Onboarding a construction leader is the employer’s job, not the hire’s. In the first hundred days the firm owes them information, access, a defined decision boundary, and one honest conversation about what they have inherited. Skip those and a good hire fails for reasons that have nothing to do with capability.
A contractor in Nashville spends four months and a placement fee finding a preconstruction leader, then hands them a live GMP submission on day three because the seat has been empty and the bid calendar did not pause.
That is not onboarding. That is a stress test conducted on someone who has not yet learned where the estimating history is kept, which subs are reliable in this market, or who actually decides what the contingency carries. The hire may survive it. If they do not, the search will be blamed.
What does onboarding a construction leader actually require?
Four things from the employer: full information about what they have inherited, access to the people and systems they need, a written decision boundary, and one honest conversation about the problems nobody mentioned during interviews. The new leader supplies the work. The firm supplies the conditions.
The information obligation is the one most firms discharge badly, and usually with good intentions. Nobody wants to open with the department’s problems. So the new preconstruction manager learns in week five that two senior estimators applied for their job, that the estimating software migration stalled eighteen months ago, and that the president reviews every number over $10M despite what the org chart says.
None of those would have stopped a good candidate accepting. All of them, discovered late, cost trust that was cheap to keep and expensive to rebuild.
Access is the second, and it is more concrete than it sounds. A construction leader needs the last two years of bid history and hit rates, the current backlog with its margins, the WIP schedule, and unmediated conversations with the people who report to them. A leader arriving into a Houston or Dallas operation needs the local subcontractor history too, because a market read does not transfer between metros. Firms that stage this over six weeks are protecting nothing and delaying everything.
What should happen in each of the first hundred days?
Split it into three phases: learn, then diagnose, then act. Roughly thirty days understanding the system, thirty forming a view, and forty executing on it. A leader pushed to act in week one is guessing, and their first guess sets the tone for everything after.
The first thirty days belong to information. The new leader should be reading the last two years of estimates or project reports, sitting in on meetings without owning them, and having a conversation with every direct report and every peer. In construction that includes people outside the org chart: the two superintendents who have been there twenty years and decide more than their titles suggest, and the owner’s rep on the largest live job.
The middle thirty are for diagnosis. This is where a genuinely good hire earns the fee, because they will see things the firm has stopped seeing. Expect to be told something uncomfortable somewhere around day forty-five, and expect it to be at least partly right.
The last forty are execution, and the firm’s job here is to get out of the way while backing the decisions it agreed to. The most common failure at this stage is a president who hired a leader to fix something and then relitigates each decision, which teaches the whole department that the new leader does not actually have the seat.
| Phase | Days | The leader’s job | The firm’s job |
|---|---|---|---|
| Learn | 1 to 30 | Read history, meet everyone, own nothing | Give full access, withhold nothing |
| Diagnose | 31 to 60 | Form a view and say it out loud | Listen without defending |
| Act | 61 to 100 | Execute the agreed changes | Back the decisions publicly |
Why does a good construction leader fail in the first year?
A construction leader rarely fails for capability. It is usually one of three reasons: an inherited problem nobody disclosed, a decision boundary that was never defined, or a predecessor comparison the firm refused to name. All three are the employer’s to prevent.
The undisclosed inheritance is the most common. Someone takes a role, finds a problem in month two that materially changes the job they agreed to, and starts taking calls in month six. By the time they leave in month nine, the guarantee window on any search has long closed and the firm concludes the hire was wrong. Selah runs this as a preconstruction manager search.
The undefined decision boundary is the most fixable and the most neglected. A chief estimator who does not know whether their number goes to the owner without the president reviewing it will either overstep and be corrected publicly, or underclaim and be judged passive. Neither is a capability problem. Both are a one-paragraph problem the firm never wrote down.
The predecessor comparison is the most human. Replacing someone well liked, particularly someone who was there a long time, means the new leader is measured against a memory that improves with time. A firm that names this openly, to the team and to the hire, converts it into an ordinary transition. A firm that pretends it is not happening leaves the new leader fighting a ghost with no acknowledgment that there is a fight.
- The undisclosed inheritance: the problem discovered in month two that changes the job
- The undefined decision boundary: what this person signs without asking, in writing
- The predecessor comparison: named openly, or fought silently for a year
What does this cost if you get it wrong?
Three things. A placement fee of $30,000 at a $150,000 base. The salary paid before the exit. And a second search in a hard market: 91.7 percent of the 1,041 contractors answering the salaried-hiring question in the 2025 AGC and NCCER workforce survey reported difficulty filling salaried positions.
The second search is harder than the first, and not only because the market is tight. The seat now has a history. Candidates ask why the last person left after nine months, and there is no good answer that does not raise a question about the firm. If the seat has to be refilled quietly at that point, it becomes confidential executive search for mandates that cannot be publicly posted, which is a harder and slower engagement than the first one was.
Against that, onboarding costs attention. A structured hundred days is perhaps twelve hours of the hiring manager’s time. Spread over three months. Plus one honest conversation in week one. The asymmetry is not close.
Who does not need a hundred-day plan?
An internal promotion into a seat the person already understands, in a firm they have worked at for years, needs a decision boundary and little else. The information and access problems are already solved, which is much of the argument for promoting from within where the bench supports it.
The exception inside the exception is worth naming. An internal promotion into a seat where the person now manages former peers is not a light-touch transition, whatever the org chart suggests. That is a decision-boundary problem in its sharpest form, and it needs more definition than an external hire does, not less.
The other case where a plan does not help is a genuine interim, brought in for one bid cycle or one troubled project. Their remit is narrow and time-boxed by design, and the interim leadership decision is a different calculation altogether.
For everyone else, the hundred days is where the search either pays for itself or does not. The firm did the hard part already: it found someone good and persuaded them to move. Spending twelve hours protecting that is not an overhead. It is the last, cheapest step of the search, and it is the one most often skipped. If you have an offer out and no plan behind it, that is worth a conversation with employers before the start date, not after.
Sources and further reading
- AGC, Associated General Contractors of America
- NCCER, construction education and research
- BLS wage data, construction managers (11-9021)
- BLS wage data, cost estimators (13-1051)