Nobody read it.

Twenty-five weeks of technique, and here is where most of it quietly dies. Not because the analysis was wrong — because the person who needed to act on it couldn't find it.

Dumping data is not reporting. Reporting is targeted clarity.

The same truth, at six altitudes

Your schedule is one model. But nobody wants the whole model.

At the top, a single line: is this project worth the room it takes in the portfolio? Below that, one page for the board — milestones, money, the date. Then a few pages for senior management, split by building or by system.

Then the monthly, for the project team: phases, indices, the forecast — everything we've built since Week 16.

Then the three-week look-ahead, for the people who run packages: what lands next, and what it needs from whom.

And at the bottom, this week, day by day: which gang, which pour, which crane. That's the foreman's world, and nothing above level five is of the slightest use to him at 6am.

Notice what does not change as you move up: the truth. It's one model, one set of actuals, one critical path. Altitude changes the resolution, never the message. The moment the board's version says something the site version doesn't, you no longer have a reporting system — you have two stories, and you'll eventually have to defend both.

Two ways to be useless

Get the altitude wrong and the report fails, however accurate it is.

Too low: you send the board a 400-line Gantt chart. What you meant was look how thorough we are. What they saw was a wall. They cannot find the decision in it, so they ask you the question directly — and now you've spent two days producing something whose only function was to be talked over.

Too high: you tell the foreman the project is 74% complete. He cannot pour 74% of a slab. That number changes nothing he will do tomorrow, so it is, operationally, silence.

The rule I'd put above the desk: a report that triggers no decision is not a report. It's a document.

So work backwards. Ask what this person actually has to decide this week — approve the money, release the crane, escalate to the client — and send them exactly what that decision needs. Nothing else. The discipline is in what you leave out.

Send them the month they can still change

Now the failure that's harder to see, because the report looks complete.

Almost every monthly report answers two questions: what did we do, and where do we stand? Progress, quantities, indices, critical path. All of it accurate. All of it about the past and the present.

And then it stops.

Nobody can change last month. What management can change is next month — and that only happens if you tell them what's coming: the risks that are hardening, the dependencies about to bite, the assumption that has quietly stopped being true.

"The steel is late" is history. "The steel is late, and unless it lands by the 12th we lose the crane slot and the delay becomes six weeks instead of one" is a decision, and it will get made this week rather than next month.

That's the difference between reporting and reporting usefully. One tells them what happened. The other hands them a lever.

Honest numbers, honest report

One last warning, and it's the one that ends careers.

There's a way to make every number on a report look healthy without lying about a single one. The team picks off the easy work — the accessible, low-risk packages that were scheduled for later. Earned value climbs. The percentage complete looks excellent.

And the critical path hasn't moved an inch.

You've seen every piece of this already: it's Week 17's out-of-sequence progress, feeding Week 19's index that cannot see the critical path. Everything is green, and the project is quietly late.

The defence is the one we built in Week 20. Report schedule performance in time, not in money. Earned schedule doesn't care how much easy work you did — it asks a colder question: how many months of the plan have you actually earned? Easy work done early can't inflate that, because the plan already knew when it was due.

Put a time-based schedule number on your report and the mask comes off. That single change will make you, quietly, the most trusted person in the room.

Both halves matter, and most of us are only good at one. The analyst who gets the maths right and the altitude wrong writes documents nobody opens. The communicator who gets the altitude right and the maths wrong is confidently, persuasively leading people off a cliff. The job is both.

Practical insight

Take your last monthly report and count the pages that describe the past.

Then count the pages that describe what is about to happen, and what someone needs to decide about it.

If the second number isn't at least as big as the first, you're writing history. Accurate, careful, beautifully formatted history — and nobody can do anything about it.

Key takeaways

✔ Dumping data isn't reporting. Targeted clarity is.
✔ One model, six altitudes: portfolio line, board page, management summary, monthly, look-ahead, daily.
✔ Altitude changes the resolution, never the message — two versions of the truth is a system failure.
✔ Too low and the board sees a wall; too high and the foreman hears nothing.
✔ A report that triggers no decision is not a report. It's a document.
✔ Work backwards from what the reader must decide, and leave everything else out.
✔ Most reports cover history and position, then skip the only part anyone can act on.
✔ Easy work done early inflates percent complete while the critical path stands still.
✔ Report schedule performance in time, not money, and the mask comes off.

What's coming next

Twenty-six weeks. A model that's honest, numbers that forecast, evidence that survives a courtroom, and a report that someone will actually act on.

And every single piece of it can still be defeated by one person who doesn't want it to be true.

Next week, the last one: the part of this job that isn't technical at all. Why the best analysis in the world loses to a relationship, what a sponsor is actually for, and how to tell a client something they have decided in advance not to hear.

Projects are built by people. We've spent twenty-six weeks on the mathematics. Let's finish with the harder half.