How food gets engineered

TL;DR. Processed food comes from a research-and-marketing pipeline. Howard Moskowitz tunes the bliss point (the sugar or salt level where you like it most) using taste-panel math. He worked on Maxwell House, Prego, and Cherry Vanilla Dr Pepper. Sensory-specific satiety explains why one-flavor drinks like Coke beat complex meals. It also explains why stores grew from 6,000 items to 33,000. Coca-Cola fights for share of stomach against water itself. Sugar hits the brain in about 600 milliseconds. That is 20 times faster than nicotine. Your stop signals never catch up. David Kessler tracks the layering and loading behind the 880-calorie Cinnabon and the 1,950-calorie Bloomin' Onion. Mark Schatzker tracks the flavor industry. It makes 1.4 billion pounds of flavor a year. The label "natural flavor" is a legal term, not a chemistry one. Marion Nestle tracks the store: $2 million slotting fees per item, with eye-level shelves sold to the top bidder. The FTC lost the 1980 KidVid fight. Companies now spend about $14 billion a year marketing to U.S. kids.

What you'll learn

• How the bliss point gets tuned, and what products it built.
• How stores use variety to break your satiety brake.
• What "share of stomach" means, and how Kraft used it on cheese.
• Why a 600-ms hit of sugar beats willpower, and why PepsiCo silenced Dana Small.
• Kessler's layering-and-loading system.
• How "natural flavor" became a legal term that means nothing in chemistry.
• Slotting fees, eye level, endcaps: the store as a profit engine.
• Why the FTC lost the 1980 fight over food ads aimed at kids.

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

The phrase bliss point describes a measurement. Howard Moskowitz built the method behind it. He earned a Harvard PhD, worked as an Army food researcher at Natick in the 1960s, and then opened a consulting shop in White Plains. The method finds the concentration of sugar, fat, or salt at which your liking peaks. That peak sits just before you start to feel full.

Companies aim for the cheap plateau just under that peak. The procedure is called conjoint analysis. Researchers change one ingredient at a time across hundreds of prototypes, run designed taste panels, and map the response surface.

Michael Moss tells the case studies in Salt Sugar Fat. Maxwell House was losing to Folgers in the 1980s. Moskowitz told General Foods's John Ruff to ship three roasts, not one. Prego got a chunky version and more than two teaspoons of sugar per half cup. Cherry Vanilla Dr Pepper was built for Cadbury Schweppes in 2004 using 61 recipes and 3,904 taste tests in four cities. The goal was not new drinkers. The goal was more drinks per existing fan.

The target is never "best taste." Companies optimize for the maximum number of repeat bites before the satiety brake engages.

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

Moskowitz uncovered a second effect at Natick by studying soldiers who got tired of their field rations. He named it sensory-specific satiety: a one-flavor food fills you up fast, while a many-flavor meal pulls you back in because each new taste wakes up fresh hunger.

The industry built its biggest products around this finding. Top sellers like Classic Coke, Doritos, and Velveeta Cheesy Skillets taste good without letting any single note dominate. Nothing in Coke signals you to stop.

The store changed too. U.S. stores held about 6,000 items in 1980. They hold about 33,000 today. Most of that growth came from line extensions. Oreo went from 6 versions to about 200 since 1990. In 1988 Kroger ran the Variety Research Program with Nabisco, Frito-Lay, Kellogg's, Coca-Cola, and General Mills. They named the "variety seeker" as the most valuable shopper.

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

Inside Coca-Cola, Pepsi was never the real rival. Robert Woodruff set the goal decades ago as "ubiquity," and Jeffrey Dunn made it explicit during his run as president of North and South America. The frame the company actually used is share of stomach: take fluid ounces away from water, juice, coffee, milk, and tea.

The unit of analysis is the heavy user. Former Coke marketer Todd Putman says executives stopped saying "customers" in meetings. Heavy users drank Coke two or three times a day and produced about 80% of sales volume. The first Coke bottle held 6.5 ounces; the 7-Eleven Double Gulp holds 64. In 2005 Coke ran Convenience Teens, a program designed to capture teens who stopped at convenience stores on their own. Dead Zone was the internal term for any area where per-person consumption fell short of target.

Dunn broke ranks in 2001. His team took him to São Paulo favelas in Brazil to plan how to grow drinks per child. He wrote memos asking for limits. He got pushed out. Kraft used the same heavy-user lens on cheese under Geoffrey Bible, the former Philip Morris CEO who ran both companies. U.S. cheese per person rose from about 11 pounds a year in the 1970s to about 33 pounds by the 2010s. That growth came from a corporate plan, not from shopper demand.

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

The variable that matters most is speed. The faster a substance reaches the reward circuits in the brain, the bigger the dopamine spike and the more durable the habit. Crack hits harder than snorted cocaine for this same reason.

Sugar in liquid reaches the brain in about 600 milliseconds, roughly 20 times faster than nicotine. Your taste buds and the trigeminal nerve start firing signals before the drink clears your throat. Stop circuits in the hippocampus and orbitofrontal cortex take longer to engage than go circuits in the striatum and hypothalamus. A processed drink can deliver reward, install a memory, and trigger craving before you know you made a choice.

Two design rules follow. Companies pick fast sugars such as high-fructose corn syrup, glucose syrups, and refined-flour carriers. They also engineer foods you can eat quickly: low chewing resistance, easy meltdown. Gail Vance Civille of Sensory Spectrum estimates that Americans now chew about 10 times per mouthful, down from about 25 a generation ago.

This is why Dana Small matters. The Yale neuroscientist accepted PepsiCo funding under Indra Nooyi's "Big Bet" initiative to test whether reduced-calorie sodas were metabolically safer. Her brain scans pointed the opposite way. A 112.5-calorie soda activated reward centers more than a 150-calorie one. The brain may register a mismatch between the sweetness signal and the calories that arrive, then dial up wanting. PepsiCo cut her funding overnight. Executive John Fletcher reportedly told her collaborator Linda Flammer, "She is dangerous."

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

David Kessler, the former FDA commissioner, spent years interviewing a senior food consultant who insisted on staying anonymous. Kessler called him a "Henry Ford of mass-produced food." The taxonomy has two parts. Layering stacks salt on sugar on fat on fat. Loading drives more fat into the main ingredient through marination, injection, breading, and par-frying.

Jerilyn Brusseau and Rich Komen built Cinnabon in 1985. The pieces: soft dough, a syrup of cinnamon and caramel at the core, cream-cheese frosting, three sugars, and a blend of Sumatran and Vietnamese cinnamons. The Classic Roll holds 880 calories. Mall vents push the smell down the hall before you can see the store. The Outback Bloomin' Onion runs about 1,950 calories. It sells as an appetizer. The cook cuts the onion into petals to add surface area, then deep fries it. Cheesecake Factory pastas run 1,800 to 2,500 calories. Claim Jumper's Chocolate Motherlode Cake is 2,150.

Kessler's source broke down 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 hold a 25% water solution. A plant batters and breads them, par-fries them, freezes them, and the restaurant fries them again. Billy Rosenthal, former president of Standard Meat, called them "prechewed."

Kessler named the result conditioned hypereating. The signs: loss of control, no real satisfaction, food on your mind all day. He estimates 70 million Americans live inside it.

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

Most processed food hides a seventh ingredient: synthetic flavor. Mark Schatzker traces the story in The Dorito Effect. Commercial gas chromatography after 1955 industrialized the process. Any flavor can be deconstructed into its components and rebuilt at parts-per-billion concentrations. The number of flavor chemicals in commercial use rose from about 700 in 1965 to over 2,200 today. The global flavor industry produces about 1.4 billion pounds a year, dominated by Givaudan, Firmenich, IFF, Symrise, and McCormick.

The consequential trick is the legal definition of "natural flavor." A flavor qualifies as natural if it is derived from a plant, animal, fungus, or microbe through extraction, distillation, fermentation, or enzymatic processes. The resulting molecule can match its "artificial" counterpart atom for atom. What makes it "natural" is the production pathway, not the chemistry.

McCormick can build a pumpkin-spice profile from about 80 aroma compounds. Each one counts as "natural" if it comes from a plant source through enzymes. The line is a marketing rule a court can enforce. It is not a chemistry fact.

U.S. herb and spice use per person rose about 500% across the 20th century. The reason is not better palates. Real food now needs help. Chicken yields tripled. Dairy yields rose tenfold. The ingredients lost flavor as the yields climbed. Fred Provenza, a range scientist at Utah State, told Schatzker that flavor is nutrient information. Aroma compounds in real food often come from amino acids, carotenoids, or omega-3 fats. When the channel gets faked, your wanting system fires without the matching nutrition.

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

Marion Nestle's What to Eat Now (2025) treats the supermarket as the environment where all the previous engineering pays off. The layout follows codified rules. High-margin sections sit at the perimeter. Milk sits at the back because the walk past 200 other products is the point. Impulse zones sit at the entrance and the checkout. The 60-inch eye-level shelf and the endcap belong to the most-promoted brands.

Shelf space is priced through slotting fees. A manufacturer often pays a chain $2 million or more per item for chain-wide placement. 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 about 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 aimed at children, based on the finding that young children cannot distinguish ads from programming. The food and broadcasting industries assembled a $16 million coalition led by lobbyist Tommy Boggs. The Washington Post labeled the FTC "The National Nanny." Congress briefly shut the FTC down in 1980, and Pertschuk lost his job.

That loss authorized about $14 billion a year in U.S. food, beverage, and restaurant marketing aimed at children. About 80% of that spending pushes fast food, sugary drinks, candy, and 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 charged Kellogg's for the Frosted Mini-Wheats "20% better attentiveness" ad. The Kellogg-funded study behind the claim showed about half the children tested had no gain at all. Kraft moved its Tang target to children aged 9 to 14, under Coca-Cola's own floor of 12. Bob Drane built Lunchables in 1988 to save Kraft's bologna sales. He designed it around how children like to put food together themselves. Drane later helped Jeffrey Dunn launch a baby-carrot project. He told Moss he carries "plenty of guilt" about Lunchables.

Industry research on the pester factor runs every day. The two named types are persistence nagging and importance nagging. Both are built into ads on purpose. Marion Nestle once asked a Kraft executive in private why the company kept marketing to children. He said they would stop if they could, "but our stockholders won't let us." On ABC, anchor Peter Jennings once asked a Kraft executive on air at what age it becomes wrong to market to kids. The answer was six. The audience gasped.

The eight techniques work as one machine. James Behnke, the Pillsbury executive, set up the secret April 8, 1999 Minneapolis meeting. Eleven food-company CEOs got briefed on the obesity crisis. General Mills's Stephen Sanger pushed back and shut the effort down. Behnke gave Moss the last 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 left the supply after 2018. Chain restaurants have to post calorie counts since 2018. Added sugars sit on the Nutrition Facts panel since 2016. Marketing to children, slotting fees, front-of-pack warning labels, and real sodium limits remain unregulated. The reason is lobbying (about $50 million a year at the federal level) and a Dietary Guidelines process that since 1977 has swapped food advice ("eat less red meat") for nutrient advice ("limit saturated fat"). Nutrient advice lets companies keep reformulating.

What about "clean label" reformulations?

That is a marketing label, not a legal one. A cereal with less sugar, less sodium, and added fiber is still NOVA Group 4. It is still built for hyperpalatability. It is still shaped by all eight techniques above.

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

Same playbook, different stage. The supermarket is harder to escape because it looks like a neutral store when most of its shelf space holds industrial product.

Are sit-down chains really worse than fast food?

Often yes on calories per dish. Cheesecake Factory pastas run 1,800 to 2,500 calories. The Bloomin' Onion runs 1,950. Claim Jumper's Chocolate Motherlode Cake is 2,150. Fast-food chains tend to cap single items under 1,000.

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

Most stay 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 recipe with a different target shopper. 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 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