Forecasting the Next Menu

If the last decade was defined by bold predictions of what will happen—meat will disappear, farms will move indoors—the next decade requires a different mindset. We need to think in terms of dependencies.

Traditional forecasting asks: “What will the future look like?” This approach produces tidy timelines and confident predictions, but it consistently misses the mark because it ignores conditionality. A better question is: “What conditions must be met for X to happen?” This shifts us from prophecy to scenario planning. Instead of declaring that cultivated meat will dominate the market by some fixed date, we ask: What regulatory frameworks, cost structures, consumer acceptance levels, and infrastructure investments must all align for that to occur?

And what happens if even one of those conditions fails to materialise? This dependency-based thinking reveals not one future, but multiple possible futures, each hinging on whether specific variables fall into place. It’s less satisfying than a bold prediction, but far more useful for anyone actually trying to build, invest in, or adapt to what comes next.

The future of food is not a single destination. It’s a branching path. The success of emerging technologies depends less on their theoretical potential and more on specific economic, biological, and behavioural conditions being met. Instead of asking “what is the future of food,” we should be asking “what must be true” for these innovations to actually reach our plates.

The Appetite Suppressants: How GLP-1 Drugs Could Reshape Food Demand

The single most significant variable for the food industry right now is not agricultural. It’s pharmaceutical.

The widespread adoption of GLP-1 agonists—Ozempic, Wegovy, Mounjaro—introduces a wild card into consumer demand that could reshape grocery store aisles. These medications suppress what some users call “food noise” and grazing behaviour. Eaters on these protocols report a distinct lack of interest in high-volume, low-satiety foods like chips, soda, and candy.

The numbers are striking. In Australia, nearly 500,000 people—almost 2% of the adult population—are now using GLP-1 medications for weight loss or medical reasons, according to research from the University of New South Wales. Australian GLP-1 sales increased almost tenfold between May 2020 and April 2025. Globally, some analysts project the GLP-1 receptor agonist market could grow from around $65 billion in 2025 to over $170 billion by 2033. Early research suggests that among GLP-1 users, food and beverage purchases drop by an estimated 6–9%, which, given current adoption rates of 2–3% in most markets, translates to a 0.1–0.3% reduction in overall food consumption, with the US leading and Europe and Asia-Pacific following.

For the food industry, this creates a specific condition for success: the pivot from selling volume to selling density. If the “family size” bag becomes less appealing, the nutrient-packed single-serve becomes the opportunity. The winning products will be those that pack maximum nutrition and satisfaction into smaller serving sizes—high-protein yoghurts, fibre-rich snacks, nutrient-dense mini-meals.

There’s also a flavour dimension worth considering. If people are eating less, they may want each bite to matter more. We might see a shift toward more flavour-intense foods, bolder seasonings, richer sauces, more complex taste profiles. And here’s an interesting economic implication: if someone’s overall food consumption drops by 15–20%, their grocery budget doesn’t necessarily shrink by the same amount. They might simply redirect that spending toward more premium, higher-quality items. A person who used to buy a lot of cheap snacks might instead buy fewer, but nicer, meals. That’s a meaningful shift in where value gets created in the food system.

Companies like Nestlé are already responding, introducing lines of high-protein, high-fibre frozen meals. Danone and Coca-Cola are launching products positioned for this “medicated metabolism” consumer. Restaurants in the US are experimenting with “Ozempic menus” featuring smaller, nutrient-dense portions.

AI in Food Development: Useful, But Not Magical

While pharmaceuticals reshape demand, Artificial Intelligence is quietly entering how we formulate supply. AI arrived with a promise to revolutionise home cooking, but the everyday eater has shown little interest in having a chatbot design their dinner. The friction of cooking isn’t a lack of ideas. It’s a lack of time.

The more realistic impact of AI will be in industrial food development—what you might call “computational formulation.” This technology allows R&D teams to model millions of ingredient combinations in seconds rather than months.

But we should be honest about the limitations. Machine learning excels at tasks with quick, easily measurable feedback loops. Think of how a recommendation algorithm knows immediately whether you clicked on a video or not. That signal comes back in milliseconds. Recipe development doesn’t work that way. You can generate a recipe from an AI in seconds, but determining whether it actually tastes good requires someone to make it, eat it, and offer feedback—a loop that takes hours or days, not milliseconds. This makes AI a useful tool for narrowing possibilities, but not a replacement for human judgement and palate.

NotCo, the Chilean food-tech company, offers an interesting case study. Their AI platform Giuseppe analyses molecular structures to find plant-based ingredient combinations that mimic animal products, identifying unexpected ingredients like pineapple and cabbage for plant milk formulations. They’ve partnered with Kraft Heinz (launching plant-based mac and cheese and Oscar Mayer hot dogs), Shake Shack (eggless custard), and most recently Barry Callebaut, one of the world’s largest chocolate manufacturers. NotCo positions Giuseppe as a tool that can cut R&D timelines significantly. But even they acknowledge the AI works best when complementing human food scientists, not replacing them.

Where AI could prove genuinely valuable is in responding to supply chain volatility. The clearest application is in precision fermentation, where AI can optimise fermentation parameters—temperature, pH, nutrient timing—in real time based on sensor data. Unlike recipe development, fermentation provides continuous feedback loops measured in hours, not weeks. This allows machine learning to actually improve outcomes faster than human operators alone.

Regenerative Agriculture: Beyond the Marketing Hype

From the lab, we move to the field, where regenerative agriculture—farming practices that restore soil health rather than depleting it—has become the darling of corporate sustainability reports. But for the everyday eater, the term remains vague, often confused with organic or just “sustainable.”

Currently, the headlines about regenerative agriculture are outpacing the acreage. While major corporations have pledged to source from regenerative farms, the transition is slow, risky, and expensive for the farmer. The movement is stuck in what might be called “pilot purgatory”—plenty of small trials, but limited scaling.

Farmers operate on thin margins and cannot risk a yield drop during the transition years without financial support. Three things would actually scale adoption: reformed crop insurance that covers regenerative transitions, functioning carbon credit markets with real prices (not the current $10–20 per tonne that barely covers verification costs), and government subsidy programmes that reward soil health metrics rather than commodity production. Without at least two of these three, regenerative practices remain a niche.

The consumer reality is also key: will shoppers pay a premium for “soil health” when they’re already squeezed by inflation? Probably not, at least not in large numbers. Regenerative practices will likely need to reach cost parity to become the standard, moving from a marketing story to a baseline expectation.

Hybrid Proteins: The Pragmatic Compromise

Finally, the alternative protein sector is realising that purity might be the enemy of progress. The early vision was a binary choice: you either eat meat, or you eat a plant. But the future likely belongs to the compromise.

We’re seeing the emergence of products that blend plant-based proteins with cultivated animal fats. Since fat is the primary driver of flavour, adding just 5% cultivated beef fat to a plant-based burger could solve the taste challenge without needing to grow expensive muscle tissue at scale.

For this to succeed, the industry has to accept a more modest positioning. The “tech” side needs to recognise that they may be an ingredient supplier, not the centre of the plate.

The Path Forward

For the everyday eater, the changes will likely be subtle. The protein in their burger might be a hybrid of plant and cell. The flour in their bread might come from a regenerative farm. The formulation of their snack might be informed by AI to work with their metabolic needs. These changes won’t feel like a revolution. They will feel like a gradual evolution of the choices available on the shelf.

The implications of these dependencies are immediate and practical. If you’re a food company executive, your strategic planning needs to account for the pharmaceutical weight-loss adoption curve in your core markets. If you’re a restaurant operator, understanding the economics of hybrid proteins matters more than committing to all-or-nothing menu changes. If you’re a farmer, the viability of regenerative practices depends on whether carbon markets materialise with real money behind them, not just corporate pledges.

And if you’re an investor, the winners won’t be the companies with the best technology in isolation—they’ll be the ones that correctly read which dependencies will resolve favourably and position themselves accordingly. The next decade won’t reward those who bet on a single future, but those who build adaptive strategies across multiple possible futures. The branching path requires branching plans.

The companies that succeed in the next decade will be the ones that stop promising to invent a new food system and start doing the hard work of improving the one we have—one ingredient, one farm, and one meal at a time.