A price on a GLP-1 landing page is a designed object. It has a term attached to it, a condition under it, a product behind it and a billing cycle none of those three necessarily share. Six presentation devices recur often enough across the sellers written up here to be counted, and none of them requires anything to be concealed — each is disclosed somewhere, in smaller type, on a different page, or in a footnote that no longer matches the table above it.
One: the from-price attached to no product
The most common device by a wide margin is also the least remarked on. 114 of 257 written-up sellers lead with “from” or “starting at”, and 143 publish a floor figure of some kind. A floor is only comparable if the thing it floors is identified, and frequently it is not: 473 sellers here publish an injection price and 91 publish an oral one, several publish a microdose line beside a standard line, and a single “from $99” can belong to any of them. The qualifier a buyer needs — which molecule, which dose, which plan length — is often simply absent rather than hidden. Two sellers with identical floors can be selling entirely different things, which is one reason prices in this market vary so much.
Two: the first-month-only figure
75 of 257 — 29.2% — carry an introductory, first-fill or new-customer figure. These take several forms: a flat dollar discount on month one, a promotional code applied once, a first-month rate that reverts, or a starter dose priced below the maintenance dose it leads to. The device is honest in structure — a month-one price genuinely is the month-one price — and misleading in placement, because a comparison surface reads the largest number on the card and a buyer stays for longer than one month. Anything sorted on a first-month figure is sorted on a rate the same seller will not be charging by the second refill.
Three: the four-week cycle labeled per month
39 of 257 bill on a four-week or weekly cadence while labeling the result a monthly rate. Thirteen four-week fills land in a calendar year rather than twelve, so the annual total is one fill higher than the label implies, and a plan described as “billed and shipped approximately every four weeks” is a different product from a plan billed on the first of the month. This is the device with the strongest claim to being an artifact of operations rather than marketing: shipping cycles are built on weeks because dosing is weekly.
Four: the struck-through anchor
33 of 257 — 12.8% — present a higher figure struck through beside the live one, or a discount code that reduces a posted rate. The anchor itself is the variable worth examining, because the struck-through number is a claim about a price the seller says it would otherwise charge, and in several cases here the crossed-out figure is the standing rate and the visible one is a one-time code applied to a first order only. A strike-through combined with a one-time code is two devices stacked, and the resulting figure is neither the standing price nor a recurring discount.
Five: the prepay rung shown as a monthly rate
109 of 257 mention prepayment somewhere, and in 22 of them the figure printed on the hero or the product card is itself a prepay rate rather than what a single month costs. The shapes recorded include a three-month block divided by three, a twelve-month block divided by twelve, and an asterisk under the pricing table stating that the price shown applies to the annual plan paid up front. Because the arithmetic is genuine division, the resulting number is internally correct and externally incomparable: it is a real monthly figure for a buyer who has already paid a year. What each rung is worth in dollars, and what a month-to-month buyer pays instead, is the subject of the prepay and membership article.
Six: the price that renders nowhere
The last device is the absence of one. 49 of the 1,040 published price rows here end with no computable figure at all, across 22 sellers, of which 12 finish with no payable row whatsoever. The usual shape is a seller that publishes a medication price and also states that enrollment in a program, a membership or a consultation is required, without publishing what the required item costs. Every number on that page is then conditional on an unpublished one, and the honest output is a blank rather than the unconditional figure.
A related shape appears in 37 of 257 write-ups: two figures for the same product that disagree with each other on the seller’s own site. A homepage carousel and a medication table quoting different numbers, a bullet list saying a consultation is free while the FAQ below it states a fee, a pricing page that no longer matches its own footnote. Where a seller contradicts itself the useful output is neither figure, which is why the methodology treats a contradiction as a finding rather than as a gap to be resolved by picking the lower number.
Why an automatic filter cannot catch any of this
Here is what should complicate any faith in a screening rule. Of the 1,040 price rows here, exactly 27 — 2.6% — are flagged as implausible standing monthly rates, and every one of them is a tail. Seven are flagged as promotional, and all seven are advertised at $25. Twenty are flagged as non-monthly packages, and every one of those falls between $629 and $1,695. A rule that asks “is this figure plausible as a month” finds the $25 teaser and the four-figure quarterly bundle and nothing else.
Every device in the taxonomy above is invisible to that rule by construction. A $99 twelve-month prepay rung occupies exactly the position a $99 standing monthly rate occupies. A four-week cycle at $175 looks like a calendar month at $175. A first-month rate of $149 is indistinguishable in a column from a permanent $149. The devices that move the most money produce the most ordinary-looking numbers, which is why a price table has to be read against the seller’s own terms rather than against the distribution — and why a per-seller price check answers a question a filter cannot.
The devices stack, and stacking is where the arithmetic goes
None of the six is exclusive, and a single card can carry three at once: a from-price, struck through against a higher anchor, applied to the first order only, on a plan billed every four weeks. Each layer is individually disclosed and individually small, and the compound effect is a figure that describes no charge the buyer will ever see. The stacking is also what makes a spot check unreliable — confirming that the four-week note is accurate does nothing about the prepay condition sitting under it.
Which is why the useful reading protocol is subtraction rather than verification. Rather than checking whether the figure on the card is true, establish what it is a figure for: which molecule and route, at how many milligrams, for how many days, under what plan length, in which month of the plan, with what else billed alongside. A figure that survives all six questions is a price. A figure that cannot answer them is a design element, and no amount of confirming the individual footnotes turns it into one.
The same pages misdescribe the product, too
Price is not the only field under pressure. A cross-sectional study of 93 business websites advertising compounded GLP-1 products for weight loss across Colorado, searched between March 21 and April 12, 2024, found that 41 of 93 referred to FDA approval in their descriptions of compounded products and 5 described them as “generic”.[1] Seven advertised oral formulations and eight advertised combinations with B vitamins. Neither of those regulatory descriptions is accurate: compounded drugs are not FDA-approved and are not reviewed by the FDA for safety, efficacy or quality before dispensing, and no generic semaglutide or tirzepatide exists in the United States. What that term actually means is set out in the approval article.
What this does not establish
Coverage is not census, and none of these counts should be read as a rate for the whole market. Of the 477 sellers here, 257 have a write-up; a device absent from a write-up may be absent from the seller’s site or may simply not have been examined. The counts above are floors on how often each device occurs, not measurements of how often it does not.
And nothing here supports a claim about anybody’s intent. Every one of the six devices is consistent with a stale footnote, an inherited page template, or a marketing team that never saw the billing system — and in several cases the qualifying condition is printed directly under the price, which is the opposite of concealment. The useful conclusion is about the reader rather than the seller: a figure on a card is not a price until the term, the product and the billing cycle attached to it are all known, and a board that ranks on the figure alone is ranking on a design decision. The sellers holding one rate across every dose are collected on the flat-pricing board.