Logo
HVAC RESEARCH DIAGNOSTIC TOOLS THE DIAGNOSTIC WHY BUILT ON TENTH THE FIELD REPORT
ORDER THE DIAGNOSTIC — $1,497

METHODOLOGY · BUILT ON TENTH

How We Source, Model, and Publish Data

Last updated April 28, 2026. The transparency anchor for every published article and every Market Diagnostic.

Primary Public Sources

Most quantitative claims in our published research and in the Market Diagnostic are anchored to verifiable public sources:

  • U.S. Bureau of Labor Statistics (BLS): Occupational Employment and Wage Statistics (OES), Employer Costs for Employee Compensation (ECEC). Source for HVAC tech wages, CSR wages, and loaded-cost burden rates.
  • U.S. Census Bureau: American Community Survey (ACS), County Business Patterns, Economic Census. Source for housing stock age, contractor density, and metropolitan demographic data.
  • NOAA National Centers for Environmental Information (NCEI): Climate Data Online. Source for cooling-degree-days and weather-related metric data.
  • Google Maps and Google Business Profile: Live capture for review counts, ratings, and map pack positions. Date-stamped at the time of capture.
  • Vendor public disclosures: SEC filings (Synchrony, Truist, Regions), official company blogs (Google, vendor announcements), and pricing pages where vendors publish list prices.
  • Industry trade publications: Contracting Business, ACHR News, ACCA member resources, where data is referenced with named sources.
  • Consumer cost guides: Angi, HomeAdvisor, Forbes Home — used for residential pricing ranges with explicit acknowledgment of self-reporting bias.

What's Cited vs. What's Modeled

We make a clear distinction between cited statistics (drawn directly from a primary source) and modeled illustrative math (built from public-knowledge industry assumptions to make a structural argument).

  • Cited statistics include the source name and link. Example: "BLS May 2024 OES data: median HVAC technician wage $27.51/hour."
  • Modeled math is labeled as illustrative and shows the assumptions behind the calculation. Example: "Modeled at 1.30x loaded-cost multiplier on a $32/hr base wage."
  • Where industry-typical figures appear without a single peer-reviewed source (e.g., HVAC callback rate ranges, attach rate ranges), we say so explicitly and frame the figure as "industry-typical" with the methodology caveat.

Mystery-Shop Protocol

Two distinct mystery-shop tracks: (1) the city benchmark of the top 7 operators in each covered metro, captured once and refreshed quarterly for our published articles, and (2) the single client-shop call performed for every paid Market Diagnostic. Paid reports compare the client's one shop call against the on-file city benchmark — competitors are not re-shopped per buyer.

  • Calls are placed by a human researcher using a homeowner persona. No AI or synthetic voice is used.
  • One call per operator per data collection cycle.
  • No appointment is booked. No truck is dispatched. No business resources are consumed beyond the inbound call itself.
  • No call is recorded. Three of our covered cities (Tampa, Orlando, Las Vegas) are all-party-consent jurisdictions; uniform no-record policy applies across all cities.
  • CSR personal names are never captured in any data column or published in any article. Observations reference "the CSR" only.
  • Observations are factual and time-stamped (e.g., "rang 7 times to voicemail at 10:14am ET on April 28, 2026"). No characterizations of CSR competence or attitude are published.

Single-call captures are illustrative of intake design and protocol on that specific call. They are not statistically representative of an operator's full performance across all inbound calls at all times. The benchmark dataset (top 7 per city) gives the comparison weight a single client-shop on its own cannot.

Vendor & Brand Naming

Articles name specific vendors and operators (Synchrony, GoodLeap, Wisetack, ServiceTitan, Frank Gay Services, Coolray, Parker & Sons, etc.) for descriptive accuracy. Built on Tenth has no affiliate relationships, receives no compensation from any vendor or operator named in any analysis, sells no advertising, and offers no marketing agency retainers.

All trademarks belong to their respective owners. Use of brand names is descriptive only; no endorsement, partnership, or commercial relationship is implied or claimed.

Correction Policy

If your operation believes any data point in a Built on Tenth article is inaccurate or misrepresents your business, email hello@builtontenth.com with:

  • The article URL and the specific claim you dispute
  • The substantiation showing the claim is inaccurate (e.g., your own current data, screenshots, time-stamped records)
  • What you'd like updated or retracted

We will investigate and update or retract substantiated corrections within 5 business days. The original publication date is preserved; the correction date and a brief note are added to the article.

Data Freshness

Every article carries a pubDate and updatedDate. Vendor pricing, review counts, and map pack data are time-sensitive — verify current values directly with each source before making business decisions. Where we know a figure is likely to drift fast (vendor pricing, regulatory changes), we date-stamp the capture explicitly inside the article.

Built on Tenth has no affiliate relationships, accepts no referral fees, sells no advertising, and offers no marketing agency retainers. Independence is the product.