Every figure on this site is either sourced and cited, or explicitly labelled as our own assumption. This page explains which is which.
Where the data comes from
Cost figures are drawn from published US cost guides and, where possible, cross-checked across more than one. Our primary sources are HomeGuide, Angi, HomeAdvisor and Fixr for consumer cost ranges; the Remodeling Cost vs. Value report for regional multipliers and fixed-scope project costs; retailer pricing from The Home Depot and Lowe’s for current material costs; and manufacturer or retailer estimating documents for material quantity ratios.
Each calculator page carries its own numbered source list. If a number appears on this site, you can trace it.
How we handle disagreement between sources
Sources disagree more often than most cost sites let on, and the disagreements are not random. Angi’s figures derive from customer surveys and run consistently below HomeGuide’s contractor-quote ranges. Some sources include disposal or preparation in a headline number and some do not.
Our approach: where sources broadly agree, we use the consensus range. Where they diverge materially, we say so on the page, explain why we think they differ, and pick the more conservative figure. A budget that turns out generous is a better outcome for a reader than one that turns out short.
Regional multipliers apply to labor only
Materials move through national supply chains at close to national prices. Labor does not. So applying a flat regional multiplier to an entire project cost overstates the regional effect, often by roughly double.
Every calculator here splits the installed rate into its material and labor components, applies the regional factor to labor alone, and recombines. The labor share used for each trade is stated in the calculator’s source code and is drawn from published material-versus-labor splits.
Our regional multipliers are derived from the Cost vs. Value report, which prices an identical specified scope in each region — giving a clean regional-to-national ratio without the scope drift that affects survey averages.
What is an assumption, not a source
Some things every contractor knows are not published anywhere usable. Where we have had to fill a gap with our own judgement, we mark it on the page with a dagger (†) or a plain statement. The current list:
- Steep-slope roofing labor uplift. Every source agrees pitch raises labor above 6:12 or 7:12; none publishes a tiered figure, because those live in Xactimate’s proprietary price list. Our 10–20 percent uplift above 7:12 is an assumption. The geometric pitch-to-area multiplier, by contrast, is trigonometry and is exact.
- Deck height uplift. Published per-square-foot deck data does not isolate elevation. Our 12 percent for raised and 25 percent for elevated decks is an assumption.
- Flooring layout uplift. Diagonal and herringbone layouts demonstrably cost more in both waste and labor; the labor uplift figures are ours.
- Gate pricing. Per-linear-foot fence data does not separate gates. Our $150–$600 per gate is an assumption — get gates quoted separately.
What we deliberately do not publish
We have working data for kitchen remodels and asphalt driveways but have not published calculators for them, because the underlying sources are too weak. Three major sources give irreconcilable kitchen cost breakdowns — labor at 15, 25 and 50–60 percent of the same job — because they are measuring different bases. Rather than average three numbers that cannot be averaged, we have left it out until we have something defensible.
Review cycle
Cost data is reviewed and re-sourced at least twice a year. The review date appears at the top of every calculator page. Construction costs move, and a cost site that does not say when it last checked is asking you to trust a number of unknown age.
Corrections
If you are a contractor or estimator and a figure here does not match what you see in the field, we want to hear about it. Write to hello@benchcost.com with the page and the number. Corrections are made promptly and the review date updated.