Skip to content
This GLP
← Research
Formats

When the headline is not the bill: a census of our own corrections

Most corrections this desk writes exist for one reason — the advertised figure is not the billed one. A few exist because no figure could be published at all.

Glenn Torres5 min read
What happened to 249 corrected price rows187 restated — the billed figure is higher50 withdrawn — no figure could be published12 corrected a field other than the priceAcross 117 sellers.

The single most common thing wrong with a GLP-1 price is that it is not the price. Across the sellers this desk has worked through, 117 carry at least one correction, over 249 separate price rows — and the largest group by far, 187 of them, exists because the figure a seller advertises is not the figure a buyer is billed. Before comparing anything on a board, it is worth knowing what kinds of gap keep appearing.

Three different outcomes

A correction is arithmetic on what a seller publishes, and it has three possible endings. The commonest, 187 rows or 75% of the corrected set, restates the figure upward: a mandatory membership added, a four-week cycle converted to a calendar month, a prepay rung separated from the monthly rate it was standing in for.

The second, 50 rows, withdraws the price entirely. Those are sellers whose published figures could not be resolved into anything a buyer could be held to — two pages giving one number different meanings, a savings line printed where a price goes, a default carried by a shared template. Publishing any of those would have put a number on a board that nobody is charged.

The third and smallest, 12 rows, leaves the price alone and corrects something else — a product recorded as an injection that is an oral, a compounded row that is actually brand. Those change what a figure describes without changing the figure.

What this is not a rate of

This counts corrections written, not sellers audited. A row enters the count when somebody read the seller, found the gap and did the arithmetic, so the denominator is the worked set rather than the 475 sellers on the roster. It would be easy and wrong to turn these into a percentage of the market.

What it does establish is the shape of the problem. The gaps are not random errors — they cluster into a small number of repeated structures, and once a reader can recognize those structures they can check for them on any page, including sellers nobody has corrected yet.

The four that keep recurring

A billing cycle that is four weeks rather than a month, which adds a thirteenth charge a year — counted separately in a month that is not a month. A required fee outside the medication price. A first-month or promotional figure printed where the standing rate belongs. And a prepay rung presented as though it were the monthly rate.

All four are checkable before paying, and all four are invisible if you read only the number. The questions that surface them are in the questions they leave open, and the broader census of what goes unpublished is in what sellers will not tell you.

Every figure above is computed from this site’s own correction records at the moment the page was built, most recently read September 2026.

Frequently asked

How often is the advertised price wrong?
This counts corrections written rather than sellers audited, so it is not a rate. What it shows is that the largest group of corrections by far exists because the advertised figure is not the billed one.
Why would a price be withdrawn entirely?
Because nothing publishable could be resolved — two pages giving one number different meanings, a savings line printed where a price goes, or a default carried by a shared template.
What should I check on a seller's page?
The billing interval, any required fee outside the medication, whether the figure is a first month or a promotion, and whether it belongs to a prepaid term rather than a month.

Sources

    Where to get it

    Best GLP-1 injections

    Every injectable seller we can verify, with the price each one publishes and an honest read of what the trials measured.

    Compare providers →

    More in Formats