Why Dashboards Lie.
Written and narrated by Serguei Poppeleer.
Your dashboard is green. Call your site supervisor and ask what actually happened today. Then look at the dashboard again.
Dashboards don’t lie on purpose. They lie by latency. They show last week’s numbers, entered by people who entered what was asked of them — not necessarily what was true. The colour is real. The picture is old.
Green is the most dangerous colour on a programme.
Often it isn’t the colour of “under control.” It’s the colour of “nobody looked hard enough.” A transformer delivery shows confirmed. A cable-routing decision shows closed. Both are green because someone typed green — not because anyone checked.
I have walked sites in airports, power plants and pipelines, on four continents, where the dashboard was green and the work had quietly stopped weeks earlier.
They are not the same discipline, and one cannot stand in for the other — no matter how good the visualisation. A beautiful dashboard built on stale, unverified inputs is not control. It is a very professional way to be surprised on schedule.
Dashboards are not the enemy. Unverified dashboards are.
The test that tells you the truth
Here is the test. When the dashboard and the site disagree, which one do you believe? If the honest answer is “the dashboard,” you are not in control of the programme. You are in a meeting about the programme.
Make it concrete: ask the site for one date, one blocker, one decision — then compare all three to the dashboard. The gaps are your real status.
Most owner-side teams have never run that test deliberately. They assume alignment because the inputs arrive on time and the format is familiar. But arrival is not verification. A date marked “confirmed” by email is not a date checked against capacity. A milestone shown amber two weeks after it slipped is not an early warning — it is a receipt.
What happened: the reporting stayed clean for months while the position on the ground drifted.
Why everyone thought it was normal: every input arrived on time, in the right format, and turned the right colour. Nothing looked wrong because the system was built to show that nothing was wrong.
Why it wasn’t: by the time the colour changed, six months of idle had stacked up that nobody had owned in real time — and it had become a €30M+ conversation. The position was recovered. The months were not.
The distance between reporting and control is where claims are born.
What to check before you trust the colour
Updated means someone typed a number. Verified means someone checked it against reality. A dashboard of updated-but-unverified inputs is fiction with good production values.
If the answer is “next month’s pack,” your reporting latency is your real risk — not the item itself.
If amber triggers a decision, an owner and a date, you have control. If it triggers a discussion, you have decoration.
The objections this raises
“But our reporting is best-in-class.”
The quality of the visualisation has nothing to do with the quality of the inputs. The better the dashboard looks, the more trust it borrows — and the more dangerous a stale input becomes, because nobody questions a polished pack.
“We’d see a real problem — people would escalate.”
People escalate what the system surfaces. If the system shows green, the instinct is to trust it, not to go digging. The dashboard does not just report the silence. It manufactures it.
AI can close the latency gap — watching the inputs that feed the dashboard and flagging when a confirmed date stops matching reality, faster than any monthly cycle. That is real value.
But automating a dashboard built on unverified inputs only produces a faster, prettier lie. You standardise and verify the inputs first; only then does acceleration help. Architecture before automation, every time.
AI can tell you the colour is wrong sooner. It cannot make a colour mean control.
The line worth keeping
A control is only real if it changes what happens next week. A green dashboard that changes nothing anyone does is not control. It’s decoration.
So the question was never “is it green?”
The question is: if it weren’t — would you know in time to do anything about it?
The colour is the last thing to know. By then it is news, not control.
Where does your delivery system actually stand?
The CONTROL diagnostic measures your governance architecture across the five dimensions of The 5Cs Framework™. Twenty questions. Four minutes. Instant result.
Start with the CONTROL diagnostic → Read the core governance report →