Construction Budget Forecasting: Why Your Spreadsheet Is Lying to You

The Problem With Monthly Cost Reports

By the time that monthly cost report arrives, the project has already moved on.

The numbers show what you spent three weeks ago. Commitments from last week's purchase orders are missing. Work confirmations approved on Friday haven't been booked. Two variation orders are still pending approval — but the work is already done. Your forecast says you're 4% under budget. Your gut says something is wrong.

Spreadsheet forecasting doesn't fail because of bad intentions. It fails because it relies on manually collected, periodically updated data to track a project that changes every day.

The Three Gaps That Corrupt Your Forecast

1. The Commitment Gap

A construction project spends money in two phases: when it commits (raising a PO, issuing a subcontract instruction) and when it pays (invoice certification, work confirmation). Spreadsheets almost always capture the latter. The former — committed but not yet invoiced — is either estimated or missing entirely.

On a SAR 150M project, the gap between actual costs and committed costs can reach SAR 8–12M at any point mid-project. If your forecast ignores committed costs, you're forecasting on incomplete information — and your cost-to-complete is structurally understated.

2. The Progress Gap

Forecasting cost to complete requires knowing how much work has actually been done — not how much has been billed.

When a subcontractor submits a 40% billing claim, your QS might certify 35% after measurement. That 5% gap matters: it changes your cost-to-complete calculation for every remaining activity in that package. If your forecast uses billing submissions as a proxy for progress, you're overstating completed work and understating the cost ahead.

Work confirmations — formal field measurements of completed quantities — are the only defensible basis for progress tracking. Without them, forecasts are guesses dressed in spreadsheet formatting.

3. The Scope Gap

Change orders expand or contract the budget. But the time between a variation being instructed, priced, negotiated, and formally approved can stretch 8–12 weeks on large GCC projects. During that window, the work may already be underway.

Spreadsheet forecasts either ignore unapproved variations — understating scope and cost — or include them at full value, overstating certainty. Neither is accurate. A proper forecast needs a pending variations register: unapproved variations tracked at their most likely value, flagged by approval status so the PM knows what's at risk.

What a Reliable Construction Forecast Actually Requires

Good forecasting runs on four data streams that must be current, connected, and granular to the WBS:

• Purchase orders — every PO raised creates a committed cost against a cost code. Your forecast's cost-to-complete for materials starts here, not at the invoice.
• Work confirmations — field-measured quantities certified by your QS. This is your progress baseline. Cost of work done = confirmed quantities × unit rates. Cost to complete = remaining quantities × rate.
• Timesheets — daily allocation of labour hours to activities. Without this, labour is the one cost line that can't be tracked in real time. On a labour-intensive MEP fit-out or finishing package, this matters enormously.
• Change orders — pending, approved, and rejected variations tracked by value, status, and the cost code they'll land against.

When these four streams feed the same system, your Estimate at Completion (EAC) is calculated from data — not estimated by a PM under pressure to show green.

How EAC Should Work in Practice

EAC = Actual Cost to Date + Estimate to Complete (ETC)

The ETC is where most forecasts diverge from reality. Two approaches work in construction:

Trend-based ETC: Use your current Cost Performance Index (CPI = Earned Value ÷ Actual Cost). If you've spent SAR 42M to complete SAR 38M of work (CPI = 0.905), your remaining budget will likely underperform at the same rate. ETC = Remaining BAC ÷ CPI. Simple, and often more honest than manually revised estimates from people who are optimistic about recovery.

Bottom-up ETC: Re-estimate remaining quantities and rates for each cost package. More accurate when scope has shifted significantly, but requires QS time and current field data.

Neither approach works if your actual costs are three weeks stale. A CPI calculated on incomplete actuals will mislead you every time.

The Warning Signals a Forecast Should Surface

A forecast that only shows cost vs budget is not doing its job. The signals that matter for early intervention:

• CPI below 0.90 on a major cost package — at-risk margin, requires re-estimation and scope review before the overrun compounds.
• Committed cost exceeding remaining budget — you're already overcommitted before the work is complete. This needs commercial intervention, not a revised spreadsheet.
• Pending variations older than 45 days without resolution — cost accumulating without approved budget. This is where delay claims begin to form.
• Labour productivity trending down — if timesheets show hours increasing faster than certified quantities, field efficiency is declining. ETC needs upward revision.
• Subcontract packages approaching ceiling — if a subcontractor is at 90%+ of their contract sum with 35% of scope remaining, a final account dispute is already forming.

None of these signals surface in a monthly cost spreadsheet. They require real-time data and configured alert thresholds — visible to the PM and commercial manager simultaneously.

Practical Steps for Better Forecasting

Step 1: Commit to WBS-Level Cost Codes from Day One

Every PO, timesheet, and confirmation must land against a specific cost code — not "general conditions" or "miscellaneous." If your cost structure is too broad, your forecasting is too blunt to act on.

Step 2: Close the PO-to-GRN Loop

Every materials PO must have a goods receipt process. Without GRN, you can't distinguish materials delivered from materials ordered. Committed cost without delivery confirmation inflates your forecast and makes three-way matching impossible.

Step 3: Run Confirmations on a Weekly Cycle

Monthly confirmations produce monthly progress data. Weekly confirmations mean you're forecasting on last week's actuals — close enough to intervene before a trend becomes a loss. Most GCC subcontract agreements permit weekly certification; use that right.

Step 4: Build the Pending Variations Register

Track every variation request from the moment it's instructed, not when it's approved. Assign a probability weight — 50%, 75%, 100% — based on commercial status. This gives you a risk-adjusted scope view that honest forecasting requires. Under FIDIC-based contracts, the notice clock is ticking from instruction; your register should reflect that.

Step 5: Review EAC at Package Level in Every Project Review

Stop reviewing cost vs budget in isolation. Start reviewing CPI, ETC, and the gap between committed cost and remaining approved budget. That's where margin is won or lost — and it's where early action makes a difference.

Key Takeaways

• Spreadsheet forecasts are accurate for the moment they were compiled. Construction doesn't pause for month-end.
• Reliable forecasting requires four live data streams: purchase orders, work confirmations, timesheets, and change orders — all connected to the same WBS.
• EAC calculated from real-time data is more useful than EAC estimated by a PM under commercial pressure.
• The signals that predict margin erosion — CPI trends, overcommitment, aging pending variations — only surface when cost data is current and granular.
• On a SAR 200M project, a 2% improvement in forecast accuracy can protect SAR 4M in margin from leakage and delayed commercial response.

Your forecast is only as good as the data behind it. Get the data right first — and structure your review meetings around what it's telling you.