Sole cost authority on $1.5B+ of capital work

Context

Three consecutive engagements, one seat: a compressor station at $335M, a four-project portfolio at roughly $200M, and a $1B capital program in execution.

In each I was the sole cost, estimating and forecasting authority — not a member of a controls team, the controls function. Project and construction managers made cost decisions on numbers I produced, and nobody above me in that chain checked the work before it reached leadership.

The mandate I gave myself

Project controls is about conveying the right information — and the right information generates uncomfortable questions.

Project managers are not always willing to show upper management what is actually happening, because what is actually happening suggests they are not fully in control. That instinct is human and it is expensive. What can’t be displayed can’t be controlled. A problem that gets attention early is a leak fixed early, at a fraction of what it costs later.

So the job is not to produce comfortable numbers. It is to produce the true one and stay in the room for the conversation it starts.

The decisions

Publishing the number that makes the meeting harder

There are a lot of uncomfortable discussions in this seat. What I have found is that people reach agreement most of the time — provided they are putting the project first rather than themselves. That distinction is the whole game. When the conversation is about the project, an unwelcome forecast is information. When it is about position, the same forecast is an accusation. Part of the work is keeping the room in the first mode.

Stopping being the cost person

The first two engagements look like the same job at different sizes. They are not.

On the first I joined in construction and demobilised at the end — one project, one project manager, one lifecycle stage, a scope that was already defined by the time I arrived.

The second was four projects at four different points in their lives — some in design, others building — with scope nothing like as well defined and a different project manager on each, every one of them working differently. Touch points multiplied and internal meetings multiplied with them. The projects grew, which drew attention, which stacked director meetings, status reviews and engineering steering committee reporting on top of the work. And I carried them from initiation through closeout rather than arriving for construction.

That demanded I stop being the cost person. On one project you can be the specialist who owns a number. Across a portfolio at mixed maturity the constraint stops being analytical rigour and becomes coordination — whether four managers with four styles are held to one standard, and whether the number still means the same thing in every room it enters.

Earning the right to be believed

I have worked most of the seats in this cycle: estimating, procurement, the field as a construction manager, business development, change control, cost engineering. That matters more than it sounds.

It means I understand why the construction manager is asking what he is asking, why the project manager pushes back where he does, and what upper management is actually worried about underneath the question they asked. When you have sat in someone’s chair, you can tell their story back to them accurately — and a person who feels accurately understood will argue with your number instead of with you.

Grounding every model in something other than my opinion

The Estimate at Completion models were adopted — and used by the Director of Engineering for capital allocation, reforecasting and portfolio prioritisation — because none of it rested on judgement alone.

Everything was grounded in research. I hold a Master’s in civil engineering and I have spent a long time learning where this industry’s data actually lives: published standards, AACE recommended practice, industry-wide benchmarks. That is not academic housekeeping. It changes the register from “I think so” to “here are the facts, and here is the basis of the position.” People come on board for the second one.

Outcomes

EAC models and cost performance dashboards adopted by the Director of Engineering and senior leadership as the basis for capital allocation, reforecasting and portfolio prioritisation — across $1.5B+ of capital value, through three engagements and two orders of magnitude of program size.

What I carry forward

A forecast is a guesstimate. That is not a criticism of forecasting — it is the definition. Forecasts go wrong when they are not grounded in real data, and they go right when the guesswork is disciplined by fact. The whole craft is using the best available information and being honest about which part is measured and which part is judgement.