Embody Review

What we couldn't verify

• No BMI threshold or minimum age is published. The intake page returns 101 characters to a raw fetch, and the eligibility questionnaire is not in the served HTML. The skill's intake-flow walkthrough would reach it, but that requires a rendered fetch, and joinem.co is recorded stance 'refused' — the sanctioned exception covers raw retrieval only.
• No dose, strength, vial size or titration schedule is published for either medication.
• The list of states served is not published; Embody directs prospective patients to email support.
• No pharmacy-classification detail is published for any of the four named dispensing pharmacies.
• No consumer-sentiment data. Trustpilot is recorded stance 'refused' with robots.txt 'Disallow: /' and terms banning automated access; no rating or count was retrieved and none is published. This is our constraint, not brand opacity — the factor is excluded, not scored low.
• Third-party accreditation status was not checked against the registries this run. No accreditation is claimed on the site, so nothing is being repeated unverified, but the directories were not queried.
• The struck-through comparison prices of $299 and $349 on the one-month product pages are not substantiated anywhere on the site.
• NAD+, sermorelin, glutathione and MIC/B12 intake pages exist but carry no product, price or clinical detail. They are recorded as discovered offerings and are NOT scored or reviewed this run — the research to score them was not done, and a thin score is worse than none.
• The CRM record lists ED as a sub-vertical for this advertiser. No ED product, intake page or mention appears anywhere in the site's own sitemap or on any retrieved page. Recorded as a CRM/site discrepancy, not as a vertical.
• The absence of any entry on the FDA's enforcement register and of federal litigation is recorded as METHOD in the relevant observations' notes, not as a citation. A search URL is a query, not a record — `classifyFdaCitation` refuses to classify one, and it is right to: citing a search result as a primary source would let "I looked and found nothing" render as a cited fact. The federal-docket check ran on 2026-09-21 against the CourtListener REST API v4 (the only sanctioned route — courtlistener.com is refused on terms), with the positive control "Mochi Health" returning 15 records, so a zero result from this run is a measurement and not a failed search. The legal entity "Modern Metabolic Medicine, Inc" returns 0 records — a measured absence. The brand-name query "Embody" returns 24,378 records, but no case caption contains the string: "embody" is an ordinary English word matching case body text, so that query supports no claim in either direction and the absence rests on the legal-entity query alone. This check did NOT run before this review was first published; it was added afterwards.
• The drug-label and regulatory records attached to the safety, dosing and approval claims (DailyMed WEGOVY, Drugs@FDA NDA 217806, the FDA compounded-GLP-1 action) are REUSED from records the corpus already holds for the same drugs — they carry no quote, because their text was not retrieved in this run. The citation is to the record; the wording of the record is not restated.
• The paid-search and affiliate-placement figures come from internal commercial records (the WebMD term analysis and the RX position log), which are not public sources and belong to no citation tier. They are recorded as METHOD in the observations' notes, not as citations — a file path on disk is provenance for a figure, not a source a reader can check, and rendering it as one would put an uncheckable citation on a published page.