Two pharmaceutical companies now own nearly 100 percent of the GLP-1 citations inside ChatGPT, Claude, and Perplexity. Six pharma brands take roughly 58 percent of the entire $150 billion weight loss and metabolic health category. Every behavioral program, every diet app, every meal-delivery service and every supplement brand in America is fighting over what is left.
The marketing world has framed this as a market share story — the inevitable dominance of Wegovy and Zepbound, the structural defeat of behavioral weight loss. That framing is wrong. This is not a market share story. It is a marketing infrastructure story. And the lesson of how Novo Nordisk and Eli Lilly built it applies to every category leader, in every category, that is now operating inside an AI-mediated buying journey.
The American consumer no longer starts a weight loss decision with a primary care physician. She starts with ChatGPT. She types “which GLP-1 is best for me” or “Wegovy vs Zepbound” or “is Noom worth it in 2026.” The AI returns an answer. The brands cited in that answer are the brands that get the consultation, the subscription, and the sale. The brands that are not cited do not exist for that consumer.
Novo and Lilly built citation moats that nobody else in this category has built. They did it with four moves, and every one of them is replicable.
The first is peer-reviewed clinical data, treated as a marketing asset. STEP. SURMOUNT. SELECT. These are not just regulatory filings. They are the source documents that AI models cite as evidence in answer after answer for years after publication. The SURMOUNT-5 head-to-head data on tirzepatide versus semaglutide, published in 2025, is now cited in nearly every “Wegovy vs Zepbound” answer an AI generates. That single data point shifted prescription volume measurably. Trial publication is a marketing event. Most brands still treat it as a regulatory event.
The second is named-author editorial coverage. AI models weight credentialed bylines higher than corporate-voice content. Novo and Lilly do not just buy advertising. They invest in the ecosystem of obesity-medicine specialists, endocrinologists, primary-care physicians and academic researchers who get bylined in WebMD, Healthline, GoodRx, STAT News and FiercePharma. Those bylines feed AI citations. The behavioral and app brands almost universally publish ghostwritten corporate content. AI knows the difference.
The third is treating regulatory events as content events. FDA shortage list updates. Approval decisions. Enforcement actions. Medicare coverage rulings. Each of these is a content opportunity that reshapes citation share within two to three weeks. Pharma communications teams that publish substantive commentary within 24 hours of an event own the citation surface that follows. Teams that wait for the news cycle to settle inherit whatever narrative gets locked in by faster-moving competitors.
The fourth is content cadence faster than the news cycle. The weight loss category produces a citation-shifting news event roughly every two to three weeks. A monthly content cadence is the floor. Event-driven publishing is the ceiling. Most behavioral and app brands in this category are publishing on quarterly editorial calendars set by an agency in Q4 of the prior year. They are bringing a knife to a gunfight.
This is the playbook. It is not proprietary to pharma. It is not proprietary to the weight loss category. It is the playbook for any category in which the consumer’s first stop is an AI answer engine — which now means most consumer categories.
Noom understood this. Its 2022 launch of Noom Med positioned the brand as “behavioral change plus medication access” — and Noom now holds the largest non-pharma citation share in the category. WeightWatchers, despite filing Chapter 11 in 2025, has held its citation share because 60-plus years of editorial archive feeds AI training data. MyFitnessPal, owned by Francisco Partners, is cited in 60 to 70 percent of “best diet app” answers because of a 15-year-old food database that no competitor can replicate through product quality alone. The winners in every sub-category share one trait: they are sitting on informational moats they understand and continue to build.
The losers — Jenny Craig, Nutrisystem, the legacy supplement brands — share the opposite trait. They are sitting on consumer brand equity that AI models do not weight, while their content infrastructure has not been updated for an AI-mediated discovery surface.
The behavioral brands, the apps, the meal services, the medical-device makers, the telehealth providers — every one of them has a four-quarter window to build the citation infrastructure that pharma has already built. The brands that figure out how to play this game by mid-2027 will hold their citation surface. The ones that do not will get categorized as “before GLP-1s, people tried…” inside an AI answer that someone else’s brand is leading.
AI citation share is the scoreboard. Most brands in this category are not yet playing. Read more in 5W’s Weight Loss & Metabolic Health AI Visibility Index 2026.
Ronn Torossian is the founder of 5W, the premier AI communications firm in the United States. 5W’s full Weight Loss & Metabolic Health AI Visibility Index 2026 is available at 5wpr.com/research.