Module 2 - Cost Management
Cost management on the exam is less about arithmetic than about judgment: knowing which estimate is appropriate at which stage, what a number actually includes, and what the CM owes the owner when the number moves.
Estimates are built using one or a blend of three classic methods. An analogous (or historical) estimate scales the known cost of a similar past project to the new one; it is fast and needs little detail, but only as good as how truly comparable the reference project is. A parametric estimate applies a statistical relationship between a project characteristic and cost, such as dollars per square foot or per bed; it is the standard method for early, low-detail estimates. A bottom-up (engineering) estimate prices individual work packages or line items from a detailed quantity takeoff and current unit costs, then sums them; it is the most accurate method but requires the most design detail and the most time to produce. As a project moves from Pre-Design toward Construction Documents, the estimating method shifts from analogous and parametric toward bottom-up, tracking the same maturing detail that drives the estimate classes below.
Estimates get more accurate as design matures, and the exam expects you to match the estimate type to the phase.
An order of magnitude or conceptual estimate is produced in pre-design, before drawings exist. It relies on historical cost per square foot, per bed, per lane mile, or a similar parametric unit. Expected accuracy is wide, commonly cited around minus 30 to plus 50 percent. Its purpose is to test feasibility, not to set a contract price.
A schematic or square-foot estimate follows at roughly 30 percent design. Major systems are identified but not detailed. Accuracy tightens to something like plus or minus 20 percent.
A design development estimate at roughly 60 percent design breaks the project into assemblies and systems, with quantities taken off where documented. Accuracy near plus or minus 10 percent.
A construction document estimate at 90 to 100 percent design is a full quantity takeoff priced at current market rates. This is the estimate against which bids are compared, and accuracy should be within roughly 5 percent.
A definitive estimate is the most detailed class, priced from complete documents. The contractor's bid is a different thing: it is the contractor's own priced proposal, not the CM's estimate, though the CM evaluates it.
The accuracy bands above are teaching values. Published systems differ and use wider, uneven ranges. AACE International's Recommended Practice 18R-97, for example, lists Class 5 (concept screening) at about minus 20 to minus 50 percent on the low side and plus 30 to plus 100 percent on the high side, and Class 3 (budget authorization) at about minus 10 to minus 20 percent and plus 10 to plus 30 percent. Rely on the pattern that the range narrows as design matures, and state the range you are using.
Why this matters for the exam: Scenario questions will describe a design stage and ask which estimate is appropriate, or will describe an owner demanding a guaranteed number at 15 percent design and ask what the CM should do. The correct answer is almost never to give false precision. It is to state the estimate class, its accuracy range, and the assumptions behind it.
Beyond matching the estimate class to the design stage, a reliable cost estimate has four characteristics, drawn from federal cost-estimating guidance and equally applicable to CM budgets. It is comprehensive: it includes all life cycle costs, is based on a technical baseline (design and scope description) that completely defines the program, and uses a WBS detailed enough that cost elements are neither omitted nor double-counted. It is well-documented: the methods, calculations, data sources, and assumptions behind every cost element are recorded well enough that someone else could trace or repeat the estimate later. It is accurate: it uses the best available methodology and data for each element, is adjusted properly for inflation and escalation, contains few if any mathematical errors, and is updated as the program changes. It is credible: it discloses its own limitations, includes a sensitivity analysis showing how the total changes if key assumptions move, and, where the stakes justify it, is cross-checked against an independent estimate. An estimate can be numerically correct and still fail this standard if it skips documentation or never discloses its assumptions.
An estimate prices the work. A project budget prices the project, which is a larger thing. A complete budget includes construction cost, design fees, CM fees, land or site acquisition, permits and regulatory fees, furniture, fixtures and equipment, testing and inspection, insurance and bonds, escalation, and contingency. A CM who reports only construction cost to an owner has under-reported the project, and the exam treats that as a failure of the CM's duty.
Two distinct contingencies appear on the exam and confusing them is a common trap. Design contingency covers the cost growth that happens as an incomplete design is completed. It is largest early and is drawn down as design matures. At 30 percent design it might be 10 to 15 percent of construction cost; by 100 percent documents it approaches zero because the scope is now fully drawn. Construction contingency covers unforeseen conditions during construction, such as differing site conditions or unanticipated coordination. It is not intended to fund owner-directed scope additions. When an owner adds scope, that is a budget increase, not a contingency draw, and a CM who quietly absorbs owner scope into contingency is hiding a real cost increase.
Allowances are a third thing again: a specified dollar amount carried in the contract for a scope item that is known to be required but not yet selected, such as a flooring allowance. When the actual selection is made, the contract price is adjusted up or down by the difference. An allowance is a placeholder for a decision, not a reserve against risk. Escalation accounts for cost increases over time between the estimate date and the time the work is actually bought. On a project buying steel eighteen months out, omitting escalation is not conservatism, it is an error.
Owners frequently choose based on first cost when the exam's correct answer is life cycle cost. A life cycle cost analysis considers acquisition, operation, maintenance, replacement, and disposal over the asset's useful life, usually discounted to present value. A more expensive mechanical system with lower operating cost may win on life cycle terms. Expect at least one question where the "cheapest" option is the wrong answer.
Why this matters for the exam
Almost every cost question on this exam is really asking one of two things: does this number match the amount of real design behind it, and does this dollar belong in the category the scenario claims it belongs in (contingency versus allowance versus escalation versus a real scope increase). Get the category right and the arithmetic mostly takes care of itself.
1. An owner asks for a cost figure at pre-design, before any drawings exist. The appropriate estimate is:
2. Design contingency exists primarily to cover:
3. An owner adds a new wing to the program during design development. The CM should:
4. A flooring line carried at a fixed dollar amount pending final product selection is:
5. In this lesson's teaching scale, which estimate class would you expect to be accurate to roughly plus or minus 10 percent?
6. A project will not purchase structural steel for eighteen months. Omitting escalation from the estimate is:
7. A mechanical system costs more to install but substantially less to operate over a 25-year life. The analysis that captures this is:
8. A CM reports only the construction contract value as "the project cost." This is:
9. An estimate built by scaling the known cost of a comparable past project to the new one is:
10. A bottom-up (engineering) estimate is generally the most accurate method because it:
11. An estimate that discloses its own limitations, includes a sensitivity analysis, and is cross-checked against an independent estimate is demonstrating which characteristic of a reliable estimate?
12. An estimate whose methods, calculations, data sources, and assumptions are recorded well enough for someone else to trace or repeat it later is demonstrating which characteristic?
13. An estimate that is numerically correct but never discloses its assumptions or a sensitivity range would still fail which standard?
14. As a project moves from Pre-Design toward Construction Documents, the estimating method typically shifts:
Take any project you know. Build a one-page budget with these line categories: construction cost, design fees, CM fees, permits and fees, FF&E, testing and inspection, insurance and bonds, escalation, design contingency, construction contingency. Populate it with rough figures. Then write two sentences identifying which single line you are least confident in and what would have to happen for that number to firm up, and note which of the four estimate characteristics (comprehensive, well-documented, accurate, credible) that line is currently weakest on. Keep this. You will re-use the budget structure on Day 7.