An unfilled construction seat has a weekly running cost that no ledger records: bids not chased, margin not protected, and the seat’s work absorbed by people who were already full. That cost usually passes the cost of the search itself somewhere around week ten.
There is a line in your accounts that gets better every week the unfilled construction seat stays empty. The salary you are not paying. It is the only number in this entire situation that anyone is tracking, and it is pointing the wrong way.
Everything the vacancy actually costs you lands somewhere else, unlabeled, in a place nobody reconciles back to the empty desk. Which is why firms that would never run a crane without an operator will run a preconstruction group without a leader for five months and call it prudence.
What does an unfilled construction seat cost per week?
Price the work the seat owned, not the salary you saved. An estimating leader running six bids a quarter on jobs averaging US$12 million, at a hit rate of one in five, is worth roughly one won project a quarter. Divide that project’s margin contribution by thirteen weeks for the weekly number.
Run it once on your own book and the argument ends. Take a mid-size general contractor working negotiated and hard-bid work across Texas and the Southeast. The chief estimator seat has been open eleven weeks. In that window the firm passed on four bid invitations because there was no capacity to price them properly, and submitted three more with numbers assembled by a project manager borrowed from operations.
Now count what that costs. Four bids not chased is 4 opportunities gone. On the pass rate above, that is close to a won job. The three that did go out carried a risk premium added by someone who was guessing, which either loses the work or wins it at a number that hurts later.
Meanwhile the saved salary sits in an account somebody reviews monthly, looking like a win. That asymmetry is the whole problem: the saving is visible and the cost is not.
Why does the vacancy feel cheaper than it is?
Because the saving is recorded and the cost is distributed. Salary not paid appears in one account. Bids not chased, margin eroded on jobs running short-handed, and overtime absorbed by the people covering all land in different places, none tagged to the empty seat. Nobody adds them up.
There is a second reason, and it is the more expensive one. The work does not stop when the seat empties. It gets absorbed, and absorption is invisible by design.
Your VP of Operations picks up the two hardest projects. Your senior estimator starts pricing the work the departed leader used to price, on top of pricing their own. For six weeks this looks like the team pulling together, because it is. By week ten it looks like something else: the senior estimator is doing a job they were not promoted into, without the title or the pay, and they have started taking calls.
- The absorbed work has a second bill. The people covering are your best people, because only your best people can cover. They are also the most placeable people you employ.
- Quality degrades before capacity does. A borrowed project manager pricing a bid does not price it badly. They price it defensively, which is a different failure and harder to see.
- The clock is not neutral. In a market where 91.7 percent of the 1,041 contractors answering the salaried-hiring question in the 2025 workforce survey reported difficulty filling positions, a seat that stays open is not waiting in a queue. It is competing.
So the vacancy is not a pause. It is a slow transfer of load onto the people you can least afford to lose, and it compounds.
When does an unfilled construction seat cost more than the search?
Somewhere around week ten to twelve for a leadership seat, on typical construction margins. Before that point the vacancy is genuinely cheaper than a rushed hire. After it, the accumulated weekly cost passes the one-time cost of a search, and every further week is pure loss with no offsetting saving.
The crossover is not a universal constant, so build it from your own inputs: the seat’s margin contribution, your bid calendar, and how much of the work is genuinely absorbable. But the shape holds everywhere, and the shape is what people get wrong.
| Week of vacancy | What is actually happening | Reversible? |
|---|---|---|
| 1 to 4 | Work absorbed, team pulling together, no visible cost | Yes, fully |
| 5 to 10 | Bids passed on, quality of pricing softening, covering staff stretched | Mostly |
| 11 to 20 | Covering staff taking recruiter calls, margin visibly eroding on live jobs | Partly |
| 20 plus | Seat now harder to fill, because candidates read a long vacancy as a warning | Rarely without a premium |
That last row is the one that surprises people. A seat open five months is harder to fill than the same seat open five weeks, because a strong candidate reads a long vacancy as evidence that the role is broken, the pay is wrong, or the person they would report to is the problem. The vacancy becomes its own obstacle.
What should you do differently while the seat is open?
Decide explicitly what you are not doing, rather than letting absorption decide for you. Name the bids you will pass on. Name who is covering and for how long. Put a date on the decision. An unmanaged vacancy is a set of choices being made by default, by whoever is tired that week.
Three moves change the arithmetic materially, and none of them is “hire faster”.
First, price the vacancy out loud, once, in a room with your operations lead and your controller. Not a precise number. A defensible one. The act of stating it converts an invisible cost into a managed one, and the decision that follows tends to be different from the decision that was drifting.
Second, protect the covering staff deliberately. If your senior estimator is carrying the departed leader’s work, either pay them for it now or tell them exactly when it ends. The failure mode here is silence, and the counteroffer conversation that follows is one you will lose.
Third, separate the urgent from the postable. If the seat can be advertised, advertise it and run a proper process. If it cannot, because the incumbent is still in the chair or because signaling trouble in preconstruction before a bid would cost you the negotiation, then you are in confidential executive search for mandates that cannot be publicly posted, and the search needs to start that way rather than after a posted attempt fails.
The firms that handle vacancies well are not faster hirers. They are firms that stopped treating an open seat as a neutral state, which is the only reframe here that matters.
Put a number on your own open seat this week. If the number surprises you, our approach to construction search starts from exactly that arithmetic rather than from a fee.
Sources and further reading
- AGC of America and NCCER, 2025 Workforce Survey, national results: AGC survey PDF
- US Bureau of Labor Statistics, Occupational Employment and Wage Statistics, Cost Estimators 13-1051: BLS OEWS
- US Census Bureau, Value of Construction Put in Place: Census construction
- AGC of America, construction data and industry surveys: AGC data