too late to recover. The Cost/Schedule Control Systems Criteria (C/SCSC), which eventually became EVMS, was the structural response.

The Three Core Values

Earned Value Management is built on three data points measured at the same point in time:

Planned Value (PV) — the budgeted cost of work scheduled. What did you plan to have accomplished by today, expressed in dollars?

Earned Value (EV) — the budgeted cost of work performed. What did you actually accomplish, expressed in the dollars you originally budgeted for that work?

Actual Cost (AC) — the actual cost of work performed. What did it actually cost to accomplish what you accomplished?

The critical insight is that EV is denominated in budget dollars, not real dollars. When you complete a task that was budgeted at $50,000, you earn $50,000 of value — regardless of what it actually cost. This creates a common unit of measurement that allows schedule performance and cost performance to be evaluated independently and simultaneously.

What the Derived Metrics Tell You

From PV, EV, and AC, all of the important EVMS performance indicators are derived:

Cost Variance (CV = EV − AC) — negative CV means you spent more than budgeted for the work you accomplished. A true cost overrun, not distorted by schedule status.

Schedule Variance (SV = EV − PV) — negative SV means you accomplished less than planned for this period. A true schedule shortfall, expressed in dollars of work rather than days of delay.

Cost Performance Index (CPI = EV/AC) — the efficiency rate at which you are converting spending into accomplishment. A CPI of 0.85 means you are getting $0.85 of earned value for every dollar spent. On large programs, the CPI tends to stabilize after the first 20% of work is complete — making it one of the most reliable early predictors of final cost.

Schedule Performance Index (SPI = EV/PV) — the efficiency rate at which you are accomplishing planned work. An SPI below 1.0 means you are behind schedule.

Estimate at Completion (EAC) — the projected final cost of the program, derived from current performance trends. The simplest formulation is BAC/CPI, where BAC is the total budget at completion.

Why This Matters Beyond Compliance

EVMS is frequently perceived as a federal compliance requirement — something programs do because a contract requires it. This perception misses the point. The reason the federal government mandates EVMS on major acquisitions is the same reason any large capital program benefits from it: it is the only method that quantifies schedule performance in financial terms and separates schedule-driven cost distortions from true cost overruns.

A project that is behind schedule will show inflated actual costs relative to planned value even if it is running efficiently. Without EVM, those inflated costs look like a cost overrun when they are actually a schedule variance. With EVM, the distinction is explicit and actionable.

The transit authority managing a multi-prime infrastructure expansion, the general contractor running a hospital construction program, and the international EPC team on a liquefied natural gas terminal all face the same fundamental problem: how do you know, with confidence, where the program actually stands? Earned value is the answer that has existed for over sixty years — and it remains the most rigorous method available.

The Prerequisite Nobody Mentions

For earned value to be meaningful, the baseline against which performance is measured must be trustworthy. A budget that was set too low, a schedule that was built to look compliant rather than to reflect real sequencing, or an earned value claim that does not reflect actual physical progress — any of these corrupts every derived metric downstream. The integrity of the baseline and the accuracy of progress measurement are prerequisites for everything else EVM produces.

This is the problem that EVMS compliance tools, including the FIE, are ultimately designed to address: not just whether the calculations are correct, but whether the underlying data is.

Peveka Solutions builds the Forensic Intelligence Engine — an AI platform that reads your contract, schedule, and cost data simultaneously and flags where your program is out of compliance before it becomes a finding.