Trust architecture

Methodology

Every number on this site is generated from one dataset under one set of rules. This page is those rules — specific enough that you could rebuild our rankings yourself from the downloads and get the same answers.

§1 · The basis

Every program is normalized to all-in monthly cost: the medication price plus every mandatory recurring fee — membership, subscription, platform, or program fees — with free-shipping and included-consultation policies noted and any per-shipment or per-visit charges added. We record this figure at two points: the advertised start (what month one costs) and the maintenance dose, anchored at 10 mg weekly for tirzepatide and 2.4 mg weekly for semaglutide — the most common long-run doses in the pivotal trials and the doses where tiered programs stop being cheap. Rankings use the maintenance figure only.

§2 · Twelve-month totals

The twelve-month total is the best available basis times twelve, including twelve cycles of any membership: the prepaid twelve-month plan rate where one exists, otherwise the monthly maintenance figure. We do not model titration-month savings into the total — real first years cost slightly less at tiered providers — because a conservative, uniform basis beats a flattering, assumption-heavy one.

§3 · Verification statuses

Every figure carries one of five statuses. Verified means our desk captured the figure on the provider's own page or checkout on the stated date. Provider-stated means it comes from the provider's published pages with our checkout capture pending. Operator-stated means the figure was supplied directly to our desk — by the provider or by the site's operator — with supporting evidence (rate cards, checkout screenshots), and our independent capture is pending — this status exists so supplied information can be published quickly without being dressed up as independently confirmed. Reported means dated third-party reporting is the best current source. Unrankable means the record can't be normalized. Statuses are displayed on every figure, never hidden in a footnote.

§4 · Exclusions

Three kinds of prices never enter a ranking. Start-only prices — a program that publishes no maintenance figure (Fifty 410's $133 is the current example) cannot be ranked, because month one is not the program. Unlabeled promotions — promotional rates are ranked only when labeled as promotional, with the standard rate stored beside them; an introductory price that applies to a fixed number of fills (like NovoCare's $199 two-fill offer) is recorded but never used as the program's normalized figure. Unverifiable outliers — figures that can't be corroborated against any live page are removed rather than published, and the removal is logged in corrections.

§5 · How winners are chosen

They aren't chosen; they're computed. At every site build, the generator takes the eligible rows for a category and applies the rule — minimum all-in monthly for "cheapest," minimum twelve-month total for annual categories, filtered subsets for "no membership" or "flat-rate." No editor assigns a priced label, which means no editor can be lobbied for one. The two standing qualitative labels (insurance navigation, pricing transparency) are argued in prose on the page that carries them, with the trade-offs attached.

§6 · Source hierarchy and cadence

First-party checkout capture outranks first-party marketing pages, which outrank manufacturer announcements, which outrank dated reporting from named outlets, which outrank everything else; anonymous aggregators and undated pages are not sources. Compounded-program rows are re-checked on a weekly cadence and brand-direct rows against manufacturer pages on publication of any pricing announcement; every row shows its capture date so staleness is visible rather than hidden.

§7 · The regulatory frame we apply

We describe compounded GLP-1s exactly as the law does: not FDA-approved, not reviewed for safety, effectiveness, or quality before dispensing, lawful to compound only within the Section 503A/503B framework — which, after the FDA resolved the tirzepatide shortage in December 2024 and the semaglutide shortage in February 2025, no longer permits copies of the approved products outside documented patient-specific need. Programs are listed because they exist and people are buying from them; listing is not endorsement, and the safety checklist in our guides applies to every listing including the cheapest.

§8 · What this site does not do

No priced ranking can be bought, no promotional rate is silently substituted for a standard rate, no start price is compared against a competitor's maintenance price, no review or star rating is invented, and no medical advice is given. Where our own verification hasn't caught up to a claim, the status says so on the claim itself.

The Value Score formula (the sticky bar's math)

The site-wide "Top picks" bar ranks programs by a published, uniformly computed Value Score out of one hundred — printed here so the bar can never be an editorial thumb on the scale. The formula, applied identically to every rankable program: seventy points scaled by price (the category's cheapest all-in monthly figure divided by the program's, times seventy — the cheapest rankable program earns the full seventy; a program at double the floor earns thirty-five); fifteen points scaled by evidentiary status (verified 100%, provider-stated 85%, reported 70%, operator-stated 60% — the bar literally scores our own confidence against us, and today's leader loses points here); ten points for a dose-flat policy; five points for carrying no membership fee. Promotional prices count as the price a buyer pays today and are labeled "promo" in the bar; programs without a rankable maintenance figure receive no score and never appear. The score recomputes at every build from the same database every other page draws from, and the disclosure page documents why a public formula matters doubly on this site.