Stop reacting to food cost volatility and start absorbing It

Oct 7, 2026|5:34 pm BST

Hospitality operators are playing ‘whack-a-mole’ with operating costs – except none of them go down. Food inflation forecasts keep rising, energy prices are on the up, and the UK now faces a potential 82% decline in CO2 supplies, with knock-on consequences for packaging shelf life and livestock processing. The tax burden on the hospitality industry is astronomical;  big names are calling the current climate “crippling, and that’s on top of the £1.9 billion in wage costs the sector has had to absorb over the last year.

Unfortunately, it’s nothing new. Most operators have been managing versions of this for years now, the difference today is that volatility isn’t a phase, it’s the status quo. Restaurant teams have to make more decisions, more quickly, with less certainty. And we need to find a way to make that work.

Reacting Isn’t Success

The dominant model for managing inventory pressure in hospitality has been to push accountability down the chain. GMs are told to control stock more tightly, hold less, reduce waste, and keep food costs within a set budget, without real insight into what that should look like.

Venues are running on instinct, with limited autonomy and undefined guidelines — which means most restaurant operations have no choice but to be reactive. If a delivery is short, a manager has to improvise; if olive oil prices go through the roof, they don’t have the authority to order an alternative, and the impact won’t be noticed centrally until the P&L is reviewed at the end of the month. In this system of devolved responsibility and unprecedented volatility, no one is set up for success.

What Does Absorbing Volatility Look Like?

The link between electricity prices and volatile gas markets needs no explanation. To combat it, the UK government recently moved older wind and solar farms onto fixed-price contracts specifically to uncouple the two. It’s a good lesson; if a price shock is structural, redesign the system to absorb it rather than react to it. The same logic applies to inventory.

That redesign means reacting centrally without overriding site-specific knowledge. This is achievable with a central intelligence layer, an artificial intelligence (AI) that connects inventory, sales, and demand data across the whole estate and highlights problems in real-time.

Rather than each site managing its own purchasing decisions, central teams set the guardrails around approved substitutions, cost thresholds, and ordering parameters. The AI then alerts a decision-maker immediately when a situation requires additional input, such as a supplier price movement or anomaly in stock levels. In practice, this means GMs receive an alert early enough to act on it, rather than leaving them buried in the back office managing an avoidable crisis when it hits.

Making It Work in Practice

Building that AI-enabled inventory management system on top of existing operations requires three main components:

1. Demand Forecasting: Orders should be driven by forecasts, not historical trends. If you know what you’re selling this week, you can order what you need and avoid tying up cash holding stock orpaying for unnecessary waste. Automating this process with AI saves managers hours in the office and changes their role to sense-checking the order against the forecast and centrally established rules, and then actioning it. The same forecast should used to inform how labour is deployed, otherwise operators risk tightening food cost and then overspending with the wrong number of people on shift to prep and serve it.

2. Dynamic Sourcing: When a price shifts or a delivery comes in short, managers need quick access to alternatives. An automated system that can surface changes quickly and shop around saves an inordinate amount of time. It also enables central teams to set approved substitutions and supplier rules in advance, so sites are working within a framework rather than improvising when something doesn’t go to plan.
3. Suggested Ordering: AI-driven Demand Forecasting and Dynamic Sourcing tell you what you’ll sell and where to buy it; Suggested Ordering tells you how much and ensures you are only buying what you actually need. In climates with high prices, over-purchasing is particularly pricey, and operators pay excessively per unit and then in waste if the item doesn’t sell. Using AI to forecast demand and identify exactly what you need to buy when ensures venues run at maximum efficiency.

Managing Volatility Now

Volatility was cyclical; we used to be able to ride out shipping disruptions, energy price spikes, or stock availability, adjusting once conditions settled. That’s no longer the case, and the cost pressures hitting hospitality are complex, structural, and here to stay.

Operators who treat the current conditions as temporary will continue to absorb damage that a better-designed system would deflect. Those who have accepted that volatility is the baseline are building operations that not only withstand it but are capable of growing through it.