Analyses job costing data for construction and facilities projects. Compares actual costs against budgeted estimates, identifies cost overruns by trade or phase, calculates earned value metrics, and produces variance reports with trend analysis for project managers.
Analyses actual expenditure against budgeted estimates for construction projects and facilities maintenance contracts. The skill processes cost data broken down by trade package, RIBA work stage, or NRM cost element, calculates earned value management metrics, and produces variance reports that highlight overruns, underspends, and forecast final account positions. Output is designed for quantity surveyors, commercial managers, and project directors who need to make informed decisions about cost control and cash flow.
Read the input cost data file containing line items with the following expected fields: cost code or NRM element reference, description, budgeted amount, committed amount (orders placed), actual amount (invoices certified or paid), forecast final cost, and the work package or trade category. Accept CSV or JSON formats.
Validate the input data for completeness and consistency. Check that all monetary values are in GBP, that cost codes follow a recognisable structure (NRM1, NRM2, or the project's bespoke coding system), and that the sum of line items reconciles to the stated contract sum or budget total within a tolerance of 0.5%. Report any discrepancies in a data validation section.
Calculate the following metrics for each cost code and for the project as a whole:
Identify the top five cost overruns by absolute variance and by percentage variance. For each, include the cost code, description, budget, forecast, variance amount, and a brief narrative explaining likely causes based on patterns in the data (e.g., provisional sum adjustments, scope changes, remeasurement differences).
Produce a cash flow comparison showing cumulative budgeted spend versus cumulative actual spend by reporting period. If historical period data is available in the input, generate a month-by-month breakdown. Flag any periods where actual spend exceeds the budgeted S-curve by more than 10%.
Generate a risk-adjusted forecast section that applies a contingency percentage (default 5% unless specified in the input data) to uncommitted budget elements. Present three scenarios: optimistic (contingency not required), expected (partial contingency draw), and pessimistic (full contingency plus a further 3% risk allowance).
Format the output as a structured markdown report with summary tables, detailed line-item breakdowns, and clearly labelled metric calculations. Include a commercial status traffic light (Green/Amber/Red) based on the overall CPI value: Green for CPI above 0.95, Amber for 0.85 to 0.95, and Red for below 0.85.
The report is produced as a structured markdown document:
# Job Costing Analysis Report
## Project Summary
- Project name, contract sum, report date, reporting period
- Overall CPI, SPI, commercial status traffic light
## Data Validation
- Reconciliation check results, any flagged discrepancies
## Cost Summary Table
| Cost Code | Description | Budget | Committed | Actual | Forecast | Variance | Variance % |
## Top Overruns
- Ranked list with narrative commentary
## Earned Value Metrics
| Metric | Value |
- EV, CPI, SPI for each major work package and overall
## Cash Flow Comparison
| Period | Budgeted Cumulative | Actual Cumulative | Variance |
## Forecast Scenarios
| Scenario | Forecast Final Account | Contingency Draw | Net Position |
## Recommendations
- Commercial actions required based on analysis