The Plan Was Fine. The Delivery Wasn't. Why Capital Projects Fail Where No One Is Looking
Almost no capital project fails on paper. The business case was sound, the budget was defensible, the schedule was achievable, and everyone in the room approved it. Two years later it is over budget, behind schedule, and nobody can quite say when it went wrong. The plan did not fail. The delivery did.
This is the most expensive misdiagnosis in capital project management: treating overruns as a planning problem and pouring ever more effort into better plans, longer schedules, and more rigorous estimates — when the value was actually lost downstream, in the long grind of execution and the records that were supposed to keep delivery honest.
Where the Value Actually Leaks Out
Think of a plan as one hundred units of intended value, approved at design. Very little of that value is lost in the planning itself. The leaks come later, phase by phase. And each one is a documentation failure as much as an execution one.
Procurement drift: Scope gets reinterpreted between the plan and the contracts. What was approved and what was actually bought quietly diverge, and no single record reconciles them.
Delivery entropy: Decisions get made in the field faster than they get recorded. Change orders lag, approvals scatter, and the project's real state stops matching any document.
The information gap: People executing cannot quickly find the current drawing, the latest approval, or the reason behind a past decision — so they re-decide, re-work, and re-litigate things already settled.
Closeout debt: The records that should have been captured all along get deferred to the end, where they are expensive, incomplete, or simply never finished.
Notice that none of these are failures of the plan. They are failures of the connective tissue between the plan and the built thing — and that tissue is mostly information.
Why Better Plans Don't Fix a Delivery Problem
The reflex response to a cost overrun is a post-mortem that produces a better plan for the next project. More contingency. Tighter milestones. More sign-off gates. These are rational responses to a planning problem. They are the wrong responses to a delivery problem.
A better plan does not help if the team executing it cannot find the current drawing, cannot confirm which change order is in force, or cannot quickly establish what was decided three months ago and why. The plan will be excellent. The delivery will leak exactly as before, because the information problem is unchanged.
The counterintuitive lever is this: the fastest way to improve delivery is not a better plan or a bigger team — it is a project whose real state is always documented and always findable. When the current drawing, the latest approval, the open decisions, and the actual spend all live in one place everyone trusts, execution stops leaking. People stop re-deciding. Drift gets caught in weeks instead of quarters.
"Great plans do not die in the boardroom. They die in the months afterward, in a thousand small decisions nobody wrote down."
The Diagnostic Question Most Post-Mortems Miss
When a project overruns, the investigation usually traces back to scope changes, market conditions, design errors, or contractor performance. Those findings are real. But they rarely go far enough. The deeper question is: when did the project's real state first diverge from its documented state? Because that is the moment the project became unmanageable — not when the contractor submitted a claim, not when the contingency ran out, but when the record stopped being honest.
For most capital projects, the answer to that question is not a date near the end. It is a month or two after construction commenced, when the pace of field decisions outstripped the pace of documentation, and nobody noticed because the schedule still looked fine.
Four Habits That Keep Delivery Honest
Capture decisions at the point they are made, not at the end of the week. A decision that takes thirty seconds to make takes thirty minutes to reconstruct three months later — if it can be reconstructed at all.
Maintain a single current-version record for drawings and scope. If the team is working from different versions of the same document, the project has already diverged from itself.
Make change order status visible in real time. A change order that exists in paper but has not been formally issued against the contract is invisible risk. It looks like a future problem until it becomes a present one.
Start the closeout file on day one. The records that prove occupancy, compliance, and warranty are built in parallel with the work, not assembled in the final week.
A Harder Question for Every Capital Programme
Before investing in another round of better planning, ask a harder question about the last overrun: was the plan actually wrong, or did the project simply lose the thread during delivery and only discover it at closeout?
For most capital projects, most of the time, it is the second one. The plan was probably fine. The real question is whether delivery left a record honest enough to prove it — and to catch the next drift before it compounds.
Source: XNM Technologies — Why Capital Plans Fail at Delivery, Not Design (https://www.xnm.ca/post/why-capital-plans-fail-at-delivery-not-design)
This content was generated by AI.
Comments