Governed Is Not Controlled.

Written and narrated by Serguei Poppeleer.

Walk into most capital programmes and the governance looks immaculate. Stage gates defined. RAID logs maintained. Steering decks polished. And still the project drifts.

Looking governed and being in control are not the same thing — and the gap between them is where projects quietly fail.

The reports are green. The packs are thorough. Yet the plan diverged from site weeks ago, three decisions are sitting open with no owner, and a vendor date everyone is relying on was never actually verified. None of that shows up in the governance, because governance describes the project — it does not run it.

There is a fear most PMO and programme leaders never say out loud: being exposed as a paperwork function. Governance perfect on paper, while the site runs on WhatsApp messages and unverified dates. The day a number turns out to be fiction, it is not the document that gets blamed. It is you.

Governance describes the project. Control changes what it does. The two get confused precisely because the paperwork looks so convincing.

The paperwork illusion

The belief underneath most programmes is simple: if the governance is right — the gates, the logs, the reporting — control will follow. So effort pours into the artefacts, and everyone mistakes a complete pack for a controlled project.

But a pack is a description of a moment. It tells you what was true when someone filled it in. It says nothing about whether the plan still matches site, whether the open decisions will close, or whether the dates will hold. Control is not the quality of the description. It is whether the mechanisms underneath are actually doing work.

A green report is not control. It is a claim — and on too many projects, an unverified one.


What control actually is

Real control is not more documentation. It is five mechanisms that do work, whatever the paperwork says. We hold them as the 5Cs: contain the scope, command one plan, close every decision, confirm vendor dates, control the cadence. Each one is a thing that happens, not a thing that is described.

Contain
The scope is written and held

One scope — inclusions, exclusions, assumptions — locked before work starts, so the line does not move quietly.

Command
One plan matches reality

A single operating plan kept matched to site week to week, not a deck that diverged by week two.

Close
Decisions resolve

Every open item carries an owner, a date and a consequence, so decisions close before they cascade.

Confirm
Dates are verified

Commitments checked against capacity and lead times, not accepted on a promise.

Control
The cadence holds

A fixed rhythm that catches slippage while there is still room to move.


How to tell theatre from control

1
Compare the plan to the site, not the deck

Ask when the plan was last matched to what is actually happening. If the answer is the last report, you have a description, not a baseline.

2
Pull three open items at random

Do they each have an owner, a date and a consequence? If not, they are not being controlled — they are being recorded.

3
Trace one date to its evidence

Pick a committed date and ask what confirmed it. If the trail ends at “the contractor said so,” the date is a hope wearing a number.

4
Read what a green status is built on

Green should mean the mechanisms are holding, not that the pack was completed. Ask what would have to be true for the status to be wrong.


The objections

“Our governance is mature — auditors love it.”

Auditors check that the process exists and is followed. That is necessary and entirely different from whether the project is under control. Mature governance and a drifting project coexist all the time; the audit was never designed to catch it.

“We do not have time to add more controls.”

The 5Cs are not more on top of governance — they are what your governance is supposed to produce. Done right, they replace theatre, not add to it. The time goes into mechanisms that work instead of packs nobody acts on.

AI Reality Check

AI raises the stakes here, because it makes the paperwork even more convincing. It can generate status packs and dashboards that look more controlled than ever — from data that may already be detached from site.

A better-looking description of a drifting project is still a drifting project. Used well, AI watches the mechanisms — flagging when the plan diverges, when a decision sits open, when a date is at risk. Used carelessly, it just renders the illusion in higher resolution.

AI can make a project look more governed. It cannot make it controlled.


What control feels like

A camp build in Equatorial Guinea — control without the paperwork

Remote site, frontier market, every standard assumption pointing to a 36-month build and likely claims. There was no elaborate governance stack. There were five mechanisms doing work: scope contained, one plan shared, decisions forced to close, dates verified, a fixed cadence.

The camp was delivered in 28 months at a fraction of typical cost, with no claims and no security incidents. The client released payment on a single note — “We have started the works.” That is what control looks like: not a thicker pack, but a project that does what it said it would.

You can have flawless governance and no control.

The reverse is what you actually want — and it rarely comes from another document. It comes from five mechanisms that hold whether or not anyone is looking at the pack.

Where does your delivery system actually stand?

The Delivery Score measures your governance architecture across the five dimensions of The 5Cs Framework™. Fifteen Yes/No questions. About four minutes. Instant result.

Get my Delivery Score — free → Back to the library →