Change Orders and Earned Value: How MBook Tracks Variation Impact on Project Billing

Ask a project director where they lose margin and the honest answer is rarely the original tender. It is the variations — the change orders agreed on site, in a meeting, over email — that expand scope and cost without immediately expanding the billing position. The work happens; the invoice does not catch up; and by the time the commercial team reconciles it at month-end, the cash-flow gap is already real.

This is the EPC and construction version of a stale number. The approved contract value on the system says one thing; the scope actually being built says another; and the gap between them is unbilled work and unrecovered cost.

It is not a reporting problem. It is that the variation never flowed through to the numbers that drive billing — so the project is invoicing yesterday's contract while building today's.

A variation order (or change order) is a formal change to the agreed scope — additional work, a design change, a quantity adjustment, a rate revision. When it is approved, several numbers should move at once: the revised contract value goes up (or down), the budget for the affected work packages changes, the forecast cost-to-complete changes, and therefore the percentage complete and the earned value change.

In a lot of mid-market EPC operations, only one of those moves promptly — usually the one someone remembers to type into a spreadsheet. The revised contract value gets updated for the client-facing summary, but the earned-value and billing mechanics still run off the original budget. Or the cost gets booked but the revenue recognition does not follow, and the project looks less profitable than it is.

The discipline that keeps a project honest is that a variation is a single event that updates every dependent number, not a note that updates one and leaves the rest to a month-end reconciliation.

Earned value management is the standard way to answer "how is this project really doing" in a single, comparable frame. Three numbers: planned value (the budgeted cost of the work scheduled), earned value (the budgeted cost of the work actually done), and actual cost (what that work actually cost). From those you get schedule variance and cost variance — whether you are ahead or behind, over or under.

The reason variations matter to earned value is the word budgeted. Earned value is measured against the budget. Change the budget with a variation and you change the baseline that earned value is calculated against. If the variation never updates the budget, the earned-value position is being measured against a scope that no longer exists — and every derived metric, including the billing position, is wrong in the same direction.

This is why variation handling is not a commercial afterthought. It is upstream of every project-health number the director relies on.

The correct sequence is short and should be automatic. A variation is raised and approved. The revised contract value and the affected work-package budgets update. Earned value recalculates against the new baseline. The billing position — what can now be claimed based on progress against the revised scope — updates with it. The next application for payment reflects the scope as agreed.

When that flow is manual, each hop is a place the number can stall. The variation is approved but the budget is not updated for a week; earned value runs off the old baseline; the billing application under-claims; and the cash arrives late for work already done. Multiply across a portfolio of projects and the working-capital impact is material.

Speed becomes structural when the flow is connected — the billing position is current because the variation updated it, not because someone remembered to reconcile it before the deadline.

MBook is built for the EPC and construction operating model, where the project — not the finished good — is the unit that carries cost, revenue and margin. A variation is recorded against the project as a first-class event: it updates the revised contract value, adjusts the work-package budgets, and flows through the earned-value calculation into the billing position.

Because those numbers are connected rather than maintained in parallel spreadsheets, the project director and the commercial team are reading the same current figures — approved contract value, cost-to-complete, earned value, and what can be billed — reconciled to the same source. The month-end reconciliation stops being the moment the truth is discovered and becomes a confirmation of what the system already showed.

None of this is exotic project-controls theory. It is the basic discipline of treating a variation as one event that updates every dependent number — delivered at a scale and price a mid-market EPC contractor can adopt without a tier-one project-controls suite.

If your variations are agreed on site and reconciled at month-end, the gap between scope-built and scope-billed is where your margin and your cash flow are leaking. The fix is to make the variation a single event that updates the revised contract value, the budget, the earned value and the billing position together.

The metric that tells you it is working is billing lag: the time between a variation being approved and the billing position reflecting it. When that lag collapses from weeks to the same day, unbilled work stops accumulating and the cash for work already done arrives on the next application, not the one after.

You are not funding a project-controls science experiment. You are closing the gap between the contract you are building to and the contract you are invoicing against.