The biggest challenge facing today’s megaprojects isn’t cost escalation. It is the hidden cost of delivering multiple megaprograms that all compete for the same people, equipment and infrastructure at the same time.
Construction has never received so many project requests and never delivered so few. Of every 10 data centres announced, only one leaves the ground. One.
The other nine remain paper cathedrals, held down by a grid connection that slips, a transformer years away, a team that cannot be hired. We assume the tragedy of the megaproject is the overrun. It is not. The tragedy is that most never exist.
I have spent my professional life on both sides of the number, building budgets for those who construct, challenging them for those who decide. What holds these megaprojects down has a name. It appears in no bill of quantities: the simultaneity premium.
“Your real competitor is not your neighbour. It is the shared calendar.”
When everyone reaches into the same drawer
Why do so few megaprojects come to life?
Because demand is concentrated all at once, like a storm. The mainstream market is slowing while a handful of programs set order books alight: data centres, nuclear and new energy, advanced factories. All enormous. All urgent. All launched at once.
In France alone, data centres are now asking the grid for more than 28 gigawatts of power against connection lead times of two to seven years.
Every program believes it is the exception – the tenth one, the one that gets built. This is optimism bias and planning’s most indestructible flaw: we keep the best-case scenarios and budget the project as we dream it, never as it will behave.
The bias is ancient. Simultaneity turns it into a powder keg. It no longer distorts your number alone. It distorts everyone else’s at the same moment.
The French grid operator’s own figures show it plainly: the data centres connected over the past three years draw, on average, barely a fifth of the power they reserved.
Why? Because these programs all draw from the same source. And the first scarcity is neither capital nor equipment. It is the high-performing professional.
Such people cannot be decreed: they are forged, program after program, and each new project tears them from other projects out of a pool that is not growing.
Then come long-lead equipment and grid capacity – the metronomes of every project. An entire industry is reaching into the same drawer, at the same instant.
The invisible price of shared desire
One phenomenon makes everything worse: insourcing. Operators build integrated teams and drain an already exhausted pool, with each new entrant judging itself more capable than the last.
The same blindness afflicts owners who refuse any outside hand, certain they know how to do what they have never done: optimism bias applied not to the number, but to oneself.
The pattern is clear: insourcing at scale is inefficient, while externalised delivery costs less to operate and delivers faster. Risk rises exactly where it was thought to be under control.
This is the simultaneity premium: the extra cost a program pays not for what it is, but for what the others are doing. No bill of quantities mentions it. It is the invisible price of shared desire.
Your budget can be beaten without a single line failing because 10 others want the same thing you do, at the same time you do. And it does worse than inflate: it blocks. It leaves nine dreams out of 10 by the roadside. Your real competitor is not your neighbour. It is the shared calendar.
The monument budget is dead
For most of my career, the cost question fitted on one line: what will this asset cost to build?
The hard part is no longer pricing the work. It is holding the number while the world moves beneath our feet. These programs are not large buildings. They are small economies, delivered against the clock.
Where do failed budgets come from? From ratio-based estimating. So many millions per megawatt, so much per square metre, a coefficient for risk.
It is fast. It is reassuring. It has the elegance of a spreadsheet and a short memory.
A ratio does not price your project. It prices the past – the average of yesterday’s projects, built elsewhere, in a market that no longer exists. Pricing a multi-billion program by ratio is navigating with the map of another ocean.
Estimating well is the exact opposite: pricing in the present. Start from the program’s basis of design. Take it component by component, at today’s prices, in today’s market, with assumptions exposed and dated. This work costs more and flatters no one.
But it alone can say, when the world shakes, what moved, by how much, and what remains to be decided. The ratio records. The baseline steers.
And here is the heresy I gladly defend: this baseline does not wait for engineering. It precedes it. We have long treated cost as the consequence of design. We draw, we price, we act surprised. On a megaprogram, that order is a machine for overruns.
The right order is the reverse: cost, set early on the basis of design, becomes the frame within which engineering optimises. Cost is not the outcome of design. It is its pilot.
Let me say it plainly: the lowest estimate wins the tender and loses the program. I have seen admirable submissions founder because their numbers had never been their own.
And the opposite: teams ordering critical equipment months before their permit, because their baseline told them what six months of waiting would cost. You do not hold a budget by defending it. You hold it by steering it.
“Pricing a multi-billion program by ratio is navigating with the map of another ocean.”
| Five mistakes that kill a megaproject budget |
| 1. Pricing by ratio. So much per megawatt or per square metre. You are pricing other people’s past, not your own project. |
| 2. Confusing precision with reliability. A number with three decimals resting on soft assumptions is still a soft number. Reliability comes from exposed, dated assumptions, not from decimals. |
| 3. The flat contingency. A uniform percentage ‘for risk’ is not risk management: it is an admission that the risks were never priced one by one. |
| 4. Pricing tomorrow’s purchases at yesterday’s prices. An estimate is valid at its date; commitment will happen in another market. Without escalation and the simultaneity premium, the gap is already written. |
| 5. Freezing the budget on decision day. The monument budget: defended, never recalibrated, obsolete within six months. You do not hold a budget by defending it. You hold it by steering it. |
“Cost is not the outcome of design. It is its pilot.”
The discipline of the winners
The tools are known: rolling forecasts, priced scenarios, early procurement. But the decisive discipline sits one level higher: measuring the simultaneity premium itself.
No longer “what does my project cost?”, but “what does what the others are doing cost me?”.
Within the International Cost Management Standard (ICMS) Coalition, where I served as Secretary General, the conviction is everywhere the same: you can only arbitrate between programs measured in the same language. To compare is already to steer.
France offers a rare laboratory: a nuclear revival, data centres competing for the same grid capacity, major public programs all at once.
The lessons carry far beyond: to Australia’s parallel transport and energy pipelines, to the Gulf’s giga-projects, to a United States discovering contention on a continental scale. The geographies differ. The drawer is the same.
Racing to build
This race is a rare moment: a generation laying the foundations on which tomorrow’s economies will run. Rarely have builders held so much consequence in their hands.
But the chance must be earned. Ambition priced honestly and steered continuously becomes achievable. Ambition resting on a frozen number becomes a gamble. And at this scale, gambles are paid in hundreds of millions and in lost confidence.
The first scarcity is human, and so is the first strategy: attracting, developing and keeping high-performing people is not a human resources policy. In this race, it is an organisation’s first asset.
The tenth project – the one that leaves the ground – will not be the one that chased the lowest estimate. It will be the one on which the team learned to read, and to price fairly, the simultaneity premium.
Price it at the start. Keep it alive as the world moves. Then the race stops being an ordeal to survive. It becomes what our era asks of its builders: that the paper cathedrals, at last, become stone.
FURTHER INFORMATION:
