Pull up the job. Marlow Ridge, 22 townhome units, mechanical contract at $438,000. The cost budget carries $396,400. Actual spend to date reads $162,500. That is 41 percent, and you are eight weeks into a nineteen-week schedule, and it reads like a job that is behaving itself.
Now count what you have already promised.
- 22 condensing units and 22 air handlers, ordered February 9 on a twelve-week lead — $148,600
- Sheet metal package released to the fab shop, roughly 60 percent cut and not yet billed — $34,200
- Line sets, grilles and registers, PO issued — $12,800
- Zoning controls and thermostats — $6,500
That is $202,100 that is approved, ordered, and has a delivery date, and none of it has hit the account. Add it to the $162,500 you have actually spent and the job is $364,600 obligated against a $396,400 budget. Ninety-two percent. You have $31,800 of room left on a job with eleven weeks to run, and the number on your screen says 41.
The number on the screen was true and useless
Forty-one percent was not a lie. Every dollar in that figure left the account. The problem is that it answers a question nobody on a mechanical job is actually asking.
You are not asking how much money has moved. You are asking how much room is left. On a trade where the equipment package is a third of the contract and gets ordered before the slab is poured, those two questions have completely different answers for four months at a stretch.
Cash-basis reporting on a long-lead trade is a rear-view mirror pointed at February.
Here is the shape of the lie. The equipment PO went out February 9. Nothing shows up on a cash-basis budget until the units get delivered and the supplier bills you, which on this job is week fourteen. For thirteen weeks the ledger reports a job with plenty of room. In week fourteen, $148,600 lands in five days and the same ledger reports a job in trouble. Nothing changed about the job. The only thing that changed is when the paperwork caught up.
The decisions you make during the blind stretch
Thirteen weeks of a comfortable-looking budget is not a harmless reporting quirk. It is thirteen weeks of decisions made against a wrong number.
On Marlow Ridge, three of them happened in weeks six and seven:
- The GC asked for eight extra returns in the corner units. You priced them at $340 apiece, felt fine about the budget, and gave it away as goodwill against a job that looked 41 percent spent. That is $2,720.
- The architect changed the ERV location in the stacked units. Two of your best installers spent nine days re-routing before anybody wrote a change order. At an $84 loaded rate across two men, that is $12,096.
- You took a second job starting in week twelve, on the assumption Marlow Ridge would free up cash and men. It did neither.
None of those three were reckless. All three were reasonable readings of a number that was missing $202,100.
What changes when Committed is its own column
Four numbers on every budget line instead of one: Baseline locked at kickoff, Current as the live plan with every change logged, Committed for what is approved but not yet spent, and Actual rolled up from real transactions. The Committed column is the one this trade lives and dies by.
Three things change the day it exists.
The equipment line stops looking empty
A line reading $148,600 Committed and $0 Actual is not an underspent line. It is money with a ship date on it. On a budget that only tracks actuals, that line looks like your best-performing item right up until it becomes your worst. On a four-number ledger it reads exactly as what it is: fully obligated, awaiting delivery, PO dated February 9, expected week fourteen.
Percent obligated becomes the number you manage to
Percent spent tells you about the past. Percent obligated tells you what you can still decide. When the GC asks for eight extra returns and you can see 92 percent obligated on the same screen, that request gets priced as a change order in four minutes instead of absorbed in a handshake. The conversation is not harder. It is just earlier.
The escalation is catchable
Committed also holds the price you were quoted. When the equipment invoice lands at $154,900 against a $148,600 PO, that $6,300 is a variance with a document behind it — a supplier who raised the price between order and ship, on a quote you still have. You either eat it knowingly or you write to the supplier that week. What you do not do is discover it four months later in a lump under Equipment.
The obligation nobody enters: your own crew
Material commitments are the obvious half. The other half is labor you have already scheduled.
Marlow Ridge has 22 units of trim-out booked for weeks fifteen through nineteen, six installers, four and a half weeks. That is roughly 1,080 hours at $84 loaded, or $90,720, and it is as committed as any purchase order. Those men are on your payroll whether or not the units ship on time. When the equipment slid eleven days on the last job, the crew did not disappear for eleven days — they went to a job that did not need six of them yet, and you paid for them twice.
Scheduled labor belongs in the same picture as scheduled steel. A mechanical job costing view that shows committed equipment but treats a fully-booked crew as a future problem is only half honest.
How Committed actually gets filled
The reason most shops do not have this column is not that they disagree with it. It is that filling it by hand is a job nobody has time for, and a Committed number that is three weeks stale is worse than none at all.
So it has to fill itself, off things people already do:
- The purchase request is raised from the truck and approved before the order goes out. Approval is what creates the Committed figure — not a separate entry somebody makes later.
- The subcontract for the fab shop posts its value to Committed the day it is signed, and Committed drops as the shop bills against it, not on a guess.
- The receipt gets photographed at the counter and read into line items, so Actual rises the same day the money moves and Committed comes down by the matching amount.
- Delivery closes the loop. Units on site, packing slip photographed, invoice matched to the PO, variance flagged if the two disagree.
Do that and the ledger is never more than a day behind reality. A line heading toward trouble gets named with its source and its date while there is still a week to do something about it, instead of appearing as a red total after the fact.
Do this before Friday
Take your largest running job. Write down every purchase order you have issued that has not yet been invoiced, every subcontract you have signed that has not yet billed, and the labor hours already committed to the schedule for the next four weeks. Add those to your spent-to-date figure and divide by the budget.
If that percentage is within ten points of the number your system has been showing you, your reporting is close enough to run a business on. If it is fifty points apart — 41 and 92 — then the budget you have been checking every Monday morning has been describing a job you finished ordering for in February.