AI is already doing a lot of heavy lifting when it comes to reservations and guest services. Where it gets more interesting is in its ability to predict the future, writes Ben Dixon, CTO and founder of Sona
When it comes to AI in hospitality, it can feel a bit like a revolution that is happening to everything, everywhere, all at once (to lean on a film title). It certainly feels the case if you listen to too many developers of this nascent technology.
Arguably, though, this coming transformation is barely anywhere just yet. As one prominent CEO told me recently: “We know we are going to have to invest in this technology in our businesses; that is no longer in debate. The problem is what, when and how. And we all know we really need help.”
What is clear is that the concept of AI serves up excitement and confusion in equal measure – with data from CGA by NiQ and Sona indicating that almost three quarters of hospitality business leaders believe AI will be an invaluable tool for their business, yet 59% also admitting to requiring further education.
In hospitality, not all AI is the same
The proliferation of AI-related terminology has created mass confusion, it seems. This is not helped by the many software partners that may have been busy rebranding existing technologies as “AI”, blurring the lines between genuine innovation and repackaged solutions.
One of the most significant developments in recent years has undoubtedly been the rise of more advanced AI models, far beyond basic chatbots, that genuinely learn. These more sophisticated AI systems possess the ability to analyze massive data sets, suggest or make independent decisions, and to even, in some scenarios, take action.
In hospitality, while AI is already doing a lot of heavy lifting when it comes to reservations and guest services, where it gets more interesting is in its ability to predict the future. Think of an AI that can build a hospitality forecasting model for each and every individual pub or restaurant within a group, based on thousands of variables, and continuously learn and develop its understanding of exactly how that specific unit reacts to all of those variables, to build ever-more accurate trading predictions for that location.
This sort of AI can act as an army of eyes in real time relaying important information as scenarios change, to trigger all sorts of adjustments in venue, including staffing levels, cook levels and service preparation. When combined with the expertise of a GM, it means hospitality venues can get increasingly closer to ending the gap between forecasted and actual sales.
The emergence of ‘true’ or ‘agentic’ AI in hospitality
What we are starting to see is a distinction between machine learning and modern, agentic AI, or ‘true’ AI. While machine learning excels at processing numerical data for predictive purposes, agentic AI goes further. It acts based on defined objectives, gathering information, making decisions, executing tasks, and adapting its actions based on outcomes.
Early AI applications in hospitality often focused on chatbots for basic guest or staff inquiries. For example, a simple HR chatbot could answer “how much paternity leave am I entitled to?” by retrieving company policies and employee records to generate a personalised response.
While these systems have become significantly more sophisticated in recent years – handling more complex requests and even personalised recommendations – fully autonomous agentic AI deployments remain rare; yet, the hospitality industry is poised to be an early adopter of these next-gen systems and tools.
Another rich seam of development for hospitality when it comes to agentic AI lies within loyalty and CRM programmes. We are facing into a future where AI will enable restaurants, pubs and hotels to be increasingly personal in their marketing and targeting of offers to guests, especially when it comes to their loyal followers. AI will enable marketers to target individual customers within their databases who are most likely to respond to an offer, most likely to book, most likely to bring their friends and most likely to spend the most, and so on.
Beyond workforce management, AI agents could optimise inventory levels, predict maintenance needs, and even manage energy consumption to reduce costs and improve efficiency. For restaurants, this could mean dynamic table management, personalised menu suggestions, and optimised kitchen operations.
These AI systems will likely provide recommendations for human approval at first, in the near term. However, as trust in the capabilities of this technology grows, we will start to see hospitality businesses moving along the automation curve, empowering AI to execute actions within pre-defined parameters. This has the potential to truly enable staff to focus on the guest and delivery of the experience.
Preparing for the Future of AI in Hospitality
Despite the hype, the reality is that agentic AI is still nascent in hospitality, to say the least. However, we know that business leaders are increasingly looking to this future and are of the mind that they cannot walk past the value AI has the potential to unlock, especially right now, in the wake of the significant cost headwinds that have just landed. We increasingly see leading hospitality businesses evaluating existing systems and identifying the functions where significant value can be driven, with the support of AI, to drive the team and guest experience, or cost benefits, revenue benefits or profit.
So, when it comes to AI in hospitality, it is not yet happening to everything, everywhere, all at once. But at the same time, it is certainly true to say, we have a feeling that we are not in Kansas, anymore.
Ben Dixon is chief technology officer and co-founder of Sona
