Sources methodology
Primary sources first. Conflicts preserved.
CompressorFinder records where a claim came from, when it was captured, and whether another source disagrees.
A spec is not durable just because it appeared once on a product page.
Source priority is explicit.
Manufacturer manuals, spec sheets, and official product pages for model numbers, flow ratings, pressure, voltage, duty cycle, tank size, weight, and maintenance schedules.
Standards bodies and compressed-air references for air-quality classes, breathing-air requirements, dryer sizing, pressure drop, and safety context.
Authorized dealer and retailer listings for current price, availability, shipping constraints, and bundled accessories.
Installer and CAC directories for service area, authorization, certification, and contact verification.
Each important claim needs enough context to re-check.
- Source URL
- The canonical page, PDF, retailer listing, or directory entry.
- Capture date
- The date the source was read or the price was observed.
- Locator
- Page number, section label, table name, or visible product block where practical.
- Hash
- A content hash for drift detection when the source format supports it.
Conflicting specs are not averaged away.
The page should prefer the most authoritative source, describe the conflict where it matters, and preserve the lower-confidence value in the research payload.
Example: if an OEM manual and a retailer listing disagree on SCFM, the manual wins unless a newer official revision explains the change.
Sources we do not trust by default.
Unsourced specs copied from comparison snippets.
Affiliate-style rankings that hide criteria.
Schema generated from generic SEO plugins instead of the structured data model.
Price claims without a visible verification date.
Installer authorization claims that cannot be tied to an OEM or distributor source.