Cost to Complete

A contractor has spent much of a project budget but still has several deliverables unfinished, so the original budget is not a reliable answer to what remains and cost to complete asks for a fresh estimate of resources needed for the outstanding work. Oracle describes estimate to complete (ETC) as the estimated cost of finishing an activity, with different calculation methods available in its project system, and PMI distinguishes ETC, the remaining expenditure, from estimate at completion (EAC), the forecast total.

These are related but not interchangeable. Start with a dated scope of remaining tasks, because a percentage of work completed is not enough if the expensive tasks are still ahead.

Estimate labour hours, materials, subcontracting, testing, delivery and closeout work still needed, and include committed costs not yet recorded as actuals under the project's accounting convention without counting the same purchase twice, reviewing open purchase orders, accruals and invoices to reconcile the cost-to-date number. Use current prices rather than simply repeating the rates in the original bid if those rates have changed.

As an illustration, actual cost to date is $1.2 million and estimated remaining cost is $0.8 million, so forecast total cost is $2.0 million. If the approved budget is $1.8 million, the forecast exceeds it by $0.2 million, but a forecast overrun is not automatically an authorised budget increase.

Separate remaining cost from remaining budget, since budget left is an authorisation limit, not proof the work can be finished for that amount. A manual bottom-up estimate may be best when scope or productivity has changed materially, while a formula based on historical cost performance may be useful for another situation, and the assumption behind the method should be stated because a continuation of current inefficiency leads to a different estimate from a plan that fixes it.

Earned value methods can combine completed work and cost trends to estimate the remainder, but their results depend on reliable baselines. Completion percentage can be misleading if measured by spending, since spending 70% of the budget does not mean 70% of the work is done.

Track approved changes so that new scope is shown separately from an overrun on the original scope, and document excluded or uncertain work because hidden commissioning and handover tasks can make a forecast falsely optimistic. Ask suppliers for updated quotes where major packages remain, since a stale quote is not a current commitment, and include contingency based on identified uncertainty without burying a known cost in a vague reserve.

A risk register can record threats to the estimate and the response owner but should not substitute for costing the base work, and work that fails acceptance criteria must be redone, so rework belongs in the remaining cost. Reforecast at key milestones or when a material scope, price or schedule change occurs, and keep the estimate date visible, since a forecast made months ago may no longer guide a decision.

Compare ETC with available cash and contractual payment timing, as a project can be profitable on paper but need funding before final payment, and coordinate time and cost forecasts because a schedule delay can raise staff and site costs even if materials are unchanged. For a portfolio, add component remaining costs only after resolving shared resources and inter-project dependencies, avoid using a single percentage for all work packages when their risk and cost mix differ, and explain what changed between forecasts (quantity, price, productivity, scope or contingency), because the result supports an early decision to adjust scope, resources or funding but is not an accounting promise that the exact final cost is known.