Evaluating Cost Estimates: Accuracy Ranges, Review Process, and Reconciliation with Budget

How architects evaluate cost estimates prepared by others during preliminary design, including understanding accuracy ranges at different design stages, reviewing estimate methodology and assumptions, and reconciling estimates with the construction budget through design alternatives.

ARE · Project Planning & Design

Overview

A cost estimator hands you a number. Maybe $12.4 million for a community library. That figure feels precise, definitive. It isn't. Every cost estimate carries an accuracy range tied to how much design information exists at the time it was prepared. At schematic design, that range could swing 15 to 30 percent in either direction. By the end of design development, the band tightens to roughly 5 to 15 percent.

For PPD candidates, the critical skill is evaluating estimates prepared by others. You won't build these estimates from scratch on the exam; you'll assess whether they're reasonable, identify what method was used, check whether assumptions match the project scope, and determine what to do when the number lands above the construction budget.

Three cost estimating methods show up repeatedly: parametric (cost per square foot or per unit), assembly-based (pricing grouped building components), and detailed unit-cost (line-item quantities times unit prices). Each method suits a different design stage. Using the wrong method at the wrong time produces misleading precision or unnecessary vagueness.

When an estimate exceeds the budget, reconciliation follows. That means comparing line items against allowances, questioning assumptions about site conditions or material selections, and evaluating whether design alternatives can bring costs in line without gutting the program. Reconciliation is analytical work. It requires weighing competing factors: client priorities, code requirements, system trade-offs, and schedule constraints.

This topic connects directly to how architects protect clients from budget surprises during preliminary design, when changes are still relatively inexpensive to make.

Essential

Cost estimates aren't facts. They're predictions built on assumptions, and your job as an architect is to evaluate whether those predictions are sound. The ARE tests this at the analyze/evaluate level because cost estimate review requires judgment, not just plugging numbers into formulas. ## Cost Estimating Methods and When They Fit Three primary methods produce construction cost estimates. Each suits a different level of design completion. Parametric estimating uses cost-per-unit metrics. Cost per square foot is the most common in building design, but cost per hospital bed, per classroom, or per parking space also qualifies. This method works best during programming and early schematic design when the building program is defined but detailed systems haven't been selected. The estimator multiplies a unit cost (drawn from historical project data or published cost databases) by the project quantity. It's fast.

Professional

Consider an architect reviewing a schematic design cost estimate for a 45,000-square-foot community health center. The cost estimator reports $18.9 million, or $420 per square foot. The owner's construction budget is $16.5 million. The gap is $2.4 million, roughly 14.5 percent over budget. The first question isn't "how do we cut $2.4 million?" It's "how confident are we in $18.9 million?" At SD, the accuracy range might be -15% to +30%. That means the actual construction cost could land anywhere from $16.1 million to $24.6 million. The point estimate sits near the low end of the possible range. This context matters: the estimate isn't wildly over budget; it's within the accuracy band, and the budget itself may have been set without accounting for the normal range of SD-level estimates. Step two: check the Basis of Estimate.

TLDR

Three estimating methods: parametric (cost/SF, early stages), assembly-based (grouped systems, SD through DD), detailed unit-cost (line items, DD through CDs). Method must match design stage.
Accuracy ranges narrow as design progresses: Class 5 at programming (-50%/+100%) down to Class 1 at complete CDs (-10%/+15%). Ranges are asymmetric, skewing toward cost growth.
Four marks of a credible estimate: thorough scope, documented assumptions, appropriate methods, and results that survive independent review.
Review checklist: scope completeness, ground rules in the Basis of Estimate, method appropriateness, contingency percentage for the stage, escalation inclusion, and spot-check unit costs against published benchmarks.
Reconciliation means comparing estimate vs. budget by the same classification system (Uniformat or MasterFormat), identifying which categories drive the overage, and evaluating design alternatives for those specific categories.
Contingency covers design unknowns (15-20% at SD, under 5% at CDs). Escalation covers time-based cost increases. They're separate items.
A point estimate without an accuracy range or confidence level gives the owner incomplete information. Always ask what confidence level the estimate represents.
When two estimates disagree, compare them organized by the same WBS to find where they diverge. Don't just average them.

ELI5

Think of a cost estimate like a weather forecast. A seven-day forecast says "around 75 degrees." It could end up being 70 or 82. Nobody panics because everyone understands that far-out forecasts are fuzzy. A same-day forecast says "76 degrees." Much more precise, because more data is available. Cost estimates work the same way. Early in design, when you've only decided on the building's size and basic shape, the estimate is fuzzy. "Somewhere between $14 million and $22 million." That's not a bad estimate. It's an honest one, because so many decisions haven't been made yet. Which mechanical system? What type of exterior wall? How complex is the roof? All unknowns. As design moves forward and those decisions get locked in, the estimate sharpens. By the time every wall is drawn and every fixture is counted, the estimate narrows to a tight band. Maybe $17.8 million to $19.2 million.

AREProject Planning & Design

Evaluating Cost Estimates: Accuracy Ranges, Review Process, and Reconciliation with Budget

2 min read249 words

Cost Estimates Aren't Crystal Balls. Here's How to Read Them.

This topic connects directly to how architects protect clients from budget surprises during preliminary design, when changes are still relatively inexpensive to make.

Want to track your progress and access more study tools?

Create a free account