Earned Value Management (EVM) answers the question every stakeholder actually cares about: are we on track for time and money, and if not, by how much? It combines scope, schedule, and cost into a single measurement framework that provides early warning signals when a project is drifting.

Earned Value Management: Track Project Health with EVM Metrics

Most project status reports are subjective. “The project is mostly on track with some minor risks.” EVM replaces vague assessments with numbers. It tells you whether a project is ahead or behind schedule, over or under budget, and how those trends project into the future.

The Three Core Measurements

Planned Value (PV) — also called Budgeted Cost of Work Scheduled. This is how much work you planned to complete by today, expressed in dollars or effort hours. It comes from your baseline schedule and budget.

Earned Value (EV) — also called Budgeted Cost of Work Performed. This is how much work you actually completed by today, valued at the budgeted cost. If you planned a $10,000 feature and it is 60% complete, the EV is $6,000.

Actual Cost (AC) — also called Actual Cost of Work Performed. This is how much you actually spent to do the work completed so far.

These three numbers generate every EVM metric.

Key Metrics and What They Tell You

Schedule Variance (SV) = EV - PV

SV > 0: Ahead of schedule (you have earned more value than planned)
SV = 0: On schedule
SV < 0: Behind schedule (you have earned less value than planned)

Cost Variance (CV) = EV - AC

CV > 0: Under budget (you completed work for less than planned)
CV = 0: On budget
CV < 0: Over budget (the work cost more than planned)

Schedule Performance Index (SPI) = EV / PV

SPI > 1.0: Ahead of schedule
SPI = 1.0: On schedule
SPI < 1.0: Behind schedule

Cost Performance Index (CPI) = EV / AC

CPI > 1.0: Under budget
CPI = 1.0: On budget
CPI < 1.0: Over budget

The indices are more useful than the variances because they are ratios. A CPI of 0.85 means you are getting 85 cents of value for every dollar spent, regardless of project size. This makes comparison across projects possible.

A Worked Example

Consider a website redesign project with a total budget of $100,000 and a 20-week timeline. At week 10:

PV: According to the plan, $50,000 of work should be done by week 10
EV: The team has actually completed work worth $42,000 (at budgeted rates)
AC: The team has spent $48,000 to complete that work

Calculations:

SV = $42,000 - $50,000 = -$8,000 (behind schedule)
CV = $42,000 - $48,000 = -$6,000 (over budget)
SPI = $42,000 / $50,000 = 0.84 (16% behind schedule)
CPI = $42,000 / $48,000 = 0.875 (getting 87.5 cents per dollar)

This project is both behind schedule and over budget. The numbers are small today but the trend matters enormously.

Forecasting: Estimate at Completion (EAC)

EVM’s real power is forecasting final costs and dates. The most common formula:

EAC = BAC / CPI

Where BAC is Budget at Completion (the original total budget).

For our example: EAC = $100,000 / 0.875 = $114,286

If current cost performance continues, the project will cost $114,286 — a $14,286 overrun. That is an early warning at week 10 rather than a surprise at week 20.

Estimate to Complete (ETC) = EAC - AC

ETC = $114,286 - $48,000 = $66,286 remaining to spend.

Variance at Completion (VAC) = BAC - EAC

VAC = $100,000 - $114,286 = -$14,286 projected overrun.

Making EVM Work in Practice

Measuring Earned Value

The hardest part of EVM is measuring EV accurately. Methods include:

0/100 rule. A task earns zero value until it is complete, then 100%. Simple but inaccurate for long tasks. Best for tasks under one week.

50/50 rule. A task earns 50% when started and 50% when complete. Eliminates arguments about mid-task progress.

Weighted milestones. A task is divided into milestones, each with an assigned percentage. “Design complete” = 30%, “Development complete” = 70%, “Testing complete” = 100%. This balances accuracy with measurement effort.

Percent complete. The team estimates actual completion percentage. This is subjective and prone to the “90% complete for three weeks” problem. Use it only as a last resort.

Choosing the Right Granularity

EVM works best when work packages are small enough to complete within one or two reporting periods. A work package that spans three months makes earned value measurement imprecise because progress estimation within it is guesswork.

EVM in Agile Environments

Traditional EVM assumes a baseline plan with defined scope — which seems incompatible with Agile’s embrace of changing requirements. But Agile teams can apply EVM concepts at the release level:

PV: Planned story points per sprint times the cost per point
EV: Actual story points completed times the cost per point
AC: Actual spending per sprint

Velocity tracking already captures half of this data. Adding cost data per sprint creates a lightweight EVM system that works with Agile cadences. The SPI maps directly to velocity trends — if the team consistently delivers fewer points than planned, SPI will be below 1.0.

Reporting EVM to Stakeholders

Present EVM data visually. An S-curve chart showing PV, EV, and AC over time gives stakeholders an immediate picture of project health. When EV falls below PV, the gap is schedule variance. When AC rises above EV, the gap is cost variance.

For status reports, lead with the CPI and SPI trends rather than raw numbers. A declining CPI tells a clearer story than a dollar figure. “Our cost efficiency has dropped from 0.95 to 0.87 over the last three reporting periods” communicates urgency better than “we are $6,000 over budget.”

EVM is not a replacement for good project management judgment — a project with perfect EVM numbers can still fail if it is building the wrong thing. But for tracking whether the plan is being executed as intended, no other technique provides the same clarity with the same objectivity.