Conceptual estimate: accuracy, methods and when to use one

Owners use a conceptual estimate to decide whether to pursue a project. Contractors are asked for one during preconstruction, often with little more than a program, a site and a target square footage.

It is not a bid. It is a screening number with a stated range, and it should be labeled that way every time it leaves your desk.

Less accurate than most people expect. AACE International Recommended Practice 18R-97, the classification most cost engineers cite, puts early estimates in two classes:

A conceptual estimate is usually Class 5 or Class 4. On a Class 4 number, the final cost could reasonably land anywhere from about 30% below it to 50% above it. That range is not a failure of the estimator; it reflects how little of the project is defined.

Label every conceptual number with its class and range. A single unqualified figure gets copied into board decks and loan applications as if it were a bid.

• Class 5: 0% to 2% project definition, used for concept screening. Expected accuracy roughly -20% to -50% on the low side and +30% to +100% on the high side.
• Class 4: 1% to 15% project definition, used for study or feasibility. Expected accuracy roughly -15% to -30% low and +20% to +50% high.

• Conceptual estimate: prepared from a program or concept sketch. Answers whether to pursue the project.
• Budget estimate (often Class 3): prepared from schematic or early design development documents. Sets the funded budget.
• Bid estimate (Class 2 or 1): prepared from construction documents and subcontractor quotes. Becomes a price someone is held to.

Most experienced estimators combine methods: square foot cost as a sanity check, systems cost for the real number.

• Square foot or unit cost: multiply gross area by a cost per square foot from comparable projects or a published cost database, adjusted for location, building type, quality level and the date of the data.
• Assembly or systems cost: price foundations, structure, envelope, roofing, interiors, MEP and sitework separately, each with a parametric unit. Slower, but it shows the owner where the money goes.
• Capacity or parametric factoring: common in industrial work. Scale cost from a known project by capacity (beds, parking spaces, gallons per day) with an exponent, because cost rarely rises one to one with size.

• Basis of estimate: what documents existed, what you assumed, and the cost data source and its date.
• Scope boundaries: land, design fees, FF&E, permits, utility fees and owner contingency, each marked in or out.
• Location and escalation: a location factor, and escalation to the expected midpoint of construction rather than today.
• Markups: general conditions, overhead, profit, bonds and insurance as separate lines.
• Design contingency for undefined scope, larger now than it will be at bid time.
• The estimate class and its low and high range.

• Using cost data from another region or year without adjusting it.
• Leaving out sitework and utilities, which can swing early numbers more than the building itself.
• Treating contingency as padding to remove later instead of the price of undefined scope.
• Presenting one number with no range.

If a conceptual number will drive a go or no-go decision, a financing request or a GMP conversation, it deserves the same documentation as a bid. Bidmere prepares budget and conceptual estimates with a written basis of estimate and a stated accuracy range. Send whatever you have: a program, sketches or a site plan.