Your WIP schedule is a personnel report that nobody reads as one. Margin fade, revised cost-to-complete and a growing underbilled column are accounting facts first, but they cluster around individual people and individual seats. Read by job and by who runs it, the WIP schedule usually names the leadership gap months before anyone in the business does.
The construction CPA sees it first, and almost never says it out loud. A quarterly WIP schedule review turns up three jobs whose forecast margin has moved the wrong way, an underbilled column that has doubled since the last statement, and cost-to-complete numbers that were clearly revised at the last possible moment. The conversation that follows is about accounting treatment. It is almost never about who is running those three jobs.
That is the gap. The people who can read a WIP schedule properly generally do not run hiring, and the people who run hiring generally cannot read a WIP schedule. Both halves of the answer exist inside the same business, in different rooms.
What is a WIP schedule actually reporting?
Contract revenue recognized to date against costs incurred to date, job by job, under percentage-of-completion. Each row carries the contract value, costs to date, estimated cost to complete, percent complete, revenue earned, and the difference between amounts billed and revenue earned. That last difference is the over- or underbilled position.
The schedule exists because construction revenue is earned across periods rather than at a point. It goes to the accountant for the financial statements, to the surety as a core underwriting input, and to the bank. Three sophisticated readers, all of whom are looking at the same rows for financial reasons.
None of them is looking at the rows as a staffing report, or costing what the next departure will run to: the turnover cost sits inside job margin, not on a recruitment line. That is not a criticism of any of them. It is simply outside the question each one was asked. The seat that should be asking it is the finance chair, which is why our note on hiring a CFO treats the WIP schedule as the interview.
What follows is a fourth reading of the same document, using nothing the schedule does not already contain.
Which WIP schedule columns carry the leadership signal?
Three: estimated cost to complete, gross margin variance since bid (fade), and the underbilled column. Each one records a judgment made by a named person on a named job. When the same person’s jobs move together and other jobs do not, the schedule is describing a person rather than a market.
Take them one at a time, because they fail in different ways and point at different seats. Selah runs this as a cost manager search.
Estimated cost to complete is the most forecast-heavy number on the page and the one most exposed to who is producing it. A project manager who walks the job, prices the remaining scope and holds an honest view of what is still coming produces a number that moves in small increments. A project manager who is overloaded, or new, or quietly out of their depth produces a number that holds flat for two quarters and then jumps. The flat quarters are not evidence that the job was under control. They are evidence that nobody re-examined it.
This is a structural feature of the work rather than a character defect. The commercial function sits where the contracts, valuations and reporting live; the execution function sits where the work is. Ask a site team what is still outstanding that has to go into the forecast and the honest answer is often that they do not know yet, because the thing that will cost the money has not announced itself. So the number gets built from what was written down, and the part that would have changed it stays on site.
Margin fade is the gap between the margin a job was bid at and the margin it is now forecast to earn. One job fading is a job. Several jobs fading in the same sector, at the same delivery method, at similar contract values, is a pattern, and patterns have owners. Regional context matters here too: a Texas contractor watching fade appear only on its Dallas and Houston work, while the same delivery method holds in the Southeast, is looking at a market question and a staffing question at the same time, and the schedule can separate them. Fade concentrated in hard-bid work while negotiated work holds usually points at estimating rather than at project management, which matters because those are different hires. A hard-bid estimator and a negotiated-work estimator are not interchangeable people, and neither is the search that finds them.
The underbilled column is the one that most reliably indicates something operational rather than commercial. Costs incurred that have not been billed can reflect legitimate timing. They can also reflect change orders performed on verbal instruction and never priced, a schedule of values that has not been updated since buyout, or billing deadlines missed while the person responsible dealt with something on site. Construction runs on verbal instruction: someone says proceed, the work happens, and the paperwork catches up weeks later or never. The risk never sits with the person who gave the instruction. It falls on whoever has to prove, months later, what was agreed, and that proof either exists in the billing or it does not.
Read together across four quarters, those three columns produce something closer to a staffing map than a financial statement.
How do you tell a hiring problem from a market problem?
Sort the schedule by who runs each job, not by job number or by size. A market problem moves most rows in the same direction at once. A hiring problem moves one person’s rows while comparable jobs under someone else hold their position.
That single re-sort is the whole diagnostic, and almost nobody performs it, because the schedule is built for a reader who has no reason to care who runs job 2417.
Take the Texas contractor above. Sorted by job number, the schedule shows four fading jobs and a market story. Sorted by who runs them, three of the four sit under one project manager who absorbed a second portfolio in the same quarter the fade began, and the fourth is a genuinely difficult job. Same rows, same numbers, two entirely different conclusions.
Sorted that way, four patterns show up repeatedly, and each points at a different seat. The project accountant is usually the person who could have flagged them first.
| What the WIP schedule shows | The usual accounting explanation | The staffing reading, when it recurs |
|---|---|---|
| Fade concentrated in hard-bid work, negotiated work holding | Competitive bidding pressure in the market | The estimating seat is priced for negotiated work and is carrying hard bid it cannot support |
| Cost-to-complete flat for two quarters, then a large revision | Late discovery of a scope issue on one job | Nobody is re-forecasting monthly; the project management seat is overloaded or vacant in practice |
| Underbilled column growing across several jobs | Timing differences and billing cycles | Change orders unpriced and the schedule of values stale; the cost control or commercial seat is missing |
| Margin holding on paper, cash conversion slowing | Owner payment behavior on those contracts | Project controls is reporting rather than managing; the schedule and the cost position have separated |
The right-hand column is not a conclusion. It is a question worth asking before treating any of these as a pure accounting matter, and it is a cheaper question than the alternative, which is discovering the same thing at renewal or at year end.
Two further tests separate the readings quickly:
- The timing test. Line up the quarter each trend started against the quarter a role went vacant, changed hands, or absorbed a second portfolio. Trends that start within a quarter of a personnel change are rarely coincidental.
- The comparable test. Find two jobs of similar value, sector and delivery method under different leadership. If one fades and one does not, the market explanation is already weak.
Neither test requires anything the business does not already have on file.
Why does this reach a search firm at all?
Because the fix for most of these patterns is a person, and because the ones that cannot be advertised are the ones that reach us. Selah Talent Partners runs confidential executive search for mandates that cannot be publicly posted, which is exactly the shape of a search triggered by a WIP review.
The reason is structural. A contractor who has just concluded, from four quarters of data, that a seated project manager is not managing the forecast cannot post that role. The posting reaches the person’s own team, their subs, and the owners on the jobs they run. Neither can a contractor post a chief estimator role while the incumbent is preparing a bid due in six weeks. The document that produced the diagnosis is the same document a surety and a bank are reading, which sets the timetable and rules out the loud version of the search.
This work is not for a contractor whose WIP schedule shows a single bad job. One job is a job. It is not for a business that needs a bookkeeper, a controller or a CFO: those are finance hires, and the seats discussed here are preconstruction, project management, cost control and project controls. And it is not for anyone hoping a search firm will confirm a conclusion they have already reached without doing the re-sort first. The diagnosis has to hold before the hire is worth making.
Where it does apply, the schedule has done something valuable. It has converted a suspicion into four quarters of dated, auditable evidence about a specific seat, which is a considerably better brief than most searches begin with. A brief built on that evidence names what the person will actually own: the forecast, the change-order position, the buyout, the bid.
If a WIP review has raised a question you would rather not put in a job posting, that is the conversation to have. Confidentially, and before the next quarter closes.
Sources and further reading
- AICPA guidance on revenue recognition and construction contractor accounting
- CFMA on WIP reporting practice and construction financial management
- Census data monthly value of construction put in place
- BLS managers Occupational Employment and Wage Statistics, Construction Managers
- BLS estimators Occupational Employment and Wage Statistics, Cost Estimators
- AGC survey 2025 AGC and NCCER workforce survey, national results