Project Controls in the AI EraCost control and earned value management · Lesson 7 of 22
Earned value management: the core metrics
Video lecture
Earned value management: the core metrics
The narrated lecture is in production
Every chapter is scripted and ready. Browse the chapters and read the full transcript now — the video will appear here when it’s published.
Chapters
Transcript of the narration, chapter by chapter.
0:00 Forty per cent spent: good or bad?
You've spent forty per cent of your budget. Is that good news or bad news? Take a moment. The honest answer is: you can't possibly know. It depends entirely on how much work you've actually done. If you've done fifty per cent of the work, you're doing brilliantly. If you've done thirty per cent, you're in trouble. Comparing spend to budget on its own is one of the most common and most misleading habits in project reporting. In this lecture you'll learn earned value management, properly. The three core values, the four derived metrics, how to read them together, and the limitation of the schedule index that catches out even experienced people. By the end you'll be able to calculate and interpret earned value for any project in a few minutes.
0:57 Why it matters
Why does earned value matter? Because it's the one technique that integrates scope, schedule and cost into a single, objective view. It answers not just 'what did we spend?' but 'what did we get for it, and was that what we planned?'. It's also a powerful early warning system: in practice the cumulative cost performance index tends to settle once a project is meaningfully underway, so early signals are worth taking seriously. And it's widely standardised. ISO 21508 gives international guidance on earned value, ISO 21512 adds implementation guidance, and in the US defence and government sector, the EIA-748 standard sets out the guidelines an earned value management system must meet. The latest revision, E, streamlines them into twenty-seven guidelines. You don't need to memorise standards. You do need to calculate the metrics correctly.
1:55 The concept: three numbers
Here are the three numbers, all measured at the same date. Planned value, PV, is the budgeted value of the work that should have been done by now. It comes from your time-phased baseline. Earned value, EV, is the budgeted value of the work that has actually been done. Actual cost, AC, is what that work actually cost. You'll also see the older names: BCWS, BCWP and ACWP. Now, here's the key idea, and I want you to really hold on to it. Earned value is measured in budget terms, not in spend. If a one hundred thousand pound work package is thirty per cent complete, its earned value is thirty thousand, whatever you actually spent. Think of it as being paid for your work at the agreed price list. It doesn't matter how long it took you or what it cost you. You earn what the plan said that work was worth.
3:02 The derived metrics
From those three numbers come four metrics, and notice that earned value appears in every one of them. Cost variance is EV minus AC. Negative means you're over cost for the work done. Schedule variance is EV minus PV. Negative means less work has been done than planned. The cost performance index is EV divided by AC. Below one means you're getting less than a pound of planned work for each pound spent. And the schedule performance index is EV divided by PV. Below one means you're behind in work accomplished. Two more useful ratios: percent complete is EV divided by budget at completion, and percent spent is AC divided by budget at completion. Here's a memory trick. Earned value always comes first. Subtract or divide by what you're comparing it with.
3:59 Worked example one: a simple check
Let's practise with a quick one. Earned value is four hundred and fifty thousand, actual cost is five hundred thousand and planned value is four hundred thousand. Cost performance index: four fifty divided by five hundred. Nought point nine. So we're over cost: ninety pence of planned work for every pound spent. Schedule performance index: four fifty divided by four hundred. One point one two five. We've done more work than planned by this date. So what's the story? Over cost but ahead. That pattern often means the team is accelerating with overtime or extra crews. Which is fine, if it's a deliberate decision. Not fine if nobody chose it.
4:47 Worked example two: Karachi Gateway Logistics
Now the realistic example from the lesson. Karachi Gateway Logistics, a fictional company, is building a warehouse. Budget at completion, two million dollars, over ten months. End of month four. Planned value, eight hundred thousand. Earned value, seven hundred thousand. Actual cost, eight hundred and fifty thousand. Cost variance: seven hundred minus eight fifty, minus one hundred and fifty thousand. Schedule variance: seven hundred minus eight hundred, minus one hundred thousand. CPI: seven hundred over eight fifty, about nought point eight two. SPI: seven hundred over eight hundred, nought point eight seven five. Percent complete: thirty-five. Percent spent: forty-two and a half. So the project has spent forty-two and a half per cent of its budget to complete thirty-five per cent of its work. A report saying 'we've spent forty-two per cent and we're forty per cent through the timeline' would have hidden both problems.
5:50 Watch me do it: an EVM table in Excel and pandas
Let me show you how I'd set this up. One row per control account, never just a project total. Columns for budget, planned value, earned value and actual cost. Then cost variance is earned value minus actual cost, and CPI is earned value divided by actual cost, wrapped in IFERROR so a zero doesn't break the sheet. Same for SV and SPI. Now I sort by cost variance, ascending. Look what happens. The project CPI is nought point nine, which sounds like a mild concern. But one account, pumps and mechanical, is at nought point eight one and driving most of the overrun. That's where the conversation starts. The lesson text has the same calculation in a few lines of Python with pandas, which is handy when your data comes out of an ERP export every month.
6:50 Reading the combination and the SPI trap
Now read the two indices together. Both below one: over cost and behind, so ask about productivity, rework or under-estimated scope. CPI below one but SPI above: probably accelerating with overtime. CPI above one but SPI below: perhaps under-resourced, or waiting on materials. Both above one: great, or progress is overstated, or the baseline was padded. Always look at trends. A CPI going nought point nine five, nought point nine two, nought point eight eight is far more worrying than a single nought point nine. And here's the trap. SV and SPI are measured in money, not time. As a project nears completion, earned value converges on planned value, so SPI drifts to one, even on a project that finishes two months late. That's why you pair it with the critical path and earned schedule, which we cover in the forecasting module.
7:52 Common mistakes, recap and try this now
Quickly, the common mistakes. Using actual cost as a stand-in for earned value: 'we've spent forty per cent, so we're forty per cent done'. Mixing units, such as earned value in hours but actual cost in money. Missing accruals, which flatter CPI. And calculating only at project level, where healthy accounts hide failing ones. So, to recap. Planned value is what should be done, earned value is what is done at budget rates, and actual cost is what it cost. Earned value comes first in every formula. CPI equals EV over AC. SPI equals EV over PV. Analyse by control account, and watch trends. Your try-this-now: take the Karachi example into month five, with planned value one million, earned value eight hundred and eighty thousand and actual cost one million and forty thousand. Recompute everything. Is CPI improving or getting worse?
Why earned value?
Comparing spend to budget alone is misleading. If you have spent 40% of the budget, is that good or bad? It depends on how much work you have actually done. Earned value management (EVM) answers this by comparing three numbers for the same point in time.
The three core values
| Metric | Also called | Question |
|---|---|---|
| Planned value (PV) | BCWS | How much work (in budget terms) should be done by now? |
| Earned value (EV) | BCWP | How much work (in budget terms) is actually done? |
| Actual cost (AC) | ACWP | What did the work actually done cost? |
The essential insight: EV is measured in budget dollars, not spend. If a $100k work package is 30% complete, its EV is $30k, regardless of what you actually spent.
The derived metrics
Cost variance (CV) = EV − AC (negative = over cost)
Schedule variance (SV) = EV − PV (negative = behind schedule)
Cost performance index (CPI) = EV / AC (< 1.0 = over cost)
Schedule performance index = EV / PV (< 1.0 = behind schedule)
(SPI)
Percent complete = EV / BAC
Percent spent = AC / BACCPI is the most important single number in cost control: a CPI of 0.85 means you are getting 85 cents of planned work for every dollar spent.
Worked example
Illustrative. Karachi Gateway Logistics (fictional) is building a warehouse. BAC = $2,000,000 over 10 months. At the end of month 4:
- PV (cumulative plan) = $800,000
- EV (budgeted value of work completed) = $700,000
- AC (actual cost incurred) = $850,000
Calculations:
| Metric | Formula | Result | Interpretation |
|---|---|---|---|
| CV | 700k − 850k | −$150,000 | Over cost |
| SV | 700k − 800k | −$100,000 | Behind in work accomplished |
| CPI | 700k / 850k | 0.82 | 82 cents of value per dollar |
| SPI | 700k / 800k | 0.875 | 87.5% of planned work done |
| % complete | 700k / 2,000k | 35% | |
| % spent | 850k / 2,000k | 42.5% | Spending ahead of progress |
The project has spent 42.5% of its budget but completed only 35% of its work. Reporting "we have spent 42.5% and are 40% through the timeline" would have hidden both problems.
Reading the combination
| CPI | SPI | Typical story | Typical questions |
|---|---|---|---|
| < 1 | < 1 | Over cost and behind | Productivity? Rework? Under-estimated scope? |
| < 1 | > 1 | Over cost but ahead | Accelerating with overtime or extra crews? |
| > 1 | < 1 | Under cost but behind | Under-resourced? Waiting on materials or approvals? |
| > 1 | > 1 | Under cost and ahead | Is progress overstated? Is the baseline padded? |
Always look at trends over several periods, not one data point. A CPI that goes 0.95, 0.92, 0.88 is a much bigger worry than a single 0.90.
A known limitation of SPI
SV and SPI are measured in money, not time. As a project nears completion, EV converges on PV, so SPI drifts toward 1.0 even if the project finishes late. A project two months late will show SPI = 1.0 at completion. That is why schedule analysis should also use the critical path and earned schedule (covered in the forecasting module).
Level of analysis
Calculate EVM at the control account level and roll up. Project-level indices can hide problems: a CPI of 0.98 overall may combine a very healthy civil package with a failing MEP package. Rank control accounts by CV and review the worst first.
Common mistakes
- Using AC as a proxy for EV ("we spent 40%, so we are 40% done").
- Mixing budget and actual units (e.g., EV in hours, AC in dollars).
- Missing accruals, which inflate CPI.
- Reading SPI without checking the critical path.
- Calculating EVM only at project level.
Template: monthly EVM table
Control account | BAC | PV | EV | AC | CV | SV | CPI | SPI | Comment
1.2 Site prep | | | | | | | | |
1.3 Structure | | | | | | | | |
1.4 MEP | | | | | | | | |
TOTAL | | | | | | | | |Hands-on: the EVM table in Excel
Columns: A Control account | B BAC | C PV | D EV | E AC
F CV =D2-E2
G SV =D2-C2
H CPI =IFERROR(D2/E2, "")
I SPI =IFERROR(D2/C2, "")
J %Comp =IFERROR(D2/B2, "")
K %Spent=IFERROR(E2/B2, "")
Totals: sum B:E, then recompute H and I from the totals (never average the indices).Averaging CPIs across accounts is a classic error: the project CPI is SUM(EV)/SUM(AC), which weights each account by its size.
Hands-on: the same table in Python (pandas)
import pandas as pd
df = pd.DataFrame({
"account": ["1.1 Management", "1.2 Civil", "1.3 Pumps & mech", "1.4 Electrical"],
"BAC": [1.2, 2.4, 5.0, 3.4], "PV": [0.60, 1.90, 2.80, 1.20],
"EV": [0.60, 1.80, 2.20, 1.00], "AC": [0.62, 1.85, 2.70, 1.08],
})
df["CV"], df["SV"] = df.EV - df.AC, df.EV - df.PV
df["CPI"], df["SPI"] = df.EV / df.AC, df.EV / df.PV
tot = df[["BAC", "PV", "EV", "AC"]].sum()
print(df.sort_values("CV").round(2))
print(f"Project CPI {tot.EV / tot.AC:.2f} SPI {tot.EV / tot.PV:.2f}")Output: project CPI 0.90 and SPI 0.86, with account 1.3 (CPI 0.81) at the top of the list. Guard against division by zero in real data (df.AC.replace(0, pd.NA)).
Standards note
Earned value is described in ISO 21508:2018, with implementation guidance in ISO 21512:2024; US government contractors work to the EIA-748 guidelines (Revision E, published in 2026, consolidates them into 27 guidelines). Your organisation's EVM system description takes precedence for local rules such as thresholds and earning methods.
How to measure success
- EVM calculated monthly at control-account level, with totals computed from summed values.
- CPI and SPI trends (at least six periods) shown on every dashboard.
- Variance explanations written for every account beyond threshold.
Key takeaways
- PV = planned work, EV = work done (in budget terms), AC = actual cost of that work.
- CV = EV − AC, SV = EV − PV, CPI = EV/AC, SPI = EV/PV; values below 1.0 are unfavourable.
- Analyse at control-account level and watch trends across periods.
- SPI converges to 1.0 at completion, so pair it with critical path analysis and earned schedule.
Check your understanding
Quick questions to lock in the lesson. They don’t count towards your certificate.
Put it into practice
Using the Karachi example, recompute all metrics for month 5 assuming PV = $1,000k, EV = $880k, AC = $1,040k. Is the CPI trend improving or worsening?
Enrol for free to save your progress
Reading is always free. Enrol to keep your place, take the final assessment and earn a verifiable certificate.