How food gets engineered

TL;DR. Processed food on American shelves is the output of a research-and-marketing pipeline. Howard Moskowitz tunes the bliss point via conjoint analysis (Maxwell House, Prego, Cherry Vanilla Dr Pepper). Sensory-specific satiety, discovered at Natick, explains why one-note products like Coke outsell complex ones and why supermarkets grew from 6,000 to 33,000 SKUs. Coca-Cola plays for share of stomach over water itself. Sugar reaches the brain in ~600 ms — twenty times faster than nicotine — and stop circuits don't catch up. David Kessler documents the layering and loading behind the 880-calorie Cinnabon and the 1,950-calorie Bloomin' Onion. Mark Schatzker documents the flavor industry — 1.4 billion lb/yr globally, with "natural flavor" as a legal category. Marion Nestle documents the supermarket itself: $2M-per-SKU slotting fees, 60-inch eye-level shelves auctioned to the highest bidder. The FTC's 1980 KidVid loss left an industry free to spend ~$14B/year marketing to American children.

What you'll learn

• The bliss point machinery and the named products it produced.
• How the supermarket weaponizes variety against an evolved satiety brake.
• What "share of stomach" means, and how Kraft extended it to cheese.
• Why a 600-ms delivery beats willpower — and why PepsiCo silenced Dana Small.
• Kessler's layering-and-loading taxonomy.
• How "natural flavor" became a legal category that means nothing chemically.
• Slotting fees, eye level, endcaps — the store as profit machine.
• Why the FTC lost the 1980 fight over child food marketing.

1. The bliss point — Howard Moskowitz and the optimization of liking

The phrase bliss point is not metaphorical. It is the output of a procedure invented by Howard Moskowitz — Harvard PhD, Army food researcher at Natick in the 1960s, then a White Plains consultant — that finds the concentration of sweet, fat, or salt at which liking peaks just before satiety kicks in. The commercial sweet spot is the plateau just below the upper bound, which is also where the cheapest reformulations live. The method is conjoint analysis: vary one attribute at a time across hundreds of prototypes, run designed taste panels, map the response surface.

Michael Moss documents the case studies in Salt Sugar Fat. Maxwell House was losing to Folgers in the 1980s; Moskowitz convinced General Foods's John Ruff to ship three roasts instead of one. Prego was reformulated around chunky style and a sugar load of more than two teaspoons per half-cup. Cherry Vanilla Dr Pepper was developed for Cadbury Schweppes in 2004 across 61 formulas and 3,904 tastings in four cities — not to win new drinkers, but to lift consumption among existing ones.

The objective is not "best taste." It is maximum repeat consumption before the satiety brake engages.

2. Sensory-specific satiety — why Coke beats the smorgasbord

Moskowitz's other Natick discovery, from studying soldiers' MRE fatigue, is sensory-specific satiety: one-flavor foods reach "enough" quickly; multi-flavor meals drive substantially higher intake, because each new flavor triggers a fresh round of wanting.

The industry's response was structural. The biggest-selling products — Classic Coke, Doritos, Velveeta Cheesy Skillets — are alluring without being dominant in any one note. Nothing in Coke says stop. And the supermarket was redesigned to weaponize variety: U.S. stores grew from ~6,000 SKUs in 1980 to ~33,000 today, mostly via line extensions (Oreo went from 6 to ~200 SKUs since 1990). The 1988 Kroger Variety Research Program — run with Nabisco, Frito-Lay, Kellogg's, Coca-Cola, and General Mills — identified the "variety seeker" as the highest-margin segment.

3. Share of stomach — Coca-Cola, heavy users, and Convenience Teens

Coca-Cola did not, internally, frame its competition as Pepsi. Traceable to Robert Woodruff's "ubiquity strategy" and made explicit during Jeffrey Dunn's run as president of North and South America operations, the frame was share of stomach: displace water, juice, coffee, milk, tea. The unit of analysis was the heavy user. Internal Coke language, attested by ex-marketer Todd Putman, ignored "customers"; heavy users — two-a-day, three-a-day drinkers — drove ~80% of volume. The original Coke bottle held 6.5 ounces; the 7-Eleven Double Gulp held 64. The 2005 Convenience Teens program studied how to capture unaccompanied adolescents at convenience stores. Dead Zone was the internal term for any geography below target per-capita consumption.

Dunn broke ranks after a 2001 Brazil trip where his team brought him to São Paulo favelas to study replicating U.S. consumption among the children there. He wrote memos calling for restraint and was pushed out. Kraft applied the same heavy-user lens to cheese under Geoffrey Bible — the former Philip Morris CEO who oversaw both companies — and per-capita U.S. cheese consumption rose from ~11 lb/year in the 1970s to ~33 lb/year by the 2010s. A corporate project, not consumer demand.

4. Speed of delivery — why a 600-millisecond hit beats willpower

The variable the industry exploits most directly is the speed at which a substance reaches reward circuits. Faster delivery produces a more intense dopamine spike and a more durable habit. Crack hits harder than snorted cocaine for this reason.

Sugar in solution reaches the brain in ~600 milliseconds — twenty times faster than nicotine — via taste-bud and trigeminal-nerve signaling that begins firing before food has cleared the throat. Stop circuits (hippocampus, orbitofrontal cortex) take longer to engage than go circuits (striatum, hypothalamus). A processed product can deliver reward, install a memory, and trigger wanting before executive control has noticed a decision is being made.

The design consequences are two. The industry optimizes for fast sugar — high-fructose corn syrup, glucose syrups, refined-flour carriers. And it engineers for fast eating: low chewing resistance, easy meltdown. Sensory Spectrum's Gail Vance Civille notes Americans now chew ~10 times per mouthful, down from ~25 a generation earlier.

This is also why Dana Small matters. The Yale neuroscientist was funded by PepsiCo under Indra Nooyi's "Big Bet" initiative to investigate whether reduced-calorie sodas were metabolically safer. Her brain scans suggested the opposite — a 112.5-calorie soda was more rewarding than a 150-calorie one, possibly because the brain registers a mismatch between sweetness signal and caloric payload and dials up wanting. PepsiCo cut her funding overnight; executive John Fletcher reportedly told collaborator Linda Flammer, "She is dangerous."

5. Layering and loading — Cinnabon, the Bloomin' Onion, the 1,950-calorie appetizer

David Kessler, former FDA commissioner, spent years interviewing an anonymous senior food consultant — a "Henry Ford of mass-produced food." The taxonomy: layering (salt on sugar on fat on fat) and loading (driving fat into the core ingredient via marination, injection, breading, par-frying).

Cinnabon, invented in 1985 by Jerilyn Brusseau with Rich Komen, was engineered through pillow-soft dough, a syrupy cinnamon-caramel center, cream-cheese frosting, three sugars, and Sumatran-and-Vietnamese cinnamons. The Classic Roll runs 880 calories; mall ventilation carries the aroma blocks before the storefront is visible. The Outback Bloomin' Onion runs ~1,950 calories, sold as an appetizer — engineered for surface area by cutting the onion into petals before deep frying. Cheesecake Factory pastas clock 1,800–2,500; Claim Jumper's Chocolate Motherlode Cake is 2,150.

Kessler's source decoded specific dishes. Potato skins are "fat on fat on fat on fat." Buffalo wings are "sugar on salt on fat on fat on fat." Chicken tenders are "UFOs — unidentified fried objects." Chili's Boneless Shanghai Wings contain a 25% water solution, are battered and breaded at a plant, par-fried, frozen, and re-fried on site — "prechewed," in ex–Standard Meat president Billy Rosenthal's word.

Kessler's clinical name for the pattern is conditioned hypereating: loss of control, lack of satisfaction, preoccupation with food. He estimates 70 million Americans live inside it.

6. The flavor industry — McCormick, Givaudan, and "natural flavor" as a legal category

The seventh ingredient in most processed food is invisible: synthetic flavor. Mark Schatzker's The Dorito Effect traces it. Commercial gas chromatography after 1955 industrialized the route: any flavor can be deconstructed and resynthesized at parts-per-billion concentrations. Synthetic flavor chemicals in use rose from ~700 in 1965 to over 2,200 today; the global flavor industry produces ~1.4 billion pounds per year, dominated by Givaudan, Firmenich, IFF, Symrise, and McCormick.

The consequential trick is the legal definition of "natural flavor": one derived from a plant, animal, fungus, or microbial source via extraction, distillation, fermentation, or enzymatic process. The molecule may be identical to its "artificial" counterpart; what makes it "natural" is the production pathway. McCormick can deliver a synthetic pumpkin-spice profile using ~80 aroma compounds, each "natural" if sourced enzymatically from a botanical precursor. A marketing distinction enforceable in court, not a chemistry one.

Per-capita U.S. herb and spice consumption rose ~500% over the twentieth century — not from sophisticated palates but because real food now requires it. Chicken yields tripled, dairy yields multiplied tenfold; the ingredients lost flavor in step. Schatzker's claim, from Utah State range scientist Fred Provenza: flavor is nutrient information — aroma compounds in real food are typically built from essential amino acids, carotenoids, or omega-3 fats. When the channel is corrupted, the wanting system fires without satiety follow-through.

7. The supermarket as profit machine — slotting fees and eye level

Marion Nestle's What to Eat Now (2025) treats the supermarket as the operating environment where the preceding engineering pays off. Layout follows codified rules: highest-margin sections at the perimeter, milk at the back (the trip past 200 other products is the point), impulse zones at entrance and checkout, endcaps and 60-inch eye-level shelves for the most-promoted products.

Shelf space is priced via slotting fees: a manufacturer pays a chain — often $2M+ per SKU for chain-wide placement — to get on shelf. The fees emerged in the 1980s and were investigated by Congress in 1999, where some witnesses testified hooded for fear of retribution. Real food has no marketing budget and pays no slotting fees, which is why industrial formulations dominate the shelves. Walmart now controls ~25% of U.S. grocery sales; the top four chains hold more than half, up from the top twenty in 2005.

8. Marketing to children — KidVid 1980, Frosted Mini-Wheats 2008, the pester factor

The largest unprosecuted regulatory failure in U.S. food policy is the 1980 collapse of the FTC's KidVid initiative. Under chair Michael Pertschuk, the FTC was preparing to restrict TV advertising of sugary foods to children, on the finding that young children cannot distinguish advertising from programming. The food and broadcasting industries assembled a $16M coalition under lobbyist Tommy Boggs. The Washington Post labeled the FTC "The National Nanny." Congress briefly shut the FTC down in 1980; Pertschuk was ousted.

The loss authorized ~$14B/year in U.S. food, beverage, and restaurant marketing aimed at children, ~80% targeting fast food, sugary drinks, candy, and unhealthy snacks. The industry's self-regulator, CARU, is funded by the companies it polices. Mexico and Chile have enacted statutory restrictions that have measurably shifted purchase patterns. The U.S. has not.

In 2008 the FTC cited Kellogg's for the Frosted Mini-Wheats "20% better attentiveness" classroom ad as a false claim — the Kellogg-funded study underlying it showed roughly half the tested children had no improvement at all. Kraft moved its Tang target to children aged 9–14, younger than Coca-Cola's stated floor of 12. Bob Drane's Lunchables, launched in 1988 to save bologna sales, was designed around children's preference for assembling food themselves; Drane later co-founded a baby-carrot effort with Jeffrey Dunn, telling Moss he carried "plenty of guilt" about Lunchables.

Industry research on the pester factor is operational — persistence nagging and importance nagging, both designed for. A Kraft executive, asked privately by Marion Nestle why the company kept marketing to children, said they would stop if they could "but our stockholders won't let us." ABC anchor Peter Jennings once asked a Kraft executive on air at what age marketing to children becomes unethical; the answer — six — drew gasps from the audience.

The eight techniques compose one machine. James Behnke, the Pillsbury executive who organized the secret April 8, 1999 Minneapolis meeting at which eleven food-company CEOs were briefed on the obesity epidemic — and where General Mills's Stephen Sanger pushed back and effectively ended the initiative — gave Moss the closing line of Salt Sugar Fat: We're hooked on inexpensive food, just like we're hooked on cheap energy.

Frequently asked questions

Why isn't any of this regulated?

Some of it is, weakly. Trans fats are out of the supply post-2018; chain-restaurant calorie labels are mandatory since 2018; added sugars appear on the Nutrition Facts panel since 2016. Marketing to children, slotting fees, front-of-pack warning labels, and meaningful sodium-reduction targets remain unregulated. The structural reason is industry lobbying (~$50M/year federal) and a Dietary Guidelines process that, since 1977, has replaced food-named recommendations ("eat less red meat") with nutrient-named ones ("limit saturated fat") that allow infinite reformulation.

What about "clean label" reformulations?

A marketing category, not a regulatory one. A "lower-sugar" cereal with reduced sodium and added fiber is still NOVA Group 4, still engineered for hyperpalatability, still subject to all eight techniques here.

Is fast food the worst category, or is it the supermarket?

Same playbook, different venues. The supermarket is harder to escape because it presents as a neutral retailer when most of its shelf space is industrial product.

Are sit-down chains really worse than fast food?

Frequently yes on calorie load per dish. Cheesecake Factory pastas run 1,800–2,500 cal; the Bloomin' Onion runs 1,950; Claim Jumper's Chocolate Motherlode Cake is 2,150. Fast-food chains tend to cap single items below 1,000 cal.

What about D2C "better-for-you" brands?

Most are still ultra-processed. A protein bar with a clean-looking list (pea protein isolate, brown rice syrup, "natural flavors," chicory root fiber) is a NOVA Group 4 formulation with a different demographic profile. The test is the ingredient list, not the brand story.

Sources

1. Moss, M. Salt Sugar Fat. Random House, 2013. Moskowitz's bliss-point work; the 1999 Pillsbury CEO meeting (Mudd, Sanger, Behnke); Coca-Cola's share-of-stomach and Jeffrey Dunn's defection; Kraft and Geoffrey Bible on cheese.
2. Moss, M. Hooked. Random House, 2021. The 600-ms sugar-to-brain delivery; the Kroger Variety Research Program; PepsiCo's silencing of Dana Small.
3. Schatzker, M. The Dorito Effect. Simon & Schuster, 2015. McCormick; "natural flavor" as legal-vs-chemical; Fred Provenza on flavor as nutrient information.
4. Kessler, D. A. The End of Overeating. Rodale, 2009. The layering-and-loading taxonomy; Cinnabon, Bloomin' Onion, Cheesecake Factory; conditioned hypereating.
5. Nestle, M. What to Eat Now. Farrar, Straus and Giroux, 2025. Slotting fees ($2M/SKU); Walmart's 25% share; the engineered Nutrition Facts panel; child marketing; the Kraft "stockholders won't let us" admission.
6. FTC v. Kellogg Company consent order, 2009. Frosted Mini-Wheats "20% better attentiveness" false claim.
7. Hall, K. D., et al. (2019). "Ultra-processed diets cause excess calorie intake and weight gain." Cell Metabolism, 30(1), 67–77.e3.

Related modules

• ← C1: What is food, really?
• C8: The ultra-processed food problem →
• C10: Reading labels →
• C12: Why willpower isn't the problem →

Related glossary terms

• Bliss point
• Sensory-specific satiety
• Share of stomach
• Slotting fee
• Natural flavorings
• Conditioned hypereating
• Heavy users
• Monosodium glutamate
• Hydrolyzed vegetable protein
• High-fructose corn syrup
• Disodium inosinate