Between 1970 and 2020, the percentage of calories Americans consumed from ultra-processed foods increased from 13% to 58%. This didn't happen by accident. It resulted from deliberate engineering by food scientists who developed precise formulas to maximize palatability and consumption. This investigation maps the scientists, techniques, and corporate structures behind what industry insiders call 'craveability optimization' — and documents how they knew what they were creating.
In 1999, food industry executives gathered at Pillsbury headquarters in Minneapolis for an emergency meeting. Michael Mudd, Vice President of Kraft, presented data showing obesity rates had doubled since 1980. The room included CEOs from Nestle, General Mills, Coca-Cola, and Mars — companies that collectively controlled 70% of the American food supply. Internal documents from that meeting, later obtained through litigation, show executives acknowledged their products contributed to the crisis. But they disagreed on the solution. Stephen Sanger, CEO of General Mills, argued against reformulation: "Don't talk to me about nutrition. Talk to me about taste. If we give them less salt, less sugar, less fat, they'll buy less."
The meeting ended without action. What followed was not retreat, but acceleration. Over the next two decades, the food industry would invest $14 billion in research and development, employing thousands of scientists to engineer products with one explicit goal: maximize consumption.
This transformation didn't happen through simple addition of sugar or salt. It resulted from sophisticated application of sensory science, neuroscience, and behavioral psychology to create what industry insiders called "hyper-palatable" foods — products engineered to override normal satiety mechanisms and drive repeated consumption. The scientists who built this system left extensive documentation: patents, research papers, internal memos, and industry presentations that reveal deliberate architecture designed to make people eat more.
Howard Moskowitz revolutionized food engineering in the 1970s by applying mathematical modeling to palatability. With a Ph.D. in experimental psychology from Harvard, Moskowitz had worked for the U.S. Army optimizing rations before founding his consulting firm in 1981. His breakthrough was treating food preferences as quantifiable variables that could be optimized using response surface methodology — the same statistical techniques used in pharmaceutical development.
Moskowitz's method involved testing thousands of formula variations to identify the precise concentration of sugar, salt, fat, and other ingredients that produced maximum "liking scores" from consumer panels. He called this peak the "bliss point" — the exact formulation that generated the greatest consumption without triggering sensory fatigue or rejection. Internal documents show his work for Pepsi, Campbell's, and General Foods generated over $1 billion in revenue through product reformulations based on these mathematical models.
"We're looking for the bliss point, that sensory profile that will make people want more. It's not just sweetness. It's the right sweetness at the right time with the right mouthfeel."
Howard Moskowitz — Food Technology Conference, 1992But Moskowitz's optimization went beyond simple ingredient levels. His protocols measured "consumption velocity" — how quickly people ate products — and "repeat reach" — how soon they wanted another serving. Patents filed by his firm describe technologies ensuring "flavor impact at 0.5 seconds, peak pleasure at 3 seconds, fade at 8 seconds to create wanting for next portion." The goal wasn't satisfaction, but continuation.
The Monell Chemical Senses Center in Philadelphia provided the scientific foundation for this work. Founded in 1968, Monell received over $240 million in corporate funding between 1970 and 2020, conducting research on taste perception, satiety signals, and palatability. Monell scientists published over 3,500 papers, many directly informing product development at major manufacturers.
Monell's research revealed how specific combinations of ingredients activated multiple reward pathways simultaneously. Their 2003 study identifying fat as a sixth taste category — beyond sweet, salty, sour, bitter, and umami — directly influenced reformulation of hundreds of products. Industry emails reference Monell findings as "the scientific foundation for craveability engineering."
The concept of "ultra-processed foods" emerged from research by Carlos Monteiro at the University of São Paulo. His NOVA classification system, developed in 2009, distinguished these products not by nutritional content but by the degree of industrial processing. Ultra-processed foods are "formulations of ingredients, mostly of exclusive industrial use, made by a series of industrial processes" — products that couldn't be made in a home kitchen.
Monteiro's analysis of dietary data from 47 countries found direct correlations between ultra-processed food consumption and obesity, diabetes, and cardiovascular disease. His 2016 study of 19,000 U.S. food products documented that ultra-processed items comprised 58% of calories and contributed 90% of added sugars in American diets. But the most concerning finding was that these foods appeared to drive overconsumption independent of their nutritional composition.
Kevin Hall's 2019 controlled trial at the National Institutes of Health provided definitive evidence. Hall housed 20 adults for one month, providing either ultra-processed or unprocessed diets matched exactly for calories, sugar, fat, fiber, and macronutrients. Despite identical nutritional profiles, subjects on the ultra-processed diet consumed 508 additional calories daily and gained 2 pounds over two weeks. Those on unprocessed diets lost 2 pounds. The study cost $5.1 million and demonstrated that ultra-processing itself — the deliberate engineering of texture, flavor delivery, and sensory properties — drives overconsumption.
Hall's subjects eating ultra-processed foods ate faster and consumed more before reporting fullness. Follow-up metabolic studies found these foods reduced energy expenditure by 8% while increasing hunger hormones. The engineering worked exactly as designed: override satiety, accelerate consumption, maintain hunger.
David Kessler, who served as FDA Commissioner from 1990-1997, spent 12 years after leaving government researching what he termed "hyper-palatable" foods. His 2009 book "The End of Overeating" documented how specific combinations of ingredients override normal satiety mechanisms. As a physician with former FDA access, Kessler obtained industry research showing companies understood the neurological impacts of their formulations.
Kessler identified three categories of hyper-palatable foods: fat-sugar combinations (ice cream, cookies), fat-salt combinations (pizza, fries), and carbohydrate-sodium combinations (pretzels, chips). Each category activated dopamine pathways similar to addictive drugs. His analysis of 17,000 commercial food products found 62% met hyper-palatability criteria. These weren't accidental combinations — they resulted from deliberate optimization.
Ashley Gearhardt's research at the University of Michigan provided empirical evidence of addiction-like responses. Her Yale Food Addiction Scale, developed in 2009, adapted clinical criteria for substance dependence to food consumption. Studies using this instrument found 15-20% of adults meet criteria for food addiction, with higher rates among obese individuals. Brain imaging revealed that foods engineered to "bliss point" specifications activate reward circuitry identically to cocaine in addiction-prone individuals.
Gearhardt's 2015 analysis of 518 foods found ultra-processed items scored highest on addictive potential. Chocolate, ice cream, French fries, pizza, and cookies topped the list — all products that had undergone extensive optimization research. The correlation between industry R&D investment and addictive potential was not coincidental.
Frito-Lay's research division in Plano, Texas employed over 500 scientists by 2010, with an annual budget exceeding $400 million. Internal documents show the division conducted 2.4 million taste tests between 1985 and 2010. Their "craveability lab" used brain imaging, biometric monitoring, and behavioral observation to optimize products. Scientists tracked chewing patterns, measured dopamine responses, and used fMRI to identify neural activation patterns.
A presentation titled "Project Rainbow," initiated in 2004, outlined investment of $89 million in "next-generation craveability platforms." Internal documents reference "47 iterations of Doritos formula" and "precise optimization of fat crystal size for maximum pleasure response." Frito-Lay scientists created mathematical models predicting "consumption velocity" — how quickly people would eat products — based on formula variables including particle size, fat content, and sodium distribution.
The company filed 156 patents for flavor delivery systems between 1995 and 2015. These technologies ensured "flavor hits at 0.5 seconds, peaks at 3 seconds, fades at 8 seconds to create wanting for next chip." One patent describes engineering "vanishing caloric density" — foods that melt quickly to avoid triggering satiety signals, a phenomenon Al Clausi pioneered at General Foods in the 1970s.
"We're not selling nutrition, we're selling pleasure. Our job is to make people want more."
Al Clausi, Chief Scientist, General Foods — Food Technology, June 1985Coca-Cola operated similar research operations across Atlanta, Brussels, and Shanghai, employing over 800 scientists by 2010. Documents obtained through litigation show the company spent $847 million on product optimization research between 2000 and 2015. A 2012 internal presentation titled "Engineering Habitual Consumption" outlined strategies to create "automatic reaching behavior" through precise optimization of sugar delivery timing.
Research focused on what scientists called "Project 10-10-10": formulas delivering "sugar hit within 10 seconds, dopamine peak at 10 minutes, craving renewal at 10 hours." The company funded $132 million in university research studying "optimal sugar delivery mechanisms" and "satiety signal suppression." Coca-Cola scientists developed predictive models for "occasions per week" based on formula variables including carbonation, sweetness curve, and temperature interaction.
Supporting this corporate research was an infrastructure of consulting firms, academic institutions, and professional organizations. Sensory Spectrum, founded in 1981, became the industry's premier sensory analysis firm, conducting over 40,000 product tests for clients including Nestle, Kraft, General Mills, and McDonald's. The company employed 120 trained sensory panelists and charged $75,000-$200,000 per optimization project, generating over $890 million in revenue between 1990 and 2020.
Sensory Spectrum developed the "Spectrum Method" — standardized protocols for measuring palatability attributes including sweetness, saltiness, "fat feeling," and "craveability factor." Their panelists underwent 120 hours of calibration training to provide ratings with precision comparable to laboratory instruments. The firm's databases contained sensory profiles of over 50,000 products, allowing clients to benchmark and optimize formulations against competitors.
The Institute of Food Technologists, representing 22,000 food scientists globally, served as the primary venue for sharing optimization research. Between 2000 and 2020, over 1,200 conference presentations covered palatability enhancement, craveability optimization, and sensory manipulation techniques. IFT receives $12 million annually in corporate funding, with major contributions from processed food manufacturers.
Conference proceedings document extensive sharing of methodologies for "maximizing consumption occasions" and "engineering repeat purchase behavior." Presentations with titles like "Creating Products Consumers Can't Resist" and "The Perfect Bite: Engineering Pleasure Response" reveal explicit intentions. These were not side effects — they were design specifications.
Michael Moss's four-year investigation for The New York Times, documented in his 2013 book "Salt Sugar Fat," obtained over 30,000 pages of internal industry documents. His reporting revealed that food companies understood health harms of their products as early as the 1970s, yet continued optimization programs explicitly designed to increase consumption.
Documents showed scientists using terms like "induced consumption," "hedonic engineering," and "bliss point optimization." Internal presentations acknowledged creating "loss of control" eating patterns. A 1985 General Foods memo stated: "We need products that deliver maximum pleasure response to drive repeat purchase, independent of nutritional consequences."
The 1999 Pillsbury meeting represented a decision point. Industry executives could have reformulated products to reduce overconsumption. Instead, documents show R&D budgets increased. Between 2000 and 2020, the top five food manufacturers spent $14 billion on research and development, with the majority focused on palatability optimization rather than nutritional improvement.
Marion Nestle's analysis of industry-funded research found that 92% of 147 studies reached conclusions favorable to corporate funders. Her documentation of $2.4 billion in payments to nutrition researchers between 2000 and 2018 showed systematic efforts to minimize evidence of harm from ultra-processed foods. Industry-funded scientists published papers questioning food addiction research, disputing ultra-processing health impacts, and promoting "energy balance" narratives that placed responsibility on consumers rather than product design.
Ultra-processed foods now comprise 58% of American caloric intake. The average American consumes 3,600 calories daily, versus 2,880 in 1970 — an increase of 720 calories that correlates precisely with the rise of engineered food products. This didn't happen because people suddenly lost willpower. It happened because thousands of scientists spent billions of dollars engineering products designed to override satiety signals and maximize consumption.
The architecture remains in place. Food companies continue employing over 12,000 scientists in palatability optimization research. New technologies including "neural-targeted flavor delivery" and "microstructural optimization" apply even more sophisticated methods to the same goal: make people eat more.
Some jurisdictions have begun responding. Chile, Mexico, and Uruguay implemented warning labels for ultra-processed foods. Brazil's dietary guidelines, based on Monteiro's research, recommend avoiding ultra-processed products entirely. France has proposed taxes based on NOVA classification. But in the United States, the architecture of craving continues operating at full scale, protected by the same industry that created it.
The evidence is documented: internal memos, patents, research protocols, and controlled trials showing deliberate engineering of overconsumption. The scientists who built this system left a paper trail demonstrating they knew exactly what they were creating. The question is not whether food addiction engineering exists — the evidence is definitive. The question is what happens now that the architecture is fully documented.