A snack company in Minneapolis is redesigning its portfolio around portion size and protein density. An airline’s fuel-planning team is quietly revising its per-seat weight assumptions. A life reinsurer in Munich is rebuilding twenty-year mortality curves it thought were stable for a generation. None of these teams talk to each other. None of them work in the same industry. And all three are responding to the same injectable molecule.
That is the part of the GLP-1 story most executives are still missing. The obesity drug class built around semaglutide and tirzepatide, sold as Ozempic, Wegovy, Mounjaro and Zepbound, has been covered almost entirely as a pharmaceutical growth story or, at best, a food-industry disruption story. It is neither, or rather it is both and something larger besides: the first clean, fast-moving example of what might be called a biological demand shock, a shift in the underlying substrate of consumer behaviour, imposed on a dozen unrelated industries simultaneously by decisions made almost entirely outside those industries. Boards that filed this under “healthcare” have filed it in the wrong place.
A market too big to stay contained
The numbers alone explain why containment has failed. Roughly one in eight American adults is now on a GLP-1 medication, and estimates put the eligible population at some 137 million people in the US alone. Analysts expect the global obesity-drug market to roughly double from about $66 billion in 2025 to $120 billion by 2030, and food and beverage forecasters now project that GLP-1 users could account for more than a third of category sales before the decade is out. A Cornell study already finds GLP-1 households cutting grocery spending by roughly 6%, enough to move same-store sales numbers at scale.
The consumer packaged goods industry has responded the way industries usually respond to demand shifts: reflexively, at the product level. General Mills has leaned into protein-forward reformulations of Cheerios and new granola lines. Nestlé built an entire new brand, Vital Pursuit, aimed squarely at GLP-1 users. Danone reworked Oikos around muscle retention messaging, and Conagra now labels select Healthy Choice meals for appetite-suppressed shoppers who still want to feel like they’re eating well. Mondelez, notably, has waved the whole thing off as a rounding error against existing protein-snacking trends. That range of reactions, from full repositioning to studied indifference, is itself informative: nobody in the category actually knows the shape of the demand curve they are forecasting against, because the variable driving it sits entirely outside their own competitive set.
The industry built to be precise about this is choosing to wait
If any sector should have moved first and fastest, it is life insurance and reinsurance, an industry whose entire commercial logic rests on getting long-run mortality assumptions right. Swiss Re’s own modelling puts cumulative US mortality reduction from GLP-1 adoption at somewhere between 2.3% and 6.4% by 2045, depending on adoption and adherence scenarios, a spread wide enough to materially reprice decades of in-force business either way. Yet actuarial guidance from reinsurers including RGA is, at the time of writing, still counselling that it may be “too early to make material adjustments to insured trend assumptions.”
That is a defensible technical position and also a strategic choice with teeth. Waiting for certainty on a variable this large is not neutral: healthier policyholders who have lost significant weight are already positioned to lapse rated policies and re-underwrite at better terms, while the remaining risk pool skews toward those who didn’t adhere to treatment or couldn’t access it, a classic anti-selection spiral that punishes whichever insurer moves last rather than whoever moves first. An industry built on pricing tail risk is currently choosing to treat a well-quantified, multi-decade shift as noise, largely because updating pricing models before competitors do carries its own commercial cost. That is a governance failure mode worth naming precisely: the appearance of prudence can itself become the riskiest strategy in the room.
Winner-take-most economics in an industry that wasn’t supposed to have them
The competitive dynamics inside pharma tell an equally uncomfortable story. Eli Lilly’s market capitalisation has climbed past $1.1 trillion, roughly 60% of global prescription obesity-drug revenue now flowing through Zepbound and its stablemates, while Novo Nordisk, the company that effectively created this category with Ozempic, has seen its market value fall by more than 60% since mid-2024 to around $159 billion. Two companies still control something like 87% of the category between them, but within that duopoly, one execution edge, largely a more tolerable dosing and efficacy profile, has been enough to transfer hundreds of billions of dollars of value in under two years.
That is winner-take-most economics, the pattern executives associate with software platforms and network effects, showing up in a regulated, capital-intensive, patent-protected industry that boards had long treated as structurally diversified and defensible. The lesson travels well beyond pharma: single-molecule or single-product dependency risk, the idea that one clinical or technical breakthrough can carry or sink an entire multiple, deserves the same board-level scrutiny that concentration risk gets in any other capital allocation context. It rarely gets it, because “pipeline risk” still reads as a sector-specific R&D concern rather than an enterprise governance question.
India’s patent cliff previews how the next shock will actually spread
The most underappreciated part of this story, and the one with the clearest executive relevance beyond healthcare, is unfolding in India. Novo Nordisk’s core semaglutide formulation patent in India expired on 20 March 2026, and within a day, following favourable interim rulings from the Delhi High Court, Sun Pharma, Dr Reddy’s and Zydus had launched generic versions, priced 50% to 70% below the originator, under brand names including Noveltreat, Sematrinity and Obeda. Because Novo’s and Lilly’s patents hold in the US and EU until roughly 2031 to 2032, India and other patent-lapsed markets across South Asia, Africa and Latin America are now on an entirely different adoption curve to the West, gated by affordability and diagnostic infrastructure rather than insurance coverage and list price.
That split matters strategically well beyond pharma margins. Any company modelling “global GLP-1 penetration,” and by extension global demand for weight-related categories from apparel sizing to travel to snack food, off a single geography’s curve will misforecast in both directions: overestimating how fast lower-income markets convert, underestimating how disruptive a 70%-cheaper, domestically manufactured version becomes once it exists. It also hands Indian pharmaceutical manufacturers a genuinely new strategic asset, not just production capacity, but the ability to act as the access valve that determines how fast a molecule class reaches more than a billion consumers, a role that will invite exactly the kind of litigation and geopolitical friction Novo Nordisk is already pursuing through further court challenges.
The demand model, not the drug, is the actual disruption
Airlines offer perhaps the cleanest illustration of how far the ripple travels: analysts now estimate GLP-1-driven weight reduction across the flying population could save US carriers on the order of $580 million a year in fuel costs alone, a line item nobody in aviation strategy would have modelled five years ago.
More molecules like this are coming, in longevity, addiction and cognition, and each will propagate unevenly across markets exactly as semaglutide has, fast where patents lapse and payers cover it, slow where they don’t. The companies that treat this as a one-off pharma curiosity will keep discovering the impact a category at a time, in the same reactive sequence as General Mills, then Munich Re, then Boeing’s fuel planners. The ones that build a standing capability for biological demand shocks, forecasting off real cohort adoption data rather than historical trend lines, treating molecule-class penetration as an explicit scenario input outside health and consumer categories, and watching patent geography as a leading indicator rather than a legal footnote, will be the ones not still relearning the same lesson when the next molecule arrives.
