A project budget is credible only when it reflects defined work, verified assumptions and clear financial authority. Cost management therefore starts before contracts are signed or expenses are booked: the team must decide what costs belong to the project, how to estimate them, how much uncertainty to allow and how later changes will be approved.

This article covers the foundation of a project cost system rather than advanced forecasting. It explains estimating, the time-phased cost baseline, contingency and management reserves, actual spending and reporting. More advanced remaining-cost and supplier scenarios are covered in project cost control.

Define what the cost-management plan controls

The plan should specify estimate units, currency, labor-rate assumptions, purchase categories, precision, approval levels, reporting period and variance thresholds. An organization must identify who owns each work-package estimate and who can authorize a revision. A plan without these decisions can accumulate thousands of line items yet provide little real control.

Consider a regional hospital replacing its appointment system. Implementation requires configuration, integration, testing, data migration and user training. Future subscription fees and normal support may belong in a larger business case, but they should not be casually combined with the approved implementation-cost baseline. A sponsor needs to know which decision a particular total represents.

Map expenses to a work breakdown structure and responsible control accounts. Accepted deliverables, not departmental labels alone, should explain why a resource is needed. Establish rules for recording employee hours, accrued supplier work, purchase orders and payments. Accounting timing can differ from delivery progress, and the cost plan should explain those differences from the beginning.

Select an estimate appropriate to the evidence

Analogous estimating uses data from comparable work; parametric estimating multiplies a credible unit quantity by a suitable rate; bottom-up estimating adds detailed resources across work packages. Three-point estimation can describe a range when uncertainty is significant. The amount of effort spent on estimation should reflect project maturity and the consequences of being wrong.

For the appointment project, suppose implementation labor is estimated at $120,000, integration software and licenses at $40,000, and infrastructure plus testing at $30,000. The planned work estimate is $190,000 before risk provisions. These numbers are illustrative: their reliability depends on labor hours, vendor quotation boundaries, interface conditions and the amount of actual testing.

Record the basis of each estimate. A vendor quote excluding security validation should not be treated as the completed delivery price. An hourly rate from an inexpensive region may not apply to a scarce specialist. The different project cost estimating methods help a reader understand why an early rough order of magnitude cannot be compared with a mature signed-off budget as though both had identical accuracy.

Ask what might invalidate the estimate: a failed interface, unavailable test data, procurement lead times or an unapproved reporting feature. A numerical total without those assumptions can create false certainty. Use an estimate range where information is immature, then narrow it as scope and quotations become better defined.

Distinguish contingency and management reserves

Contingency supports identified uncertainties within the approved scope. Management reserve supports unforeseen work under separate authority rules. A common PMI-aligned budgeting approach includes authorized contingency within the time-phased cost baseline but excludes management reserve from that baseline; the reserve still contributes to the overall project budget.

If the planned work costs $190,000 and identified integration risks justify $20,000 contingency, the cost baseline is $210,000. If the sponsor also holds $15,000 in management reserve, the total project budget becomes $225,000. No extra earned work has occurred merely because reserve money exists, and the reserve cannot be silently spent without the prescribed authorization.

Identify the underlying risks rather than adding an unexamined ten percent. A possible data-conversion failure could need extra specialists and test cycles. The project risk management process connects probability, consequence, response and the remaining provision. Not every uncertainty deserves the same reserve treatment.

Build a time-phased baseline from real milestones

For a plant-control upgrade, a $600,000 equipment purchase order should not automatically appear as six evenly spaced monthly costs. The agreement might require a deposit at design approval, a factory-acceptance payment and a final amount after commissioning. A credible time-phased baseline places expected expenditure against those contractual events and connects acceptance criteria to the payment schedule. Otherwise, a dashboard can show an artificial spending variance that reflects billing timing rather than actual delivery problems.

Keep the authorized budget distinct from cash-flow forecasts. An advance payment may use cash before the equivalent work has earned value, while an accepted deliverable may have an invoice that arrives later. The sponsor needs a coherent explanation of both, not one misleading curve presented as every financial measure at once.

A project may have a $210,000 cost baseline but incur very little spending during requirements approval and much more during equipment procurement or integration. Distribute the budget according to actual planned work and acceptance dates, not by dividing it evenly across calendar months. The resulting baseline enables planned-value comparisons and realistic funding requests.

Cash payments may occur before or after work is delivered. A supplier deposit can be due well before its accepted integration milestone, while employee costs can accrue before payroll settlement. Show planned earned work, committed purchases and cash obligations separately so the steering group can spot a potential funding shortfall rather than a misleading performance variance.

Preserve the original approved baseline and its later authorized revisions. Changing last month’s figures retroactively may erase the evidence of slippage. A change should identify scope, reason, financial effect, schedule effect and authority. The report must distinguish the baseline being measured from a newly forecast completion cost.

Reconcile invoices, commitments and accomplished work

Use three views in review: money already paid, commitments created by approved purchase orders, and the cost forecast of the remaining required work. A project may appear comfortably under budget in the accounting ledger while large supplier commitments are not yet invoiced. Conversely, an unusually early invoice does not establish that more work was accomplished. Compare all three against accepted milestones and the authorized baseline, and explain differences before recommending a change.

An actual cost ledger records incurred spending; a cash ledger records payments; procurement records committed obligations. These are related but different. If an order is signed today and delivery is due next quarter, the business has a commitment but may not yet have a corresponding accepted deliverable or accrued invoice.

Earned value (EV) records the budgeted value of accomplished work, not the amount already spent. Planned value (PV) records budgeted scheduled work and actual cost (AC) records expenditure. An earned-value performance review can compare them when the scope is measurable. Spending 70% of the budget is not proof that 70% of the deliverables have been accepted.

Reconcile each material variance against the relevant work package and reporting cut-off. A late invoice can make a project look under budget; an early payment can distort the opposite direction. The project manager and finance lead should disclose significant recognition differences rather than adjust completed-work percentages to improve a dashboard.

Control scope and funding changes

Suppose a hospital sponsor requests an additional analytics interface after design approval. Estimate the new integration, testing, training and ongoing operational consequences. The relevant authority can approve, defer, reject or redefine the request. Until approval, it should not be hidden inside the original cost baseline and described as a performance problem.

When management reserve is approved for a legitimate change, the organization may move the authorized portion into the revised cost baseline using its prescribed procedure. Preserve the previous record. A cost forecast does not automatically change the authorization to spend, and a team should not treat an updated spreadsheet as a substitute for approval.

Schedule changes also affect money. Overtime may protect a release date but increase cost; a delayed vendor decision may extend contractor assignments. An apparent cost-saving reduction in testing can introduce safety, security and customer-acceptance risks. Evaluate the complete delivery consequence rather than optimizing one financial number in isolation.

Report cost health with clear decisions

At each review, present original authorization, current approved baseline, actual costs, committed obligations, remaining estimate, available reserves and material financial risks. Include accepted outputs and likely business value so a green financial report does not conceal an unusable system. Name the specific decision the sponsor is being asked to make.

Variance thresholds should reflect scale and consequence. A $5,000 variance in a $20,000 work package can matter more than the same amount in a multimillion-dollar program. A serious security or compliance issue may require escalation before any financial threshold is crossed. Explain the root cause and the owner of the corrective action.

At closeout, reconcile accruals, contracts, released reserves, final delivered work and operational handover. Compare final figures with both the original and approved revised baseline, then record which estimating assumptions failed and which controls helped. Cost management is most valuable when it makes spending decisions transparent enough to improve the next project.