# Methodology and limitations

## Inputs and substantive review
The baseline is the immutable 295-record GB-coded database export of 14 September 2026, not a fresh measurement of a whole national market. The earlier phase assigned 60 priority records; the expansion assigned the other 235. Every original ID now has one recorded outcome, including blocked/unattempted and unresolved cases. The seven additional flagship candidates are a separate purposive supplement, not replacements that retroactively verify inaccessible original supplier listings.

All worker-marked usable candidates were reviewed against retained primary own-product HTML, selected-option payloads and full descriptions; earlier included records were reviewed again. Independent reviewers read the surrounding material rather than approving merely because a hash, price substring or HTTP 200 existed. Direct review decisions include literal primary excerpts and file hashes in the private audit trail. The public table contains factual extracted observations and caveats, not those full supplier documents. The source_notes column states the final review finding, not the acquisition worker's provisional verdict; original notes remain unchanged in private source ledgers. Review caveats remain intact. Three further own-product molecular-reference conflicts identified at final consistency review were excluded rather than guessed to be rounding or peptide-versus-salt conventions.

## What can be included
An observation needs a coherent named product/selected nominal-size option, positive finite displayed GBP price, positive unambiguous labelled quantity, product-bound supplier in-stock claim and coherent tax classification. Price ranges, unrelated products, upsells, coupons and crossed-out original prices do not substitute for the selected one-unit offer. Explicitly labelled assay results are not nominal-fill measurements. No chemical-name, plus-sign or positive sub-mg blacklist is applied as a sampling rule; the frozen original export's earlier at-least-1mg selection still limits that cohort.

Unresolved own-product quantity, molecular-identity or explicit same-context tax contradictions stay excluded. Unreadable/blocked sources and unknown stock cannot be promoted. An exclusion is not a fraud, closure, out-of-stock, wrong-price or supplier-quality finding and does not remove a live website record.

The narrower inclusion claim is a named product and nominal size, not exact SKU/batch or chemical verification. Clear SKU/batch disagreement must remain visible. Separate bulk options, other product links, product/shipping weight and assay measurements are not automatically competing nominal sizes: their own context must be read. Molecular reference tables are not silently repaired or converted between salts, counter-ions, hydration, fragments, DAC modifiers or PEG forms.

## Prices, taxes and comparability
Prices are supplier-displayed item observations at the per-row capture time. Shipping and checkout totals are not calculated or verified. Included/excluded/unknown tax flags remain distinct; no VAT rate, gross-up or all-in price is invented. Generic tax-at-checkout wording plus conditional VAT terms need not prove that tax is added to the displayed price: a coherent unknown classification is permitted. The included semaglutide row preserves this uncertainty and its unrelated unverified assay/marketing discrepancy.

There are **no pooled price statistics** in this packet. No price averages, medians, minima/maxima, market shares or supplier rankings. GBP per nominal mg is an individual arithmetic ratio, not a scientifically harmonised or like-for-like value comparison. TB-500 fragment/full-length/unspecified labels are not pooled. CJC DAC/no-DAC/unspecified labels, kisspeptin forms, salts and other catalog families are not declared chemically interchangeable. NAD+ is retained as an original catalog compound, not called a peptide. Supplier FOX-04/FOX04 spelling is preserved rather than silently corrected.

## Arithmetic
Price and nominal quantity are parsed as exact decimals and converted to an exact rational ratio. The numerator and denominator are exported. The displayed GBP-per-nominal-mg figure is rounded half-up to two decimals using integer rational arithmetic, with no binary floating-point rounding and no rounding-before-aggregation (there is no aggregation). Labelled mass is not assay-adjusted, salt-adjusted or guaranteed net active content. Historical rounding reconciliations concern the earlier database snapshot only.

## Geography, timing and evidence preservation
GB is the original database country code, not independent proof of headquarters, dispatch origin or UK availability. Supplementary UK-facing rows have first-party delivery/location evidence but no independently verified legal domicile. Current source dates are not backfilled into database timestamps. No data was reacquired during this saved-source review and no source retrieval timestamp was refreshed. Known earlier host/access errors remain honestly identified, not fabricated as new retrievals.

Every source file frozen for this run must match its recorded SHA-256 before assembly. Candidate/review IDs must reconcile exactly, all required reviews must be complete, all inclusion checks must be explicit booleans, and all decisive quotes must exist in their hash-matched evidence. Those integrity checks support provenance; they do not replace the substantive reading or certify supplier truthfulness.

## Commercial disclosure and rights
Prepared for Peptide Scanner, the commercial platform operating the index. Its first-party disclaimer captured on 14 September 2026 stated that some outbound links are affiliate links and it may receive referral fees; it also stated that those relationships do not influence rankings, data or editorial content. This packet is not an institutional audit of the business or its commercial arrangements. Product-source links omit known affiliate/tracking query parameters while retaining functional selection parameters.

Captured policy references: https://peptidescanner.co/disclaimer/ and https://peptidescanner.co/review-policy/ . No supplier or press was contacted, no inbox or checkout was tested, and no press desk is designated here.

Cite Peptide Scanner, the dated sample, the relevant source URL and per-row retrieval time. Keep methodology and limitations beside any chart. These are factual extracts and original figures; full supplier pages, photos, reviews, logos and COAs are not licensed or distributed by this packet. No new blanket open-data licence or third-party rights clearance is asserted. Public deployment was not performed by this local preparation task.
