Cost Engineering & Fact Verification Methodology
HouseCostData is designed as an authoritative data intelligence product for American homeowners. We reject the generic affiliate model of calculating prices via arbitrary city multipliers.
Principle 1: Zero-LLM Numerical Factual Layer
Large Language Models (LLMs) are strictly forbidden from generating, estimating, or adjusting numbers in our factual layer. Every number displayed must trace to an authoritative government source snapshot (such as the U.S. Bureau of Labor Statistics or U.S. Census Bureau) or a versioned deterministic equation with published inputs.
Principle 2: Separation of Wages from Contractor Billing Rates
A common flaw on home improvement blogs is multiplying employee W-2 wages by an arbitrary multiplier and presenting the result as a verified local rate. We maintain strict conceptual separation:
- EMPLOYEE_WAGE_OBSERVED: W-2 hourly wages reported by employers to the BLS OEWS survey. Published with percentiles (10th, 25th, Median, 75th, 90th).
- CONTRACTOR_BILLING_RATE_OBSERVED: Actual customer invoice rates billed by licensed trade contractors. If no authoritative public survey exists, this field is flagged as DATA_UNAVAILABLE.
Principle 3: Why Dollar Calculators are Delayed Until Verified
To generate a legitimate commercial price estimate, a system requires three verifiable pillars:
- Empirically validated contractor billing rates for the metropolitan area.
- Legally licensed wholesale bill-of-materials pricing for trade equipment.
- Verified labor hours based on standardized trade unit standards.
Because wholesale electrical switchgear feeds and commercial contractor billing surveys are currently unresolved dependencies, HouseCostData refuses to publish speculative dollar ranges. We lead with verified local market data instead.