Tools already calculate earned value, the skill now is reading it

Any decent tool calculates CPI and SPI instantly. Knowing how to divide stopped being the skill. The skill is reading what that variance is saying and deciding before it turns into a crisis.

Why the arithmetic stopped being the problem

For years, earned value was taught as arithmetic. Three input numbers, four output formulas, and whoever remembered them passed. That made sense when you were the one doing the maths.

Today the calculation arrives already solved. And when the number shows up on its own, the real problem appears, which is what you do with it. If you need to refresh where the figures come from, we walk through it with a numeric example in earned value explained clearly. This post starts where that one ends.

What each index is saying

Both indices are ratios and both compare against the value of what you actually delivered.

CPI divides earned value by actual cost. It answers how much value you are producing for every unit you spend. A CPI of 0.85 says every unit spent is generating eighty-five cents of useful work.

SPI divides earned value by planned value. It answers what share of the work you planned to have finished by today is actually finished. An SPI of 0.90 says you are at ninety per cent of what was committed for today.

In both, 1.0 is the waterline. Above it, better than planned. Below it, worse. That is everything the number tells you on its own, and it is less than people think.

The four boxes come from the standard definition of the two indices, with 1.0 as the reference. The diagram splits no figures the source does not state.

Two projects with the same CPI and opposite decisions

This example is ours, not PMI’s, and it shows why the number is not enough.

Two projects close the month with a CPI of 0.85. Identical.

In the first, the gap comes from an outside consultancy that had to be hired in the first six weeks and was not in the baseline. That spend already happened and does not repeat. The last three readings were 0.71, then 0.79 and now 0.85. The index is climbing because the normal work is in fact performing. Here the sensible decision is to leave the scope alone, update the forecast of final cost and keep the same cadence.

In the second, the gap comes from every iteration costing more than estimated, because the team ended up smaller than planned and the difference is being paid in overtime. The last three readings were 0.94, then 0.89 and now 0.85. The index is falling because the problem is structural and grows on its own. Here the honest conversation is about scope or about budget, and waiting another month makes it more expensive.

Same number, opposite decisions. The 0.85 decided nothing. What decided was understanding where it came from.

Where the quick read breaks down

• SPI gets read as if it were time. It is not. It is built from money, not days, and it has an arithmetic trap at the end. When the project finishes, earned value and planned value both land on the total budget, so SPI returns to 1.0 even if you delivered late. Right when you would most want a warning, it stops warning.
• A number without a trend. A single index is a photograph. Three in a row are a film, and the film is what tells you whether you are climbing out of the hole or falling into it.
• The average that hides the problem. A project with two workstreams can show a healthy CPI because one offsets the other. The total looks fine and inside there is a fire.
• The threshold nobody defined. Is 0.95 a problem? It depends on the threshold you agreed, and if nobody agreed one, every meeting argues it again.

How this shows up on the exam

Before you allocate your study hours, one fact is worth having. The 2026 exam content outline does not name earned value. Not earned value, not EVM, not CPI, not SPI. You can download the full document and search, they are not there.

What it does name is the job earned value solves. In the Process domain, which takes 41% of the exam, there is a task dedicated to evaluating project status that asks you to develop project metrics, measure and analyse, update those metrics and communicate status. The schedule task asks you to analyse schedule variation. The finance task asks you to monitor financial variations and work with the governance process. And in Business Environment you find defining success metrics and outlining governance escalation paths and thresholds.

Translated, earned value is not exam content by name. It is one of the techniques you use to do a job the outline does name. That is why the exam will not ask you for the formula. It will put an index on the table, inside a scenario, and watch what you do with it. That is exactly the mindset the new exam asks for, applied to a topic with numbers in it.

The machine hands you the reading. What nobody will hand you finished is the decision.

Join the waitlist

At The PM Architect we write the PMBOK study guides in five languages, with real classroom cases and exercises that come with an explanation. They will be free. Leave your email on the home page and we’ll let you know the moment they’re ready.