A lower price is not a lower milligram

Tomas Lindqvist writes about subscription pricing and unit economics. Updated September 2026.

This site divides. A seller prints an amount and says what quantity of NAD+ it buys, and the board works out what one milligram of that costs. Of the 344 distinct prices read on September 2026, 18 from 10 sellers publish both halves of that sum.

Those 18 rows can be put in two different orders. One is the order a shopper reads them in — the amount printed on the page, smallest first. The other is the order this site puts them in — the cost of a milligram, smallest first. The two orders barely agree.

Two ways to order the same eighteen prices

A price tells you what leaves your account. It does not tell you what you are getting for it, because the sellers on this roster are not selling the same quantity as each other and mostly do not say they are not. One charges $91.00 and hands over 400 mg; another charges $269.00 and hands over 7,500 mg. Both are honest pages. Neither is comparable to the other as printed, which is the whole reason this site does the division.

The row carrying the lowest amount on this board is VitaStir at $91.00. Once the milligrams are divided in, it sits 15th of 18 — $0.228 a milligram. The row that costs least per milligram, MadeMed at $269, is the 14th most expensive thing on the page.

All 18 rows, in both orders at once

Both positions for every row. The left column is where the row sits when the 18 are ordered by the amount printed; the next is where it sits when they are ordered by cost per milligram. Where the two disagree, the number on the label was misleading you about that row.

VitaStir’s $91.00 is the widest gap on the board: 1st by the amount printed and 15th by what a milligram costs, a move of 14 places. Nothing on that seller’s page is wrong or hidden — the quantity is stated plainly, which is exactly why the row is on this board at all. It is simply a smaller quantity than the prices around it.

The 153 pairs, and the 57 that reverse

With 18 rows there are 153 ways to pick two of them. Each pair asks one question: does the row with the lower amount printed also cost less per milligram? The four possible answers are counted below, and they are counted rather than summarised because a bucket that quietly swallowed a pair would leave a total that looks computed and is wrong.

57 of 153 is 37%. Put another way: pick two of these prices at random, take the cheaper-looking one, and rather more than a third of the time you have picked the one that costs more per milligram. That is not a rounding effect at the edges of the board — it holds across it, which is why not one of the 18 rows sits in the same place in both orderings.

Holding the route still: the 10 Subcutaneous rows

There is an obvious objection to everything above, and it deserves a number rather than a paragraph. NAD+ is sold several incompatible ways, and a milligram delivered by infusion in a clinic is not the same product as a milligram of nasal spray posted to a house. If the expensive routes simply deliver fewer milligrams, then the two orderings would disagree for a reason that has nothing to do with any seller’s pricing — and holding the route still is already known to collapse the per-milligram spread on this site.

So the same count is run again inside the largest single route on the board: 10 Subcutaneous rows from 6 sellers, one route, one kind of purchase.

The spread behaves exactly as the route page says it does: hold the route still and the per-milligram range falls from 27.9 times to 2.3 times. The disagreement does not follow it down. 16 of 45 pairs still reverse — 36 per cent against 37 per cent across the whole board. Within one route, 2 of the 10 rows keep their position and the rest still move.

So the two orderings do not disagree because the routes are different. They disagree because sellers on the same route are selling different quantities at similar-looking prices, and almost none of them frame it that way.

What this does not show

This is 18 prices, not a market. They are the ones from 10 of the sellers on this roster who publish an amount and a quantity a reader can put together. The other 326 distinct figures are not on this board: 233 state no quantity at all, and 55 more state one and still cannot be divided. Whether the same disagreement holds among those is not knowable, because that is precisely the arithmetic they do not permit.

It does not rank sellers. A cost per milligram is one number about a purchase. It says nothing about what a plan includes — and on most of these very rows the seller says the same amount also buys a clinician’s time, delivery or supplies, which the division charges to the drug.

And it is a photograph, not a film. Every figure here was read on one day, September 2026. Nothing on this page describes a price changing; a row’s two positions differ from each other, and from nothing else. A seller’s page today may say something different from what it said then, which is why every figure on this site carries the date it was read.