Earned Value Management: The Complete EVM Guide

Earned value management (EVM) is a project performance measurement methodology that integrates scope, schedule, and cost into a single, objective picture of where a project stands and where it is heading. It was developed by the US Department of Defense in the 1960s, formally standardised as ANSI/EIA-748 in 1998, and is now used across defence, aerospace, software development, and infrastructure programmes worldwide.

The core idea is deceptively simple. Traditional cost tracking tells you what you spent. EVM tells you what you got for what you spent. I've sat in project reviews where a cost report showed spend "on track" while the EVM picture told a completely different story. Nobody was lying. They just didn't have the tools to see what was actually happening.

Each section below links to a dedicated spoke page with deeper detail and worked examples. For 100+ EVM term definitions, visit the EVM Definitions hub.

1. What is Earned Value Management?

Earned value management integrates scope, schedule, and cost into a single, objective picture of project health, answering four questions at any point in the project: How much of the work is done? What should it have cost? What did it actually cost? Based on current performance, what will the whole project cost?

EVM was created because cost reports alone are misleading. Knowing you've spent 40% of your budget tells you nothing if you don't know whether you've delivered 30% or 50% of the work. EVM fixes that by converting physical progress into pound terms. The methodology dates to 1967, standardised as ANSI/EIA-748 in 1998, and is now embedded in PMI's PMBOK Guide and the APM Body of Knowledge.

2. The Three Core Measurements: PV, EV, and AC

• Planned Value (PV): the budgeted cost of work scheduled to date (also called BCWS).
• Earned Value (EV): the budgeted value of work actually completed, calculated as BAC x % Complete (also called BCWP). You're measuring physical achievement, not expenditure.
• Actual Cost (AC): everything you've spent to deliver that completed work.
• Budget at Completion (BAC): the total approved project budget. BAC doesn't change unless scope changes, and every forecasting formula uses it as its anchor.

Compare EV against PV and you know whether you're behind or ahead of schedule. Compare EV against AC and you know whether you're over or under budget. That's the entire diagnostic logic of EVM.

3. The 7 Essential EVM Formulas

The master reference table. Every formula you need, in one place.

CPI above 1.0: under budget. SPI above 1.0: ahead of programme. The spoke page has full interpretation tables with threshold guidance for each metric.

4. Interpreting CPI and SPI: What the Numbers Mean

CPI and SPI compress the entire health of a project into two ratios. CPI = EV / AC. SPI = EV / PV. Above 1.0 means under budget or ahead of programme. Below 1.0 means overspending or behind. A CPI of 0.85 means you're spending £1.18 for every £1.00 of value. Research from the US DoD found CPI rarely recovers by more than 10% once it drops below a threshold at the 20% completion mark. If CPI is 0.85 at the one-quarter mark, build your forecast around 0.85, not a hoped-for recovery.

SPI has one critical flaw: it converges to 1.0 as the project approaches completion regardless of whether it finishes on time, making it unreliable after roughly 70% completion. Use time-based SPI(t) or programme float analysis for late-stage schedule tracking.

5. Cost Variance and Schedule Variance

Variances give you the same information as CPI and SPI but in pound terms, useful for reporting to project boards. CV = EV - AC. SV = EV - PV. Positive means under budget or ahead of programme; negative means the opposite. If EV is £6.4M and AC is £7.1M, CV = -£700,000: you've spent £700K more than the work is worth. Like SPI, SV converges to zero at project completion regardless of whether the project finishes on time. Use it for trend analysis in the early and middle phases only.

6. Forecasting Final Cost: EAC, ETC, and TCPI

Variances tell you where you are. Forecasts tell you where you'll end up. There are four EAC variants, each based on a different assumption about the remaining work. Always calculate Variant 1 first.

The baseline forecast. Given the CPI stability rule, this is usually the right assumption after 25% completion. Variant 2 (AC + remaining work at budget) is only valid when you've genuinely fixed the root cause of past overspend. Variant 3 (AC + [(BAC - EV) / (CPI x SPI)]) applies when programme delays are also driving cost: common on construction projects where schedule slippage has a direct cost consequence. Variant 4 (AC + bottom-up ETC) is for major scope changes when the mathematical models no longer reflect reality.

ETC = EAC - AC. How much more will you spend? VAC = BAC - EAC. Negative VAC means over budget at completion. TCPI = (BAC - EV) / (BAC - AC). TCPI above 1.10 is very challenging. Above 1.20 and the budget is effectively unrecoverable.

7. Tracking Progress with S-Curves

An S-curve plots cumulative PV, EV, and AC against time. It gets its name from the characteristic shape of a project spend profile: slow at the start (mobilisation), steep in the middle (peak production), flattening at the end (close-out). Three curves on one chart. The gaps between them tell you everything: the gap between PV and EV is schedule variance; the gap between EV and AC is cost variance.

S-curves also reveal trajectory. A project where EV is catching PV is recovering. A project where AC is accelerating away from EV is deteriorating. The trend matters more than the snapshot, which is why you need at least three periods of data before the chart becomes useful.

8. How to Implement EVM: A Step-by-Step Guide

EVM is straightforward in theory. Getting it running on a live project (with subcontractors who've never heard of CPI, a programme three weeks late, and cost data across six spreadsheets) is different. Five steps:

1. Build the Work Breakdown Structure (WBS). Structure by project, then discipline or package, then work package, then activity. Too few activities and the data is useless; too many and the measurement overhead kills the system.
2. Cost-load the programme. Assign your total budget across activities so that at any date you can state the planned value. Without a cost-loaded programme, you have no PV and can't run EVM.
3. Assign a measurement method to every work package. Choose from: 0/100, 50/50, percent complete, units complete, level of effort, or cost ratio. Get this wrong and your EV data is fiction from day one.
4. Set the baseline and start measuring. Lock BAC, the PV curve, and measurement methods. Begin the monthly cycle: measure physical progress, calculate EV, pull AC from the finance system, run the formulas. Expect the first few cycles to be rough.
5. Report and act. CPI and SPI are only useful if someone acts on them. Flag anything below threshold, recalculate EAC, update the S-curve, track the trend.

9. EVM in Construction: Practical Adaptations

EVM was developed for aerospace and defence, where scope is defined upfront and progress is relatively predictable. Construction is different: weather stops work, subcontractors fluctuate, design changes arrive mid-pour. EVM handles all of this, but the implementation differs.

Measuring Physical Progress

The three most reliable measurement methods are: units complete (metres of pipe installed, piles driven, m² of slab cast: objective, auditable, impossible to game); weighted milestones (budget weights assigned to defined completion points); and disciplined percent complete tied to physical quantities on site. Avoid subjective progress claims wherever you have something countable.

The NEC4 Context

On NEC4 Option C and D, EVM maps naturally onto the contract mechanism. Three differences from generic EVM matter. The Accepted Programme (clauses 31 and 32) is your PV baseline: get it accepted before running EVM. Compensation events update the Prices when implemented, changing BAC: re-baseline the PV curve after each implementation. And the pain/gain share schedule means the Contractor's financial exposure isn't the same as the headline overrun. An EAC showing £4M over target on 80/20 pain share means £800K contractor exposure. Most project reports don't adjust for this, and the board draws the wrong conclusion every time.

10. 12 Common EVM Mistakes

EVM fails on projects for predictable reasons. Here are the twelve most common pitfalls, each covered in depth in the full guide.

1. Measuring EV by spend instead of progress. CPI looks fine on paper while the project bleeds money.
2. Mismatching EV and AC cut-off dates. A seven-day gap creates phantom variances.
3. Not updating BAC after scope changes. If BAC doesn't update, every metric downstream is wrong.
4. Using subjective percent complete. Tie progress to physical quantities wherever possible.
5. Treating SPI as a schedule replacement. SPI measures overall efficiency, not critical path risk.
6. Ignoring the CPI stability rule. After 25% completion, forecast from the current CPI, not a hoped-for recovery.
7. No WBS alignment with contracts. If WBS doesn't match subcontract packages, reconciling EV against AC becomes a monthly nightmare.
8. Including committed costs in AC. Committed costs are a liability: including them suppresses CPI.
9. Rolling up to project level only. A project CPI of 0.95 can mask one package at 1.10 and another at 0.78.
10. Starting EVM too late. EVM at 60% completion, when the forecast is already grim, is not useful.
11. Using Level of Effort for production activities. Applied to production, EV always tracks PV and hides every problem.
12. Reporting without narrative. A CPI of 0.91 in a table tells a number. A sentence explaining why tells what to do next.

11. Worked Example: EVM on a UK Infrastructure Project

CV = -£1,540,000 (over budget). SV = -£1,400,000 (behind programme). CPI = 0.871. SPI = 0.881. EAC (Variant 1) = £32.1M. EAC (Variant 3) = £34.9M. VAC = -£4.1M. TCPI = 1.096.

The project is 37% complete by value at the 42% programme point. With pain share 80/20 (Client/Contractor), the Contractor's exposure on the £4.1M overrun is approximately £820K. That's the number that should go to the board, not the headline figure.

12. EVM Report Template and Monthly Period Close

A good EVM report answers five questions on one page: Where is the project now (PV, EV, AC)? How efficiently is it performing (CPI, SPI, trend over last 3 periods)? What will it cost to finish (EAC, ETC, VAC)? Where are the problems (by work package)? What's being done (actions against below-threshold packages)?

Monthly is standard, aligned with interim valuations. Fortnightly on high-risk projects above £50M. The trend arrows matter more than the snapshot numbers: a CPI of 0.94 heading up is a different situation from 0.94 heading down.