Methodology & Transparency

How every score is built.

We publish the inputs, the verification chain, and the exact formula behind every composite rating. No black boxes, no sponsored re-rankings.

01 · Data Auditing

Every rating starts with published, verifiable evidence.

We ground each rating in published research and verified manufacturer specifications — not spec-sheet claims or sponsored content. Sources that have been retracted, or that we cannot independently verify, are set aside before they inform a score.

Where a clinical claim carries a citable source, we record it — so the evidence behind a rating can be traced rather than taken on trust.

02 · Consumer Review Verification

Aggregated real-world sentiment from verified consumer platforms.

For each unit, consumer sentiment is aggregated from public review platforms and independent review sources using a documented deep-research synthesis process. The output is normalized to a 1–5 scale with two decimal precision and recorded as the Consumer Rating on every rating card.

AI aggregation handles scale; the sources are the authority. Platforms included are independent of manufacturer relationships. Consumer Ratings are planned to undergo a multi-model adversarial review pass as the methodology matures — divergence between models flags units for manual re-examination.

03 · Mathematical Formula

The composite is a weighted index, not an opinion.

Every total rating is the deterministic output of a four-axis weighted index. The weights never change for a sponsored placement and are published below.

Total Rating = 0.40 · Efficacy + 0.30 · Safety + 0.20 · Build + 0.10 · UX

Quality and Consumer ratings are symmetric around the composite: mean(Quality, Consumer) ≡ Total Rating.

40%Scientific Efficacy

Thermal Output, Heat Distribution

30%Safety & Purity

EMF Mitigation, Non-Toxic Materials

20%Build Quality

Insulation, Warranty

10%User Experience

Controls, Ergonomics

04 · Editorial Independence

Affiliate disclosure, never affiliate re-ranking.

SaunaScience may earn a commission when you click a link and make a purchase. Those commissions never influence the rankings — which products appear, in what order, or how they score. The rating methodology runs independently of any commercial relationship.

Every outbound link is rendered with rel="sponsored" per FTC guidance.