Earned Value Management (EVM): CPI, SPI and Forecasts

What you'll learn

Define planned value (PV), earned value (EV) and actual cost (AC) in plain English.
Calculate cost and schedule variances (CV, SV) and performance indexes (CPI, SPI).
Forecast the final cost with three EAC formulas and know when to use each.
Use TCPI to test whether the original budget is still realistic.
Turn the numbers into a clear status message and a recovery decision.

📥 Practice this lesson

The EVM sheet contains this lesson's example. Type the percent complete of each work package and the actual cost; CV, SV, CPI, SPI, EAC, ETC, VAC and TCPI are calculated for you.

project-management-template.xlsx · EVM sheet

Earned value management (EVM) measures project performance by comparing three numbers: how much work you planned to do by now, how much work you actually did, and how much you actually spent. Together they tell you whether you're ahead or behind schedule, over or under budget, and where the project will probably end up. In the last lesson you saw that comparing planned and actual spending alone can mislead. EVM fixes that by adding the missing piece: the value of the work completed. Here you'll run it on the Denver app project from the budget lesson, at the halfway point.

3inputs: planned value, earned value and actual cost

1.00is the target for both CPI and SPI; below 1 means worse than planned

$220Kis the forecast final cost in this lesson's example, against a $176K budget

Gather PV, EV, AC
Calculate variances
Calculate indexes
Forecast
Decide and report

Step 1: the three numbers

Planned value (PV)

The budgeted cost of the work scheduled to be done by the status date.

Where from: the S-curve (time-phased cost baseline).

Earned value (EV)

The budgeted cost of the work actually completed by the status date.

How: each work package's budget × its percent complete.

Actual cost (AC)

What you really spent on the work completed, by the status date.

Where from: timesheets, invoices and accounting.

Plus one more: the budget at completion (BAC), the total cost baseline. For the Front Range Fitness app, BAC is $176,000 over six months, February through July. (The $16,000 contingency has been allocated to the riskiest work packages, which is why they add up to the full baseline.) The status date is the end of April, month 3 of 6.

Calculating earned value

Work package	Budget (BAC)	% complete	Earned value
UX design	$16,000	100%	$16,000
Back end and API	$50,000	60%	$30,000
iOS app	$40,000	35%	$14,000
Android app	$40,000	25%	$10,000
QA and launch	$30,000	0%	$0
Total	$176,000	39.8%	$70,000

From the S-curve, the planned value through April is $79,200. Accounting reports an actual cost of $87,500. So: PV = $79,200, EV = $70,000, AC = $87,500.

Percent complete must be honest

EVM is only as good as the percent complete. Many teams use simple rules instead of guesses: 0/100 (no credit until done) for small packages, 50/50 (half at start, half at finish), or measurable milestones ("API endpoints done: 12 of 20"). "It's 90% done" for three weeks running is a red flag.

Step 2: variances (in dollars)

Cost variance      CV = EV − AC = 70,000 − 87,500 = −$17,500  (over budget)
Schedule variance  SV = EV − PV = 70,000 − 79,200 = −$9,200   (behind schedule)

Negative means bad for both. The team has spent $17,500 more than the value of the work it delivered, and it has completed $9,200 less work than planned by this date.

Step 3: performance indexes (ratios)

Cost performance index      CPI = EV ÷ AC = 70,000 ÷ 87,500 = 0.80
Schedule performance index  SPI = EV ÷ PV = 70,000 ÷ 79,200 = 0.88

Indexes are easier to compare than dollar variances. A CPI of 0.80 means the project gets 80 cents of work for every dollar spent. An SPI of 0.88 means it's progressing at 88% of the planned rate. Here's how to read any combination:

CPI ≥ 1, SPI ≥ 1Under budget, ahead of schedule

Action: Check that estimates weren't padded; keep going.

Best case

CPI ≥ 1, SPI < 1Under budget, behind schedule

Action: Often understaffed: spending less because fewer people are working.

Add capacity?

CPI < 1, SPI ≥ 1Over budget, ahead of schedule

Action: Paying for speed (overtime, extra crew). Is the speed needed?

Slow the spend?

CPI < 1, SPI < 1Over budget and behind

Action: Escalate now with options; the gap rarely closes by itself.

Front Range Fitness

1. A project has an earned value of $45,000 and an actual cost of $50,000. What's the CPI and what does it mean?

1.11: under budget
0.90: over budget, getting $0.90 of work per dollar spent
0.90: behind schedule

CPI = EV ÷ AC = 45,000 ÷ 50,000 = 0.90. CPI measures cost efficiency, not schedule; for schedule you'd need planned value (SPI = EV ÷ PV).

2. EV is $60,000 and PV is $75,000. What's the SPI?

0.80: the project is progressing at 80% of the planned rate
1.25: ahead of schedule
−$15,000

SPI = EV ÷ PV = 60,000 ÷ 75,000 = 0.80. −$15,000 is the schedule variance (EV − PV), the same information in dollars.

Step 4: forecast the final cost

The most useful question for the sponsor is "where will we end up?" The estimate at completion (EAC) answers it. Which formula you use depends on what you believe about the rest of the project:

Assumption	Formula	Front Range Fitness
Past cost performance will continue (most common)	EAC = BAC ÷ CPI	176,000 ÷ 0.80 = $220,000
The overrun was a one-time event; the rest goes as planned	EAC = AC + (BAC − EV)	87,500 + 106,000 = $193,500
Cost and schedule problems both continue (catching up costs money)	EAC = AC + (BAC − EV) ÷ (CPI × SPI)	87,500 + 106,000 ÷ 0.707 ≈ $237,414

Then derive the rest from the most likely EAC (here, the first one):

Estimate to complete   ETC = EAC − AC  = 220,000 − 87,500  = $132,500
Variance at completion VAC = BAC − EAC = 176,000 − 220,000 = −$44,000
Rough time forecast        = 6 months ÷ SPI = 6 ÷ 0.884 ≈ 6.8 months

Values are cumulative at the end of each month. The red line (AC) runs above the green line (EV): over budget. The green line runs below the blue line (PV): behind schedule.

3. A project's BAC is $200,000 and its CPI is 0.80. If cost performance continues, what's the EAC?

$160,000
$240,000
$250,000

EAC = BAC ÷ CPI = 200,000 ÷ 0.80 = $250,000. Multiplying instead of dividing ($160,000) is the most common mistake: a CPI below 1 must make the forecast go up.

Is the original budget still achievable? TCPI

The to-complete performance index (TCPI) tells you how efficient the remaining work must be to finish on the original budget:

TCPI = (BAC − EV) ÷ (BAC − AC) = (176,000 − 70,000) ÷ (176,000 − 87,500)
     = 106,000 ÷ 88,500 ≈ 1.20

The team has been working at 0.80 and would need to jump to 1.20 for the rest of the project. That's not a plan, it's a wish. When TCPI is well above the current CPI (a common rule of thumb is more than about 0.1 higher), stop promising the original budget and present a new forecast.

4. BAC is $100,000, EV is $60,000 and AC is $70,000. What's the TCPI, and what does it tell you?

0.75: the remaining work can be done less efficiently
1.33: every remaining dollar must deliver $1.33 of work to finish on budget
0.86: the project is on track

TCPI = (100,000 − 60,000) ÷ (100,000 − 70,000) = 40,000 ÷ 30,000 = 1.33. With a current CPI of 60,000 ÷ 70,000 = 0.86, reaching 1.33 is very unlikely.

Step 5: decide and report

Numbers aren't a status report. The PM's message to the sponsor, in four sentences:

Where we are: 40% of the work is done; 45% was planned by now, and we've spent 50% of the budget.
Why: the payment API integration took twice the planned hours (risk A from the risk register happened).
Where we're heading: about $220,000 and roughly 7 months if nothing changes, versus $176,000 and 6 months.
Options: move in-app payments to version 1.1 (saves about $30,000), add $44,000 of funding, or a mix. Decision needed by May 7.

Note what the $8,800 management reserve can and can't do: it covers only a fifth of the forecast $44,000 overrun. The overrun is too big for reserves, so it's a sponsor decision about scope, money or time, which is exactly why you want EVM to flag it at month 3 instead of month 6.

Common EVM mistakes

Using "percent of budget spent" as percent complete.

EV comes from work done, never from money spent; otherwise CPI is always 1.

Multiplying BAC by CPI to forecast.

Divide: EAC = BAC ÷ CPI. A CPI below 1 must raise the forecast.

Hoping a low CPI will recover on its own.

Check TCPI; if it's far above CPI, re-forecast and decide.

Reporting only indexes.

Translate them into dollars, dates, causes and options.

Practice exercise

The Dayton garden center has a cost baseline (BAC) of $304,500 after the approved change. At the mid-project status date, the planned value is $152,250, the work is 45% complete and the actual cost is $130,000.

Calculate EV, CV, SV, CPI, SPI, EAC (= BAC ÷ CPI), ETC and VAC, and describe the project's situation in one sentence.

See solution

EV  = 45% × 304,500        = $137,025
CV  = 137,025 − 130,000    = +$7,025   (under budget)
SV  = 137,025 − 152,250    = −$15,225  (behind schedule)
CPI = 137,025 ÷ 130,000    ≈ 1.054
SPI = 137,025 ÷ 152,250    = 0.90
EAC = 304,500 ÷ 1.054...   ≈ $288,889
ETC = 288,889 − 130,000    ≈ $158,889
VAC = 304,500 − 288,889    ≈ +$15,611

(Use the unrounded CPI; rounding it to 1.05 gives $290,000.) In one sentence: the job is spending efficiently but running 10% behind schedule, the classic pattern of a crew that's too small. With about $15,600 forecast under budget, the contractor can afford to add a crew to the critical path and recover the opening date.

Summary

PV = planned work, EV = completed work (budget × % complete), AC = actual spending.
CV = EV − AC and SV = EV − PV; negative means over budget or behind schedule.
CPI = EV ÷ AC and SPI = EV ÷ PV; 1.00 is on plan, below 1 is worse than planned.
EAC = BAC ÷ CPI is the typical forecast; ETC = EAC − AC and VAC = BAC − EAC.
TCPI shows the efficiency needed to hit the budget; if it's far above CPI, re-forecast and give the sponsor options.

Next lesson: Leading the Project Team and Resolving Conflict. You'll build a team charter, guide a team through its stages and handle conflict with the right approach for each situation.