There’s been no shortage of conversation about AI in hospitality over the past 18 months. New tools, new promises, and a growing sense that something significant is changing.

But if you speak to most operators, the reality feels more grounded.

Margins are tight, teams are stretched, and systems don’t always talk to each other. Much of the day-to-day still relies on manual work, spreadsheets, and workarounds that have been in place for years.

In that context, AI isn’t really the starting point. Data is.

From fragmented systems to a single source of truth

Hospitality has historically grown in layers. New systems get added over time – POS, delivery platforms, menu tools, finance systems – each solving a specific problem, but rarely designed to work seamlessly together.

The result is familiar to anyone who’s run a multi-site operation. The same product exists in multiple places with different descriptions. Pricing gets updated in one system but not another. Teams spend hours rekeying information just to keep things aligned, and confidence in the underlying data is often lower than anyone would like to admit.

This is the foundational problem, because AI only works as well as the data behind it. Without a clean, connected foundation, it doesn’t matter how advanced the model is, the output will always be constrained by the quality of its inputs.

What’s starting to change is that more operators are building that foundation properly. A single, consistent view of products, menus, pricing and locations, shared across every channel – not as a reporting layer, but as an operational one. Once that exists, AI becomes far more practical.

AI in action: where it’s already making a difference

The most useful applications of AI in hospitality today aren’t futuristic. They’re operational.

Pricing is a good example. For years, most operators have worked on a handful of annual price changes, not because that’s optimal, but because anything more frequent is difficult to manage. It’s time-consuming, high-risk, and hard to coordinate across multiple systems.

AI changes that dynamic. With the right data in place, it becomes possible to model how different price points might affect both volume and margin, drawing on actual performance data rather than instinct or broad assumptions. One group operating across thirty-plus sites recently cut the time spent on pricing reviews by more than half, simply by centralising their product data and building decision rules on top of it. The insight wasn’t revolutionary; the speed of acting on it was.

The same logic applies to menu management. Updating menus across multiple channels has traditionally been slow and manual, often involving several teams and tools. A connected data model allows those updates to happen in near real time, reducing errors and ensuring consistency everywhere the customer interacts with the brand.

Over time, this starts to shift how decisions are made, less reactive, more predictive. Less about fixing issues after the fact, more about anticipating them before they affect the customer.

Agility becomes the advantage

Hospitality has typically operated in a fairly static way. Pricing set at a brand or regional level. Menus updated periodically. Changes rolled out in batches. That approach made sense when systems were limited and the cost of change was high.

It makes less sense in a world where demand can shift daily, shaped by weather, local events, or a change in customer behaviour that only becomes visible in the data weeks later.

What AI enables is a different operating model, one where changes can be tested in a small number of locations, assessed against real outcomes, and scaled where they work. Pricing adjusted in response to actual conditions rather than quarterly assumptions. Menus that evolve continuously rather than through periodic overhauls. Crucially, this approach doesn’t increase risk; it reduces it, because changes are smaller, more controlled, and easier to reverse.

Efficiency that frees up the people who matter

One of the more immediate benefits is straightforward: time. There is still a significant amount of manual administration across the industry – updating menus, reconciling pricing, and checking data between systems. It’s necessary work, but it sits at the bottom of the value chain.

By automating much of this, AI allows teams to focus on higher-impact work: commercial strategy, customer experience, and the decisions that actually differentiate a brand. That’s particularly important in multi-site operations, where consistency becomes harder to maintain as the estate grows and the operational surface area expands.

The results from operators who’ve made this shift are measurable: faster menu updates, improved pricing accuracy, and greater confidence in the numbers being used to make decisions. Perhaps more importantly, better alignment between what the business intends commercially and what actually happens on the ground.

The role of people doesn’t go away

There’s a tendency to frame AI as a replacement for human decision-making. In hospitality, that misses the point.

The sector is built on judgement, on understanding customers, on reading a room, on knowing when to follow the data and when to challenge it. What AI does is provide better inputs into that process. It surfaces patterns that might otherwise take weeks to spot. It models scenarios quickly. It reduces the effort required to test an idea before committing to it at scale.

But the decisions themselves, what to prioritise, how to position the brand, how to balance experience against margin, remain human. The operators who get the most value from AI will be those who treat it as infrastructure for better thinking, not a substitute for it.

Looking ahead: from optimisation to orchestration

The next phase of AI in hospitality is likely to be less about individual use cases and more about how everything connects. Personalisation will improve as businesses develop a clearer picture of customer behaviour across channels. Systems will become more integrated, reducing the friction between front-of-house, back-of-house, and digital platforms. Decision-making will continue to shift towards a more continuous, data-led rhythm rather than the periodic reviews that have defined the industry for decades.

At the centre of this is the idea of orchestration — bringing together data from across the business into a single model that can be acted on in real time. Menus, pricing, locations, channels: all of it connected, consistent, and available to the teams who need it. This is where platforms like Openr are starting to play a role, not as another system in the stack, but as the connective layer that allows AI to operate on solid ground.

The operators building that ground now aren’t just setting themselves up to be more efficient. They’re building the capacity to move faster, experiment more, and adapt to a market that has consistently rewarded those who can act on information before their competitors do.