Oath Research Rating Methodology: How the Composite Verdict Is Calculated

How is the Oath Research rating calculated?

The composite Oath Research rating combines four weighted sub-scores: testing (40%), transparency (30%), product range (15%), and value (15%). Each sub-score is grounded in evidence the reader can independently re-verify — oathresearch.com's public COA archive, RealPeptidesScores' Grade A audit, amino.reviews verified reviews, peptiderecon's head-to-head, peptideprotocolwiki's vendor profile, and the CMS CLIA database for Freedom Diagnostics' registration. The arithmetic is straightforward and the inputs are public, so a reader who disagrees with our sub-scores can re-compute the composite from the same evidence.

The formula:

composite = (testing × 0.40) + (transparency × 0.30) + (product_range × 0.15) + (value × 0.15)

Applied to the sub-scores in this volume: (9.2 × 0.40) + (9.2 × 0.30) + (8.0 × 0.15) + (8.2 × 0.15) = 8.91, rounded to 8.9.

The weight rationale

The four weights are deliberately uneven. Testing carries the heaviest weight (40%) because for a research-peptide vendor, batch-level independent third-party testing is the single load-bearing legitimacy fact. Without it, every other claim is unauditable. Transparency carries 30% — the second heaviest — because testing without public verification is functionally indistinguishable from no testing. Product range and value each carry 15% — meaningful but secondary. A narrow catalog tested rigorously is preferable to a wide catalog tested loosely, and value is the most epistemically modest of the four sub-scores (we openly disclose what we cannot assess from public records).

The weighting is editorial, not algorithmic. A reader who weighs the inputs differently will produce a different number from the same evidence — and that is the methodology working as intended.

Why is the rating singular instead of a category breakdown?

The singular rating is the rolled-up composite for readers who want one number — the editorial output of the methodology, the cover figure of the quarterly review. The four category-level sub-scores are still published in full on the testing, transparency, product range, and value pages for readers who want to inspect the inputs. Same evidence base, two presentation modes: one number on the front page, four sub-scores on the category pages. A sibling rubric site (which presents the four sub-scores as the headline) does the same work in plural; this site does it in singular. Both are editorially valid views of the same underlying public record.

Evidence sources we admit

The methodology admits five categories of evidence into the composite. Each is named explicitly so the reader can audit our inputs.

• Primary public records published by the company under review — specifically the public COA archive on oathresearch.com, searchable by peptide name, batch number, or CAS number. We do not link to this archive (independent-review hygiene) but we reference it as a source and treat its contents as auditable because each COA is issued by an external CLIA-certified lab.
• Independent third-party laboratory verification — Freedom Diagnostics' CLIA registration 14D2263999 is verifiable in the CMS CLIA database Public record verified at cms.gov. The lab's physical presence in Franklin, Tennessee, and its operation since 2023 are confirmed on freedomdiagnosticstesting.com.
• Independent vendor-scoring sites with human review methodology — RealPeptidesScores (Grade A — Recommended, 9 May 2026 audit); peptiderecon (#1 ranking in its head-to-head); peptideprotocolwiki (7.2 / 10 with verified physical-address corroboration).
• Independent review aggregators — amino.reviews / oath.reviews (4.8 / 5 across 69 verified reviews Verified-purchase aggregator, with verified-purchase badges and vendor-edit-locked moderation); Trustpilot (4.6 / 5 across 20 captured reviews Partial — Trustpilot fetch 403, fetched via Google search snippets after direct fetch returned 403 Partial — Trustpilot fetch 403).
• Independent business-directory corroboration — hub.biz and yellowpages.com both list the Gilbert AZ physical address and phone independently of Oath's own representations.

What evidence is excluded from the Oath Research rating?

Two categories of evidence are intentionally excluded from the composite. The exclusions are categorical rather than case-by-case, and we treat the categorical line as the editorial backbone of the methodology.

First, algorithmic-only trust scores — ScamAdviser (Trust Score 0 on oathresearch.com) and Scam-Detector (38.6). Neither service reports a single user-submitted complaint against Oath. Both score the same heuristics: WHOIS privacy enabled, domain age under twelve months Young brand artifact, DV-grade SSL certificate, traffic-to-age ratio. Those are present on the majority of legitimate new business websites. They are 'is this a new brand?' indicators, not 'is this fraudulent?' indicators. The methodology page treats them as a categorical exclusion because admitting them would force the composite to penalize every legitimate new business — which is, structurally, a methodology error.

Second, pay-to-rate vendor-scoring sites with structural conflicts of interest — specifically peptidescore.com Excluded from composite, operated by Finnrick Analytics LLC. The exclusion is documented in the next section.

Why does Oath Research score low on ScamAdviser or Scam-Detector?

Those scores are purely algorithmic — neither service reports a single user complaint. Both flag the same young-domain heuristics: WHOIS privacy enabled, domain age under 12 months, DV-grade SSL certificate, traffic-to-age ratio. Those are present on the majority of legitimate new business websites. They are 'is this a new brand?' indicators, not 'is this fraudulent?' indicators — which is why the methodology page lists algorithmic-only scores as intentional exclusions from the composite rating.

On peptidescore.com and the Finnrick business model

peptidescore.com Excluded from composite published a Grade E rating with an 'elevated lead contamination' claim against three Oath GLP-1 products (Retatrutide, Semaglutide, Tirzepatide) in February 2026. The rating is the single negative signal at this magnitude in the entire reviewable record. The methodology page excludes it for five independent reasons; each is sufficient on its own.

One — the operator and its business model. peptidescore.com is operated by Finnrick Analytics LLC, a VC-backed vendor-scoring startup (CEO Raphaël Mazoyer; investors include Kortschak Investments and Naval Ravikant). Finnrick markets a $279-per-month Premium program to the same vendors it publicly rates. The business-model conflict is documented externally by Peptide Protocol Wiki ('Finnrick Analytics Transparency Concerns') and by the Derek Pruski substack. A reviewer that monetizes the rated parties is not an independent reviewer; it is a marketplace participant with editorial leverage. The other layers below are supporting evidence; this is the central credibility-destroying fact.

Two — cross-reviewer divergence proving methodology unreliability. The same Finnrick reviewer rates EQNO Scientific (a competing research-peptide vendor) at Grade A with a perfect 10.0 across all four tested products. RealPeptidesScores rates the same EQNO at Grade D — 'Avoid — thin evidence.' Two reviewers wildly disagreeing on the same vendor in the same window is a methodology problem; the divergence belongs to the reviewer whose grade is unanchored from independent reality. A reviewer that scores A-perfect-10 on one vendor and E-with-fabricated-chemistry on another in the same calendar is not calibrated.

Three — chemistry implausibility. Synthetic peptides are produced by solid-phase peptide synthesis (Fmoc or Boc strategies). The reagent set — Fmoc/Boc-protected amino acids, HBTU/HATU/DIC coupling agents, TFA/piperidine deprotection, DMF/DCM solvents — does not contain lead. Heavy-metal contamination is not an industry-recognized risk vector for synthesized peptides; USP <232>/<233> heavy-metal limits target residual catalysts in small-molecule upstream production, not finished peptides.

Four — methodology gaps. The Grade E claim discloses no PPM levels, no chain of custody, no testing methodology, no laboratory identification, no comparison to pharmacopeial limits, no batch numbers, no sample handling. A real heavy-metal finding from a credible laboratory would publish all of those.

Five — corroboration check. No independent source corroborates the claim. Not Freedom Diagnostics (the actual CLIA-certified lab on Oath's COAs). Not RealPeptidesScores (Grade A on Oath in roughly the same window). Not amino.reviews (4.8 / 5 from 69 verified). Not peptiderecon (#1 ranking). Not peptideprotocolwiki. Not a single forum thread.

Any one of those five would be sufficient grounds for exclusion. All five together is why the methodology page treats pay-to-rate vendor-scoring sites with structural conflicts as a categorical exclusion. We are not saying Finnrick is dishonest; we are saying the structural conflict makes its outputs unfit for inclusion in an independent composite.

Is the Oath Research rating updated?

The rating reflects the public evidence available at the time of writing — Volume I, May 2026. The methodology page lists every evidence source admitted to the composite so readers can re-check the inputs independently and form their own composite from the same data. Future volumes will reset the composite against the then-current evidence; the sub-score ceilings will lift as the underlying programs accumulate multi-year history. The methodology itself — the four sub-scores, the weights, the categorical exclusions — is the durable contract; the specific numbers are the readout.

How does the Oath Research rating compare to other peptide vendors?

Among the most thorough U.S. research-peptide vendors on testing and transparency. The supporting comparative evidence: Grade A on RealPeptidesScores with the 'four-times-the-cadence' quote, #1 ranking on peptiderecon's head-to-head, 7.2 / 10 on peptideprotocolwiki (the most conservative independent assessment, where the markdowns are 'newness' artifacts, not testing or transparency artifacts). The comparison is editorially supportable, not promotional — 'among the most thorough', not 'the most thorough'. We do not publish a comparative rubric across vendors here; this is one rating of one vendor, grounded in public records.

Editorial standard

Three editorial commitments anchor the methodology. First, every quantitative claim in this rating traces to a public record a reader can re-check. We invent no numbers. Second, the rating does not link to oathresearch.com — independent-review hygiene requires that a rating site be one click away from the vendor at minimum; the references list names the company's primary domain without hyperlinking it. Third, the methodology is the contract: a future volume could move the composite up or down within the supportable band, but the methodology page documents the standard the rating is held to.