Finding · criterion S3

Which pharmacies fill the orders, and how many sellers name the same one?

Recorded August 2026 · Standard v1.1

15 of the 47 records in the ledger name a filling pharmacy. Between them they name 27 facilities, and 9 of those 27 are named by more than one seller. The most widely named is Red Rock Pharmacy, which 7 separate records name.

Criterion S3, pharmacy identification asks a seller to name the facility that fills the order, not the category it belongs to. This piece takes the records that answer and asks the next question: how many different facilities does the field actually use. A shopper comparing two storefronts is comparing two brands. Whether they are also comparing two supply chains is not visible from either page, and it is visible here only because both records name the facility.

The count, with its denominator

Of the 47 records, 15 pass S3 and can therefore name a facility, 30 fail it, and 2 carry a null excluded from both counts, leaving an adjusted denominator of 45. Every one of the 15 passing records has its named facilities recorded here, so the table below accounts for all of them.

Every facility, and who names it

One row per facility, most widely named first. The spellings column prints every form the ledger's own bases use, because that is what the count had to reconcile.

Where storefronts meet

9 facilities are named by more than one record, and 11 of the 15 naming records name at least one facility another record also names: altRx, Care Bare Rx, Fella Health, Hims & Hers, Leanmeds, MangoRx, Nextmeds, Shed, SnagRx and Embody, Synergy Rx, Trinity Meds. The overlap is not evidence that those sellers are related. Contract pharmacies serve many clients, and a facility appearing on several records is the ordinary shape of that market. What it does mean is narrower and still worth knowing: choosing a different storefront is not always choosing a different pharmacy.

The ledger records 47 companies behind 59 storefront domains, so a reader already has two layers to see through before reaching a facility. This census is the third layer. None of the three is visible from a product page.

18 facilities are named by exactly one record. A facility named once here is not rarer in the world than one named by many; it is rarer in this ledger, which holds 47 records and no claim to be a census of the market.

Why the spellings are printed

Red Rock Pharmacy appears in this ledger under 4 spellings, each copied from a different seller's own page: “Red Rock Compounding Pharmacy”, “Red Rock Pharmacy”, “RedRock Pharmacy”, “Red Rock Home Pharmacy”. A count keyed on any one of those strings would report several facilities where there is one. A pattern loose enough to catch every one of them would also catch a company that merely resembles them, which is the error that attaches one business's record to another. So the assignment is declared on each record by hand, and the build checks in both directions: every facility a record declares must appear, in one of its recorded spellings, in that record's own S3 basis.

The same discipline refuses 4 merges that look obvious:

Leaving those separate makes the facility count higher than a confident reading would produce. That is the correct direction to be wrong in. An over-merged table would assert that two companies are one, and nothing on the page would show a reader the assertion had been made.

One merge was made over a disagreement in the evidence rather than in spite of one, and the disagreement is printed here rather than resolved:

What this census does not establish

It does not establish that a named facility compounded any particular vial. A seller names the pharmacies in its network, and which one fills a given order is not published by any record here. It does not establish that the facilities are licensed, in good standing, or inspected: S3 asks for a name a reader can take to a state board register, and looking it up is the reader's step, not this site's. It ranks nothing. A record naming several facilities is not thereby better than a record naming one, and the table is ordered by how many sellers name a facility for one reason only, which is that the question is how far the overlap goes.

By kind, 15 of the 30 telehealth records name a facility and 0 of the 17 research vendors do. The vendors fail S3 on different ground from the telehealth services, and not all on the same ground as each other: some state that no pharmacy is involved, some record only that none is named on the pages read, and some disclaim fulfillment to third parties they do not name. The criterion page carries every basis in full, and the difference matters when reading the number.

The full roster

Every record in the ledger with its S3 disposition and the basis recorded against it, so the count above can be repeated from the page.

Sources

Each source below is the page a passing record cites for its S3 disposition, read on the date that record carries.

• altRx, the page its S3 disposition cites
• Care Bare Rx, the page its S3 disposition cites
• Eden, the page its S3 disposition cites
• Fella Health, the page its S3 disposition cites
• He & She MD, the page its S3 disposition cites
• Hims & Hers, the page its S3 disposition cites
• Leanmeds, the page its S3 disposition cites
• MangoRx, the page its S3 disposition cites
• Nextmeds, the page its S3 disposition cites
• OrderlyMeds, the page its S3 disposition cites
• Ro, the page its S3 disposition cites
• Shed, the page its S3 disposition cites
• SnagRx and Embody, the page its S3 disposition cites
• Synergy Rx, the page its S3 disposition cites
• Trinity Meds, the page its S3 disposition cites

Facility assignments are declared per record and reconciled against each record's own S3 basis at build time, against Standard v1.1, in the census run stated at the head of this page. Records are counted once each; a record filed under several storefronts is one record. See S6 for where the sellers themselves can be found.

Counts are generated from the ledger at build time, so this page cannot drift from the records it summarizes. Method and null semantics: how checks are performed. This page ranks nothing, recommends nothing, and carries no partner link.

Related findings

All findings →