Segment 1 — The Cold Open

Walk a US grocery aisle and pick up a name-brand can of powdered infant formula. A typical tub — often somewhere around a pound and a half of beige powder — commonly sits in the mid-twenties to the high thirties of dollars, and specialty cans climb from there. Wave a magic wand, arrange the raw lactose, vegetable oil, and milk protein into that same powder for free, and the atoms themselves are worth a few dollars a kilogram. The rest of the price is not the physics of feeding a baby. It is what we have piled on top of a very ordinary set of nutrients.

Segment 2 — Why It Costs What It Costs Today

Infant formula is one of the most tightly specified foods on earth, and the tightness is not a fashion. A newborn may take in nothing else for months, so the recipe is treated as a medical specification rather than a grocery mix. Federal rules set floors and ceilings for protein, fat, linoleic acid, and a long list of vitamins and minerals in every hundred calories, and a plant that wants to sell a new powder must register, follow infant-formula manufacturing practice, test both nutrients and the dry-powder pathogens that have historically caused the scares, and wait out a notification process before the first can ships. After earlier failures of composition and quality, that legal structure was built so this food could not quietly drift, and inside the industry the resulting line — wet-blend the oils into a dairy stream, homogenize, heat-treat, spray-dry, nitrogen-flush a can, hold the lot for release testing — feels like the only adult way to do the job.

The commercial structure on top of those rules is what makes the shelf price feel normal. A handful of manufacturers have long occupied most of the American aisle. Pediatricians recommend the labels they saw in training. Hospitals send families home with a particular can in the discharge bag, and that first can is a switching cost no economist has to explain to a sleep-deprived parent. A large public nutrition program buys formula for a large share of US infants under sole-source state contracts: one brand wins, often by offering a deep rebate, and that brand becomes what clinics and endcaps treat as the default. Marketing, medical detailing, and the liability posture of a product that cannot be allowed to fail accrete on top of a spray-dried powder that is, at bottom, milk sugar, vegetable fat, and dairy protein. From inside the category none of this looks like a puzzle. It looks like caution. Caution plus concentration plus a buyer who will not experiment is how a commodity dairy blend becomes a pediatric premium.

Segment 3 — The Magic Wand Number & The Idiot Index

Start with a kilogram of finished cow’s-milk powder, because that is the unit the commodities are actually sold in. Read a typical label the way an engineer would. Something like half to a bit more than half the weight is carbohydrate, and in a standard formula that carbohydrate is almost entirely lactose. A little over a quarter is fat, almost always a blend of vegetable oils — palm, coconut, soy, sunflower, or the high-oleic cousins of those same oils. A little over a tenth is protein, as whey and casein from cow’s milk. Moisture, minerals, and a vitamin-and-lipid premix make up the rest, and they weigh almost nothing.

Now price those atoms as commodities, and flag every figure as a working estimate rather than a invoice. Lactose is a cheese-industry byproduct, not a rare earth. Industrial edible lactose commonly moves at roughly a dollar to a couple of dollars a kilogram, which puts the lactose inside one kilo of formula somewhere around fifty cents to about a dollar-fifty. The oil blend is grocery-store chemistry. Commodity vegetable oils often trade at roughly a dollar to a few dollars a kilogram, so a quarter-kilo of that fat is something like thirty cents to a dollar. Protein is the expensive macronutrient, and even here we are still in the dairy complex: whey protein concentrate and skim-milk solids, not a fermenter drug. Take a wide band — several dollars to a bit over ten dollars a kilogram of protein-rich dairy powder, depending on grade and the year — and the protein in one kilo of formula still lands nearer a dollar than a ten-dollar bill.

The micronutrient premix and the specialty long-chain lipids added for DHA and ARA punch above their weight. They are a tiny mass fraction and a real purchase-order line, which is why a magic-wand number that ignored them would be cheating. Even being generous — treating those premixes as costing as much again as the bulk lactose, oil, and protein put together — the raw-material floor for a kilogram of powder still sits in the low-to-mid single-digit dollars. A rough magic-wand estimate of about two to six dollars a kilo is honest enough to reason with. That is the theoretical floor: the price of the stuff if arranging it were free.

The shelf is a different number. A name-brand can holding roughly six to nine hundred grams and selling for something in the mid-twenties to around forty dollars is asking, in round numbers, something like thirty-five to sixty dollars per kilogram, and a small hypoallergenic can can run higher still. Divide that finished price by a two-to-six-dollar material floor and the Idiot Index — finished cost over material cost, offered here as arithmetic you can redo, not as a measured factory metric — lands somewhere around eight to twenty for ordinary name-brand powder. Store brands compress the ratio. Extensively hydrolyzed and amino-acid formulas stretch it, because those products have genuinely done more work to the protein. The integer is not the point. The point is that you are many times the atoms, and not because the atoms are exotic.

Where does the multiple actually live? Walk the can backward. Spray drying is real industrial work: a concentrated liquid is atomized into a tower of hot air so water flashes off in seconds, and the dryer, the evaporators upstream of it, and the energy bill are not free. They are also a mature cheese-plant technology, not a rocket engine, and they do not, by themselves, turn two dollars of solids into forty dollars of powder. Wet blending, homogenization, and heat treatment add capital and yield loss in the same bounded way. Pathogen control is a heavier line than outsiders expect, because powdered formula is not sterile. The organism that has driven the worst modern recalls lives in dry environments and is a genuine infant hazard, so plants run environmental monitoring, hold-and-release testing, and a quality system that looks more like a drug plant than a powdered-milk shed. That costs money. It does not cost a tenfold markup on lactose.

Packaging is visible and over-explained. A steel or composite can, a scoop, a foil diaphragm, and a nitrogen flush are a few dollars of converted metal and plastic, not the gap. The specialty lipids and the vitamin ticket are a bigger share of the ingredient bill than their grams suggest, and they are also a branding handle: once “DHA” is on the front of the can, the product has stepped out of the dairy case and into the pediatric aisle. Regulatory overhead and the time to qualify a new dryer or a new recipe are a carrying cost. Capacity that takes years to stand up will not compete away a margin the way a new snack line will.

The rest of the multiple is not on the factory floor. It is in the channel. Brand advertising and pediatric detailing teach a frightened buyer that these beige powders are not substitutes for one another. The discharge bag and the sole-source public contract lock that lesson in. Wholesale and retail take their cut on a high list price. The rebate structure on those public contracts can actually push list prices up rather than down, because the fight is over the rebate, not over the unsubsidized shelf tag the non-program parent pays. Ready-to-feed liquid makes the same logic louder: you are then paying to sterilize and to ship water, and the Idiot Index against the dissolved solids gets worse, which is why powder is already the cheap format and still not cheap. Add those stages and you can see a path from a few dollars of commodities to a thirty-dollar can without once discovering a missing expensive atom. The waste is in process, packaging, trust machinery, and lock-in — and the lock-in is the thickest layer.

Segment 4 — The First-Principles Opportunity

A from-scratch redesign would not start by inventing a new baby food. It would start by telling the truth about the one we already have: this is a nutrient-specified dairy powder, and the lactose and whey in it often began as the leftover of cheesemaking. The first move is to co-locate the dryer with those streams and to run formula as a high-care product on a dairy-solids campus rather than as a standalone pharmaceutical object. Store-brand cans already prove that a lower price is manufacturable from the same USDA-grade inputs and the same federal nutrient table. What has to be true for that move to cut toward the floor is boring and specific — inspected capacity next to a whey silo, a wet process that stays inside the pathogen spec, and a purchase contract for oils and premixes that does not treat every additive as a brand exclusive.

The second move is the can itself. Multi-serve pouches and simpler barrier packs already exist in other markets and in some domestic lines; they do not change the powder and they do shave conversion cost, cube, and shipping mass. What has to be true is oxygen and moisture barrier good enough that the fat does not oxidize and the powder does not cake, plus a filling line that keeps the dry product dry. That is packaging engineering, not a new theory of infant nutrition, and it will not by itself collapse an Idiot Index of ten. It is the easy second cut once the plant is in the right place.

The third move is the channel and the rules that freeze it. Sole-source public contracts, hospital discharge kits, and a notification-and-inspection timeline that makes a new dryer a multi-year bet are why a handful of labels can hold a pediatric premium on a cheese byproduct. A redesign that wants the cost to fall has to bid the public program on landed cost rather than on rebate theater, put a generic can in the discharge bag, and make inspected capacity something you can add as if it were a dairy dryer with a special quality system, not a once-a-generation factory. The prize is not a race to an unsafe floor. It is a store-brand-like price on every standard can, with the remaining premium reserved for the products that have actually done more chemistry.

The hard parts are real and should be named without romance. Cronobacter in low-moisture powder is not a public-relations problem; heat that would reliably sterilize the powder also damages the nutrients the statute requires, so the control has to be the process and the building, not a terminal kill step. Parents will pay for trust, and they should — a hungry infant is not a place to A/B-test a brand. Extensively hydrolyzed and amino-acid formulas for cow’s-milk allergy have a genuinely higher material and process bill, because enzymes, filtration, and free amino acids are not skim-milk powder. Liability follows any new plant for decades. None of that justifies a double-digit Idiot Index on the ordinary can. It justifies treating the ordinary can as dairy, and the specialty can as the place where the extra money actually bought extra work.

Segment 5 — The Lesson

When the customer cannot walk away — a hungry newborn, a discharge bag, a state contract — the gap between the can and the atoms stops being a manufacturing residual and becomes a measure of lock-in. A nutrient table is a recipe constraint, not a price floor: hitting every milligram of iron and linoleic acid the government named does not require a luxury label, only a dryer, a premix, and a lab that can prove the batch. Tomorrow, one concrete example of this kind of thinking or one more industry that is still waiting for it. The question worth carrying into the aisle is this: who files the next infant-formula notification for a dryer that already sits beside a whey silo, and what is the first signal they are serious — a public-program bid at a landed cost the incumbents cannot match, or a store-brand shelf tag that finally stops shadowing the name brands?

💬 Reply to this email — Patrick reads every one.

Share: X · LinkedIn · WhatsApp

Forwarded this email? Subscribe here — it's free.