Overrun discovered only at close—hours, purchases, and expenses must be visible same-day under one project ID

On Friday they said the budget was enough, but on Monday the finance department said it had already been exceeded.

The team meeting was Friday afternoon. The project manager pointed at the chart: 'We've spent 62%, the progress is about the same, no big issues.' On Monday, finance will reconcile payables and inventory, plus two weeks of unreported hours and one in-transit subcontract, and the EAC will directly exceed the budget. No one suddenly 'overspends' over the weekend — overspending happens every day; it's just that hours are on personal timesheets, purchases are in departmental expenses, travel is reconciled at the end of the month, subcontract progress payments are on a separate sheet. Each of the four streams flows in its own way, and the project manager goes by feeling.

The primary goal of project cost management isKnow the deviation on the same day, it's not about totaling the accounts when the project closes. The project management plan consolidates the estimates, budget, accounting, final accounts, and material and labor costs into the same project ledger precisely to ensure that labor is recorded as work is reported, procurement is entered as commitments, and expenses go into their respective accounts. Weekly meetings review the EAC, rather than waiting for project closure to hear the verdict. This is especially true for fixed-price contracts: contract revenue is locked, and any daily cost deviation eats into gross profit.

If the three sets of accounts don't reconcile, any budget percentage is an illusion.

A deeper layer: the budget structure is not aligned with the WBS. The budget is divided by major categories, while execution is assigned by tasks; if they don't match, task-level consumption cannot be tracked. Without change control, if the scope increases but the budget remains the same, the EAC will inevitably explode—cost issues are often delayed bills for scope problems. When resources are overloaded, both overtime and rework will simultaneously drive up labor costs; the cost dashboard and the load dashboard should be able to indicate each other, rather than reporting their indicators separately.

Evaluation criteria: Select any ongoing project and determine whether it is possible to provide, on the same day, the occurred costs, in-progress commitments, ETC, EAC, and deviations relative to the budget. If the "same-day version" cannot be provided, the cost board is just decorative.

The completeness rate of collection comes before the detailed rate.

Don't start with the full set of complex earned value calculations. When the collection rate is below an acceptable level, any SPI/CPI is a precise error. The rate model can be rough at first (comprehensive rate by job level); rough but complete is better than precise but incomplete. Priority to connect:

1. Unified Project Number: Purchase requests, orders, reimbursements, and outsourcing contracts are strictly required to be linked to a project; purchase requests without a project number are approaching zero.
2. Scheduled Work ReportingOn a daily or weekly basis, approve or reject work hours without task assignment; if there are no work hours for two consecutive weeks but the task is not closed, treat it as a data quality warning first, not a cost issue.
3. In-transit commitmentOrders placed but not yet warehoused are counted into committed costs, and budget consumption must not always lag behind actual obligations.
4. Late invoices use provisional estimates: Accrual entry recorded, offset upon invoicing; if the provisional estimate is not offset for a long time, it indicates a disconnect between the business and the invoice.

When integrating with the financial system, the project side aggregates purchase receipts and payables, but the general ledger still follows financial rules; if there is a discrepancy, first map the project numbers and accounts. Refusing integration but demanding precise project cost makes material and labor costs unable to be reliably aggregated—shadow collection and weekly reconciliation can be done first, but it must be acknowledged that the financial accounts will still lag. If mobile reporting can reduce reporting friction, priority should be given to making it "fillable"; if it cannot be filled, any cost dashboard is just spinning its wheels.

The budget versions must be traceable: adjustments from the rough estimate to the budget, budget change orders, and final account differences should be linkable by project number. Without versions, project closure can only argue about 'what was reported initially,' making it impossible to improve the template for the next period. The account structure should be mapped as much as possible to WBS work packages, or at least to a manageable level; if the mapping is too coarse, overspending can only be attributed to 'a big lump sum,' making it impossible for management to decide whether to cut scope or add personnel.

ETC / EAC: Ask once a week, 'How much more will it cost to finish?'

The Estimate to Complete (ETC) and Estimate at Completion (EAC) can be quite straightforward: update the remaining task person-days × rate, unexecuted procurement/outsourcing, and contingencies for known risks on a weekly basis. EAC = actuals (including provisional estimates) + ETC. When EAC consistently exceeds the budget, scope reduction, change requests for additional budget, or stop-loss measures should be triggered—instead of waiting to be surprised until all invoices have arrived.

Recommended threshold (can be adjusted by industry):

• Budget consumption is noticeably faster than progress → yellow light, check scope and rework;
• EAC over budget → red light, hold cost review;
• Purchasing in progress Spending ≥ Budget → Freeze non-urgent purchases;
• The timeliness rate of work hour reporting is too low → Any cost percentage should be considered unreliable for now.

Only looking at the percentage spent is the most common false sense of security: the table may show 30% remaining, but purchases in transit and unrecorded work hours could have already used it up. During final accounting, the material and labor cost summary should be able to answer whether the overspending is mainly in labor, outsourced purchases, or expenses; if you cannot answer by category, it indicates that the budget structure and accumulation keys are still misaligned. The approved scope increase should simultaneously generate a budget adjustment order; otherwise, the board will penalize 'those executing according to the change.'

ETC updates should not be left for the finance department to fill in at the end of the month. The project manager knows the remaining work best, and the finance team knows the accounting rules best — weekly updates should be led by the former, with the latter verifying the mapping, so that it doesn’t turn into two separate sets of completion estimates talking past each other. If contingency funds never enter ETC, the "potential overruns" in the register will never be reflected in the cost indicators; conversely, when EAC worsens, one should also check whether there are corresponding risk items and mitigation actions recorded. If cost and risk are invisible to each other, both sets of records will just become meaningless paperwork.

Conducting a 'mock surplus' survey on projects under construction is very useful: list projects where budget consumption is lower than progress, but may show a deficit after including work in progress and provisional estimates. These projects are the most likely to be reported as healthy in weekly reports, and the most likely to suddenly explode next month. Making the survey results public is more effective than sending another email saying 'please pay attention to costs'.

Three tables: Fixed Total Price, Budget Learning, and Combination Meeting

For fixed-price projects, it's even more important to watch the EAC: the business may have won the contract, but execution could be losing money. During the estimation phase, use historical similar projects to calibrate person-days and the proportion of outsourced work; after the final account, write deviations back into the next estimation template to form a learning loop. Without this step, the company will repeatedly experience 'post-mortem explosions' on similar projects.

When it comes to time and materials with clients paying by ticket and no internal assessment of gross profit, management requirements are weak and system investment can be postponed; once project profit and loss is assessed, immediately enforce collection discipline. When the work-hour reporting rate is persistently very low, first address discipline and convenient reporting, then discuss the dashboard. When the WBS is unstable and changes are uncontrolled, first implement change admission, otherwise budget items are just for show.

The committee will recommend arranging three tables side by side: Change/Baseline Off-Hours, EAC Variance, and Resource Red Flags — handling 'what to do, how much it will cost, and who will do it' in the same decision. Only after aligning with the contract amount and purchase warehousing will the project profitability be eligible for operational review, instead of being merely influenced by emotions during project closure retrospectives. If outsourced progress payments are still in the project manager's personal spreadsheets, the EAC will systematically be underestimated: subcontract agreements must be entered under the same project number, with provisional estimates confirmed based on progress, to avoid 'separately accounting for outsourcing' that hides the true completion costs.

The costs of JEDP-PMS are calculated down to the work package level, integrating labor hours, procurement, and expenses into a single account, and supporting EAC/ETC alerts. The project account is not the financial general ledger: posting and invoicing still follow the financial system, while the project side is responsible for aggregation and visibility of deviations. Report metrics should come from the WBS, timesheet reporting, and expense aggregation, rather than creating a separate schedule-cost comparison table. Phase II is often integrated with timesheet reporting, budget, and procurement reconciliation; if financial integration is refused but precise project costing is required, first acknowledge the accuracy limit and focus management actions on complete aggregation and weekly EAC updates.

Department expenses exceed the budget: the company can't win fake projects that succeed

Another type of explosion is more hidden: the project schedule shows attractive gross profit, while overruns are recorded under departmental expenses or general accounts. The project manager meets the targets, but the company's profits are eroded. The governance method is not moral condemnation, but enforcing project numbers on purchase requests and reimbursements, putting over-threshold expenses without a project into an exception list, and randomly auditing large expenses without project numbers in combined meetings. Hiding project overruns in departmental expenses makes deliveries look good in the short term, but distorts the pricing model in the long term—similar future projects will be quoted at a loss again based on the wrong historical data.

If the working hours of personnel across projects are allocated all at once at the end of the month, profitable projects will take the blame, and loss-making projects will look better. In the resource plan, first set the available percentage, and during the execution period, fill in according to actual input. The approval chain rejects working hours with 'no task assignment'—this is the same core issue as resource overload management: if the real input can't be seen, neither workload nor cost can be controlled.

This week, combine into one page; fake balances usually fail immediately.

Pick a project that "feels almost finished" or "still seems to have a lot left," and compile this week’s approved work hours, in-transit purchases, unreimbursed expenses, and unadjusted provisional amounts onto a single page, aligning them by the same project number. Most teams will realize on the first merge that "suddenly" isn’t really sudden. Then establish two rules: no project number, no purchase request; overdue work hours don’t get a green light in the weekly meeting. Two weeks later, check whether the EAC can still be read out during the weekly meeting—if it can’t, first review the collection gap list, then discuss the detailed model.

If, after consolidation, it is found that overspending is mainly in a certain category, then what should be modified is the budget template synchronized with the revised budget, rather than just scolding execution waste. If the numbers still don’t match the finance data after consolidation, prioritize mapping project numbers to categories, rather than blaming the 'system inaccuracy' first. Cost can only be controlled when it is visible early; when it is invisible, all 'enhancing cost awareness' is just a slogan. Project closing explosions can be turned into weekly deviation updates—provided that the three sets of accounts are willing to reside under the same project number.

• PMS Project Plan: Full Cycle Cost
• Scope Change Control
• Resource Overload and Load
• Milestone Acceptance and Payment Collection