Two numbers, one diagnosis.

There are only two operations. You subtract, or you divide. And earned value always goes first.

Subtract, and divide

Subtract, and you get a variance — a gap, in money.

Divide, and you get an index — a rate, as a decimal. Below 1.0 is bad. That's the whole convention.

Compare what you built to what you paid, and you learn about cost. Three hundred thousand of building, 380 thousand of invoices: you're 80 thousand down. Every dollar bought 79 cents.

Compare what you built to what you promised, and you learn about time. Three hundred against four hundred: a hundred thousand of promised work simply isn't there. You're building at three-quarters of the planned rate.

So: 0.79 and 0.75.

Two numbers, and they've just told you more about this project than a forty-page monthly report.

What the numbers actually mean

Here's the translation worth memorising, because it's what makes these useful rather than academic.

0.79 on cost doesn't mean "we're 21% over." It means: for every dollar we spend, we get 79 cents of building. It's a rate of conversion. It's your project's exchange rate between money and concrete.

0.75 on schedule means: we're producing work at three-quarters of the rate we said we would. Not "we're 25% late" — a rate.

And rates are the point, because rates continue. Unless something changes, next month you'll convert money into building at the same lousy 79 cents. This is why an early CPI is so brutally predictive: it isn't a snapshot of a bad month, it's the exchange rate your project runs at.

Which raises the obvious question — if we keep going like this, what will the whole thing cost? That's next week.

Four projects, four phone calls

Two indices means four combinations. And every combination is a genuinely different project with a genuinely different problem.

Fast but expensive is the project buying its schedule with money — overtime, extra crews, air-freighted steel. Sometimes that's exactly right. But it should be a decision, not something you discover in a report.

Cheap but slow is the one people misread most. The cost looks great! It looks great because you haven't done the work. You're "saving money" the way you save money by not turning up. Find the missing crews.

Both above 1.0 is either genuinely excellent, or your estimate was soft and everyone's coasting. Worth knowing which.

And both below — where our project sits, at 0.79 and 0.75 — is the honest word: trouble. Slow and expensive. This isn't a project having a bad month. It's a project whose plan no longer describes it.

The point of no return

There's an uncomfortable rule of thumb worth carrying, and our project is standing right on top of it.

We're 20% off on cost and 25% off on the work volume. And beyond about 15% deviation, a construction project does not come back to its baseline. Not with overtime, not with a recovery plan, not with a stern meeting.

The maths is unforgiving: to recover, you now have to perform better than you originally promised — on the same site, with the same crews, in less time than remains. If you couldn't hit the plan when things were normal, you will not beat the plan while things are broken.

This is where the profession divides. One planner writes "recovery plan in place" and hopes. The other says out loud: the baseline is gone; here are three ways forward and what each one costs.

The second one is doing the job. And they are far more comfortable saying it in month six than in month eleven.

What SPI cannot see

Now the caveat, and it's a big one. Do not skip it.

Everything we just calculated is measured in money. Even the schedule index. SPI counts the dollar value of missing work — and it has absolutely no idea which work is missing.

Two projects, both reporting SPI 0.90. Both "10% behind."

On the first, the missing work is landscaping — forty days of float, nobody waiting on it. The building hands over on time.

On the second, the missing work is the structure. Everything downstream is standing still behind it. Six weeks late, and getting worse.

Identical number. Completely different project.

So here is the rule: SPI tells you how much work is missing. Only the critical path tells you whether it matters. Read them together, always. An earned value report without a critical path beside it is half a diagnosis — and it's the half that can't see the delay coming.

There's a way to fix this properly, by measuring schedule performance in time instead of money. We'll get to it next week.

CPI and SPI are two of the best numbers in construction. They are still only numbers. They will tell you that your project is slow and expensive with beautiful precision — and they will never tell you that the reason is a subcontractor who stopped answering the phone in March.

Practical insight

Calculate your CPI once. Just once, on the package you're most worried about.

Take the physical percent complete, multiply by the budget, and divide by what's actually been spent. If it comes back below 1.0, you now know your exchange rate — and you can multiply it out to see where the whole package lands.

Then look at where the missing work sits. On the critical path, or in the float? Because that answer, and not the index, decides whether this is a cost problem or a crisis.

Key takeaways

✔ Subtract for a variance, divide for an index — earned value always comes first.
✔ CPI is your exchange rate: how much building each dollar actually buys.
✔ SPI is your production rate against the promise, not a percentage of lateness.
✔ Rates continue — which is why an early CPI predicts the ending.
✔ Four index combinations, four different projects: fix the right one.
✔ "Cheap but slow" isn't saving money — it's work that hasn't happened yet.
✔ Past roughly 15% deviation, the baseline is gone. Say so early, with options.
✔ SPI counts missing work in money; only the critical path knows if that work matters.

What's coming next

We know exactly how this project is performing. We still don't know how it ends.

Next week we forecast. If a project converts money into building at 79 cents on the dollar, what does the finished job cost — and what efficiency would we need, from tomorrow, to still hit the original budget? That last number is the most honest thing in project controls, because it tells you when a recovery plan is arithmetic and when it's a fantasy.

And we'll fix SPI's blindness with a measure that finally works in days instead of dollars.