A preconstruction estimating exercise for a finalist should take two hours, use a redacted set from a job you have already built, and test judgment rather than throughput: the assumptions a candidate states, the scope gaps they find, how they carry risk and contingency, and how they explain the number. It should never be live bid work.
Two hours with a redacted set of drawings will tell you more about a preconstruction leader than three rounds of conversation. A well-built estimating exercise is the closest you will get to watching a candidate do the job before they have it, and most contractors either skip it or quietly turn it into free estimating.
Both mistakes cost you the evidence. Skip it and you hire on how well someone talks about bid day. Turn it into production work and the strongest finalists, the employed ones, decline. Here is the kit, in order.
What should a preconstruction estimating exercise test?
It should test judgment on an incomplete set, not takeoff speed. Score four things: which assumptions the candidate writes down, which scope gaps they find, how they size risk and contingency, and how they explain the number to someone who has to sign it.
Takeoff is the part software increasingly covers. Pricing was never the bottleneck in preconstruction. Scope definition was.
The FHWA cost guidance asks that assumptions be documented, “including changes and what is and what is not in the estimate”, and that contingency be quantified as specific risk elements. Written for state transportation departments, it reads like a rubric for a chief estimator.
Tying the exercise to the seat also keeps it defensible: the EEOC testing guidance treats work samples as a selection procedure that must be job-related. For an industrial estimator, job-related means craft hours and productivity factors, not a building set priced by the square foot.
Which exercise, though, depends on which seat.
Which estimating exercise fits which seat?
Match the exercise to the decision the seat owns. A chief estimator gets a bid-day scope-gap review across subcontractor numbers. A preconstruction manager gets a conceptual budget review at schematic design. A director gets the budget review plus a short owner presentation.
- Bid-day scope-gap review, for a chief estimator: redacted subcontractor quotes for two trade packages, plus the spec sections. Level them: what each excludes, which exclusions are money, which sub to carry.
- Conceptual budget review, for a preconstruction manager: a schematic set and the budget your team submitted at that stage. Mark it up: what is missing, what is light, what the contingency covers. See our conceptual estimator note.
- Owner-facing variant, for a director of preconstruction: the same review, then ten minutes presenting it to two colleagues playing owner’s rep and architect.
The scope-gap review rewards hard-bid instincts; the budget review rewards negotiated ones. Those are often different people, as our piece on hard-bid estimating sets out.
Choosing the exercise is the easy part. Building the packet is where contractors leak things.
How do you build the packet from a real set?
Use a job you have already built and closed out. Redact the owner, the architect, the address and your unit rates, but keep the real budget and the real buyout. The buyout and the change-order log become your answer key, which no invented case study can give you.
A completed job knows how the story ended. If the candidate flags the curtain wall allowance as light, the change order that followed tells you whether they were right. Miss the site utilities exclusion that hurt you at buyout, and that is a scored miss, not an opinion.
Pick the job for the work you are hiring into. A finalist for a negotiated healthcare pursuit in Nashville should see a healthcare set, not the tilt-wall distribution center you built outside Dallas. Twenty sheets and the relevant spec divisions are plenty; a full set signals that you want takeoff. Then decide how long they get, which is where most exercises go wrong.
How long should the estimating exercise run?
Two hours, in person or on a video call, finished in the room. A take-home with no limit tests who has a free weekend, and the employed finalist you most want is usually the one who does not.
Split the two hours deliberately: 15 minutes to read the set and ask clarifying questions, 60 minutes to work, 30 minutes walking the panel through the result, and 15 minutes of panel questions. That is 120 minutes, every one of them observed.
The question window matters. A candidate who asks which delivery method the budget assumed, or whether a geotechnical report exists, has shown you how they start every pursuit. Those questions are scored.
Keep it for finalists. The OPM assessment guidance says work samples are costly to administer and fit best when few applicants are expected to have critical competencies, and that applicants tend to see them as fair. Two or three finalists for a chief estimator chair is that case.
Two hours of evidence is only useful if the panel reads it the same way.
How do you score the exercise?
Score each of four dimensions from 1 to 4, independently, before the panel talks. Do not score the total against your own budget. A number that lands close for the wrong reasons is worth less than one that misses for reasons the candidate stated.
| Dimension | Strong evidence | Weak evidence |
|---|---|---|
| Assumptions stated | A written list before any number: delivery method, escalation date, what the set omits | Assumptions surface only when the panel asks |
| Scope gaps found | Names each exclusion and prices it, or says where it is carried | Finds gaps but cannot say which are money |
| Risk and contingency logic | Contingency tied to named risk items, sized and explained | One percentage on the total, justified by habit |
| How they talk about the number | Gives a range and says what would move it | Defends a single figure, or will not commit |
The third row has a floor: the BLS estimator profile treats allowances for wasted material, bad weather and shipping delays as ordinary parts of any estimate. A finalist should name the specific risk this specific set carries.
Panels underrate the fourth row. FHWA notes an early estimate can be expressed as a range or with a confidence level; a candidate who does that for your panel will do it for your owner. Write one sentence of evidence beside every score, as our interview scorecard does.
What keeps the exercise confidential on both sides?
Protect both directions. You redact the owner, the address and your unit rates. The candidate must not bring, cite or reconstruct a current employer’s estimates, rates or bid results. Put both rules in writing, so a well-meaning finalist never hands you a trade-secret problem.
Contractors forget the second rule. The federal trade secret definition covers business and engineering information, compilations included, that the owner keeps secret and that has value from not being generally known. A competitor’s unit-rate history fits comfortably. A finalist who offers “what we carried for MEP coordination on our last hospital” is trying to help; stop them politely. This is not legal advice, so have counsel read the instruction sheet.
On a replacement search with the incumbent still seated, a set from a job they priced can identify your firm and the seat. Use one they never touched.
The rest of the risk is self-inflicted.
What not to do
Do not send a live bid. Candidates can tell, usually from the date on the title block, and the good ones will conclude your preconstruction department is stretched thin enough to recruit its own bid day. They will be right.
Do not grade it against your own number. Your number was also an estimate. The change-order log is the answer key; your original budget is just an earlier attempt.
Do not keep the output. The marked-up set belongs to an assessment, not to your next pursuit.
None of this is complicated. Three rounds of conversation tell you how a finalist talks about bid day; two hours with a closed job, built before the first one walks in, tell you how they run it. The exercise belongs in the brief before a single candidate is approached, because confidential executive search for mandates that cannot be publicly posted leaves no room for a second shortlist. The preconstruction manager page sets out how that seat is searched.
Sources and further reading
- OPM assessment guidance, work samples and simulations
- EEOC testing guidance, employment tests and selection procedures
- FHWA cost guidance, major project cost estimating
- BLS estimator profile, Occupational Outlook Handbook
- Trade secret definition, 18 U.S.C. 1839