How We Rank: Our Methodology
A rating here is a single number out of 10, and it is the weighted average of the six listed below. The weights are fixed. They are the same for a company that pays us and a company that has never heard of us, and none of the six inputs has a commercial term anywhere in it. Money does reach this site, and where it lands is set out under Scores vs. Placement further down. A register that prints only the first half of that is not telling you much.
A score rests on what the seller publishes and we could read, and a gap is never filled in from guesswork. A seller that shows no comparable monthly price has its Value held at 3 rather than scored against the others. One that publishes no state list has its Accessibility held at the midpoint and misses the Trust point that publishing earns. One that does not name its pharmacy misses the Trust points for naming one. That is not a penalty for being secretive so much as a refusal to certify something we could not see, and it applies to a seller paying us the same way it applies to one that has never contacted us.
Two carve-outs matter. Effectiveness moves only on the clinical evidence for the drugs being sold, so a seller cannot lift it by being easier to audit — the trials are the trials whoever is dispensing. And Trust moves only on things we can document: FDA warning letters, whether the pharmacy is named, whether the price comes before the intake, whether a state list is published, and how much of the record we could corroborate. Star ratings on review platforms and BBB letter grades are not inputs, because we cannot check how they were collected.
How each number is computed
No one here assigns a seller a score out of ten. Each of the six is a function of things the seller has published, or has declined to publish, and the register recomputes all of them every time it builds. That is the point: you should be able to arrive at our number yourself, and tell us when you cannot.
| Dimension | What it is a function of |
|---|---|
| Value | The seller’s cheapest standing monthly price for a compounded GLP-1, against the median for that drug and form across the whole register. Half the median scores 10, the median scores 5, double scores about 1. Membership fees and first-month rates are excluded from the comparison; a seller publishing no price is capped rather than scored, because that is a fact about disclosure and not about value. |
| Effectiveness | Which molecules the seller will actually dispense, ranked by the trial evidence behind them. It is a property of the drug, not the vendor, so two sellers shipping the same compounded tirzepatide score the same. Nobody earns it by being easy to audit. |
| Trust | FDA warning letters on record, whether the dispensing pharmacy is named, whether a price appears before the medical intake, whether a state list is published, and our own verification confidence. Opacity costs points here because in this market it is the reliable predictor of an unpleasant surprise. |
| Accessibility | States of the 50 the seller publishes coverage in. Publishing no list is held at the midpoint, never zero — we will not assert that a company serves nobody because it did not tell us where it operates. |
| Support | What the monthly figure actually covers: medication, the consultation, shipping, lab work, coaching. A seller who states too little for the bundle to be assessed is held at the midpoint. |
| Experience | Read off the seller’s own pages, not from how the site looks. The heaviest term is whether a price is visible before you hand over health information. A required video visit or baseline bloodwork costs half a point each, because a slower intake is often the more careful one. Where we do not know when the price first appears, this is held at the midpoint. |
Every review page shows the inputs behind its own six numbers, so the arithmetic is on the page rather than in a policy. Nothing commercial is an input to any of them: not whether a seller pays us, not which network they run on, not what they pay. Where money does change what you see is set out below, and it is placement, never a score.
Value
25%The seller's cheapest standing monthly price for a GLP-1 (a compounded one where it sells one), against the median for the same drug and form across the sellers we compare. First-month rates and fees without the medication are left out
Effectiveness
25%Trial evidence for the molecules the seller actually prescribes. Its own outcome marketing counts for nothing here
User Experience
15%Where the price first appears, before or after the medical intake, and whether the intake requires a live video visit or baseline bloodwork
Trust & Safety
15%FDA warning letters, whether the seller names the pharmacy that fills its prescriptions, whether a price appears before the intake, whether it publishes a state list, and how much of its account of itself we could corroborate
Accessibility
10%How many of the 50 states the seller publishes coverage in. A seller that publishes no list is held at the midpoint
Support
10%What the monthly price covers: the medication, clinician visits, shipping, lab work and coaching
How We Evaluate Compounding Pharmacies
Whoever writes your prescription for a compounded GLP-1 did not make the vial. A compounding pharmacy did, under contract, and it is the party whose sterility record and licensing you are actually relying on. So we keep a second register for them. Our compounding pharmacy directory carries a rating built from five things: what the facility is accredited for (PCAB, cGMP, registration with FDA as a 503B), what regulators have put on its file (warning letters, Form 483 observations, recalls), how many states license it to dispense, whether it makes semaglutide and tirzepatide or only one of them, and how many years it has been running.
The two ratings never touch. Contracting with a well-rated pharmacy earns a telehealth company no points, and a poorly rated one costs it none. What we publish instead is the link between them, so you can weigh the prescriber and the facility filling the order on their own separate records.
Scores vs. Placement
Score is independent. Placement is not always. The same formula runs over the whole register, and a commercial relationship is not one of its inputs. Two companies backed by the same evidence land on the same number, whether one of them pays us or neither does.
Order is a different matter, and here is where a paying company gets something. Partners are grouped ahead of everyone else in the reviews directory and in the comparison tool. They lead the high-intent boards: the semaglutide, tirzepatide, compounded semaglutide and compounded tirzepatide lists, the injection and shot lists, fitness apps, and the ED, TRT and HRT boards. The homepage partner cards, the floating partner widget and the partner pop-up are drawn from the partner pool and led by what each partner earns us rather than by rank. The homepage picks must score 7.0 or above, which earnings alone do not buy past. A partner marker sits on the listing itself in the reviews directory, on the boards, and on each of those three surfaces. The comparison tool carries its disclosure once, at the page level, rather than per row.
Two boards are held out of that on purpose. The cheapest-semaglutide and cheapest-tirzepatide lists promise price order and nothing else, so putting partners at the head of them would be a lie the reader has no way to catch. They stay in strict price order. So does the full A–Z index, which is alphabetical and carries no commercial weighting at all. A company we earn nothing from is scored by the identical rubric and shows its full number wherever it appears. The revenue side is itemized on our Affiliate Disclosure, and walked through start to finish under Nature of Reviews & Rankings.
A rating moves when the evidence under it moves: a fresh price capture, a trial readout, a change in how a company treats the people who have already paid it. If a rating here looks stale or plain wrong to you, send us the figure and where you found it. A specific correction is worth far more to us than a general complaint, and we would rather hear it than not.
Who runs this register, and how it pays for itself, is on the About GLP Watchdog page.
Sources & methodology — as of September 2026
- 1.GLP Watchdog — GLP-1 Pricing Index 2026 (our independent dataset)— GLP Watchdog.
- 2.NovoCare — Wegovy cash-pay and savings program— Novo Nordisk (NovoCare).
- 3.LillyDirect — Zepbound self-pay program and cash price— Eli Lilly and Company (LillyDirect).
- 4.STEP 1 Trial — Once-Weekly Semaglutide in Adults with Overweight or Obesity (Wilding JPH et al.)— New England Journal of Medicine.PMID: 33567185.
- 5.SURMOUNT-1 Trial — Tirzepatide Once Weekly for the Treatment of Obesity (Jastreboff AM et al.)— New England Journal of Medicine.PMID: 35658024.
- 6.PCAB — Pharmacy Compounding Accreditation Board Standards— Accreditation Commission for Health Care (ACHC) / PCAB.
- 7.FDA — Compounding and the 503A Pharmacy Framework— U.S. Food & Drug Administration.
- 8.FDA — 503B Outsourcing Facility Registration and Requirements— U.S. Food & Drug Administration.
- 9.FDA — Warning Letters Regarding Compounded GLP-1 Drug Products— U.S. Food & Drug Administration.
- 10.AACE/ACE Comprehensive Clinical Practice Guidelines for Medical Care of Patients with Obesity (Garvey WT et al., 2016)— American Association of Clinical Endocrinology.PMID: 27219496.
- 11.ADA — Standards of Care in Diabetes (2025)— American Diabetes Association.