A project director called me into his office and slid a report across the desk.

“Tell me what’s wrong with this.”

It was an EVM report. Beautifully formatted. Colour-coded dashboards.

CPI sitting at 1.02.

SPI at 0.98.

On paper, a project running almost perfectly. I asked him one question. “Who signs off on percent complete each month?”

He paused.

“The contractor’s site manager.”

That was the problem.

The numbers weren’t wrong because someone had made a mistake. They were wrong because percent complete – the single most important input in EVM, was being reported by the person with the most to gain from reporting it high. An activity that was genuinely 40% complete was being reported at 65%.

The EVM output was clean.

The project was in trouble.

This is the thing about Earned Value Management that nobody talks about enough.

Used properly – it’s one of the clearest early warning tools in project controls.

Three numbers tell you everything:

– BCWS – what you planned to spend by now.

– BCWP – the value of work actually completed.

– ACWP – what you actually spent.

From those, SPI and CPI.

SPI below 1.0 – you’re behind.

CPI below 1.0 – you’re overspending.

Simple. Powerful. Objective. But here’s what EVM doesn’t tell you.

Why.

An SPI of 0.85 tells you you’re behind.

It doesn’t tell you whether the critical path has shifted.

Or which package is the problem. Or whether recovery is still possible.

That interpretation requires a human who understands the schedule, not just the metric.

And the second thing EVM can’t tell you:

The future – unless the inputs are honest.

Percent complete is the most manipulated number in project controls.

If the team reports 60% on an activity that’s 40% complete, the EVM output is wrong.

The CPI and SPI are wrong.

The forecast to completion is wrong.

Garbage in. Elegant garbage out.

EVM is only as reliable as the physical progress measurement behind it. That’s the discipline most teams skip.

The project director stared at the report for a moment. Then he said: “So we’ve been making decisions based on numbers that aren’t real.”

Yes.