Cost estimating methods and the four types of estimates
Estimates are built differently depending on how much is known. Early on, you compare to similar projects. Later, you price every item. Knowing which method fits the stage, and stating its expected accuracy, keeps expectations honest. AACE International's estimate classes give everyone a common language for that.
- Analogous
- Parametric
- Assembly
- Detailed bottom-up
The four common types
| Method | Inputs | Effort | Typical use |
|---|---|---|---|
| Analogous (comparative) | Cost of similar past projects, adjusted for size, location and time | Low | Feasibility, early budgets |
| Parametric | Cost drivers and ratios: $/sf, $/bed, $/key, $/ton | Low to medium | Conceptual budgets |
| Assembly (systems / unit cost) | Quantities of assemblies × unit costs (wall type, roof system, MEP system) | Medium | Schematic and design development |
| Detailed (bottom-up) | Full takeoff of every item, priced with labor, material and equipment | High | Bids, GMPs, check estimates |
Names vary by source. Some lists call the "four types" preliminary, detailed, quantity and bid estimates. The four above are the underlying techniques.
AACE estimate classes
For building and general construction, the relevant AACE recommended practice is 56R-08 (rev. August 2020). The often-cited 18R-97 applies to the process industries. 56R-08 classifies estimates from Class 5 to Class 1 by how mature the project definition is:
| Class | Project definition | Typical purpose | Typical method | Expected accuracy (low / high) |
|---|---|---|---|---|
| 5 | 0%–2% | Functional area or concept screening | SF or m² factors, parametric models | −20% to −30% / +30% to +50% |
| 4 | 1%–15% | Schematic design or concept study | Parametric, assembly-driven models | −10% to −20% / +20% to +30% |
| 3 | 10%–40% | Design development, budget authorization | Semi-detailed unit costs with assembly-level line items | −5% to −15% / +10% to +20% |
| 2 | 30%–75% | Control or bid/tender | Detailed unit cost with forced detailed takeoff | −5% to −10% / +5% to +15% |
| 1 | 65%–100% | Check estimate, bid/tender, change order | Detailed unit cost with detailed takeoff | −3% to −5% / +3% to +10% |
AACE stresses that these ranges assume appropriate contingency, and that a Class 5 estimate on a repeat project with good cost history can be as accurate as a Class 3 on a novel one.
Parametric vs. detailed
Parametric (Class 4): historical $285/sf for similar offices, adjusted +6% for time and location
40,000 × $285 × 1.06 ≈ $12.1M → range ≈ $9.7M to $15.7M using −20% / +30%
Detailed (Class 2, at bid): sum of trade bids + general conditions + fee = $12.9M
Check: $12.9M is inside the parametric range. If it weren't, look for scope gaps or quantity errors first.
What they are
A preliminary estimate is built before design is complete, usually with analogous or parametric methods. It supports feasibility and budget decisions. It should state its basis, class and expected accuracy, and it should never be used as a bid.
Match the stage
Use analogous or parametric methods when drawings are schematic, assembly methods as systems become clear, and detailed estimates when documents are complete. Subcontractor bids are almost always detailed estimates. State the method and class on the estimate. People remember the number, not its basis.
Using methods together
Experienced estimators use more than one method on the same bid. The detailed estimate produces the number, and a parametric check (cost per square foot or per unit for similar work) tells you whether it's in a sensible range. Subcontractors rarely bid from a parametric number, but they use one to screen opportunities and sanity-check the final price.
Where each method misleads
- Analogous and parametric fail when the reference project differs in ways the parameter doesn't capture. A lab and an office of the same area cost very different amounts per square foot.
- Assembly estimates fail when the assembly library is out of date.
- Detailed estimates usually fail through missed scope, not bad math.
- Unit cost databases (e.g. RSMeans) need location and time adjustment, and they don't know your crews' productivity.
Detailed estimates, faster
RenderDraw's AI takeoff supports detailed estimates by proposing quantities with the source sheet on every row for estimator review. Pricing, method and class stay with you. Get a free sample takeoff.
Questions estimators ask.
What are the four types of estimating?
Commonly analogous, parametric, assembly (unit cost) and detailed bottom-up.
What are AACE estimate classes?
For building and general construction, AACE 56R-08 classifies estimates from Class 5 (0–2% project definition) to Class 1 (65–100%), with expected accuracy narrowing from about −20/−30% to +30/+50% at Class 5 down to −3/−5% to +3/+10% at Class 1.
Is 18R-97 the right AACE practice for buildings?
No. 18R-97 covers the process industries. Building and general construction projects use 56R-08.
What is a preliminary estimate?
An early estimate made before design is complete, usually with analogous or parametric methods, labeled with its basis and expected accuracy.
Which method is used for subcontractor bids?
Detailed bottom-up estimates from a full takeoff, sanity-checked with a parametric number.
Move from ratios to real quantities faster.
Send a set. RenderDraw proposes quantities with sources so detailed estimates start sooner.
- Upload the plans, specs and schedules, plus an addendum if you have one.
- RenderDraw reviews the set and the scope you want tested.
- We follow up with extracted quantities, flagged items and next steps.
Where these numbers come from
We checked this guide against the standards, manufacturer documents and industry references below (last reviewed November 2026). Figures marked as rules of thumb are typical ranges, not requirements; your project documents, local code and your own cost history govern.
- AACE 56R-08 Cost Estimate Classification System, building and general construction (sample/TOC PDF) — class matrix and accuracy ranges