← Insights

Projects

Project budgets that forecast, not just record

July 2026 · 6 min read

Most project budgets are bookkeeping wearing a spreadsheet. A column of planned figures, a column of what has been spent, a total at the bottom, and a variance that goes red only once the money is already gone. It is an accurate record of the past and almost useless for the one thing a budget is for: telling you, while you can still act, that you are heading for an overspend.

The gap is not more detail. It is one missing column and a shift in what "spent" means.

The column everyone leaves out: committed

Track a project budget on planned-versus-actual alone and you are blind to the most dangerous money on the project - the money you have promised but not yet paid. You raise a purchase order for fifty thousand of equipment. The invoice will not land for two months. On a planned-versus-actual sheet, that line still reads healthy, because "actual" is zero. You have already spent the money; the spreadsheet just does not know it yet.

Add a committed column and the picture becomes honest. Three numbers per line, each answering a different question:

• Planned - what you budgeted for this line.
• Committed - what you have obligated through orders and contracts but not yet been invoiced for.
• Actual - what has genuinely been paid.

Committed is the early-warning column. It moves the moment you sign, not the moment finance processes the invoice, so it tells you about trouble weeks before actual would.

Variance done properly

Once you hold all three, variance can stop being a rear-view mirror. The useful definition is commitment-aware:

Variance = Planned - Actual - Committed.

Now a line you have over-committed on turns red now, while you still have options - re-scope, negotiate, move budget from a line that came in under. Wait for actual-only variance and the same red arrives after the invoice, when the only option left is to explain it. Same money, months of difference in when you find out.

Forecast at completion

The question a sponsor really asks is not "what have we spent?" but "where will this land?". Answer it on every line with a forecast at completion - a running estimate of the final cost. A sound, simple rule: take the higher of the planned figure or the sum of what is already actual plus what is committed. That way a line tracking under plan forecasts to its budget, and a line already blowing through it forecasts to the honest higher number rather than optimistically snapping back to plan.

Roll those line forecasts up and you have the single most valuable figure a project budget can produce: a live estimate of the total final cost, updating every time you enter a commitment. That is the number that lets you go to a steering group with "we will land four per cent over unless we act" instead of "we are on budget so far" - two sentences that age very differently.

Make overspend impossible to miss

All of this is worth nothing if it hides in a wall of figures. Two finishing touches:

• A red flag over 100 per cent. Any line where consumption passes its budget should flag itself, automatically, the instant it does.
• A rollup you can read. Total planned, committed, actual and variance; overall per cent consumed; the count of lines over budget; and spend by category, so you can see where the money is going, not just that it is going.

A budget built this way stops being a monthly reconciliation chore and becomes a forward instrument - the thing that tells you about the overspend while it is still a decision rather than an apology.