A traveller now opens a chatbot, describes a weekend, and an AI agent goes off and books the hotel without a human touching any part of it. Agentic commerce has arrived in hospitality, and somewhere a panel is being convened about what it means for the future of the industry.
Less is at stake in the guest experience than the panel thinks, and more is at stake in the delivery of hospitality technology than most operators have registered. Those are two different points, and it’s worth keeping them apart.
On the guest side: the booking was never the hard part. A machine can find the cheapest room in the city. It cannot read a table going quiet, rescue a soured evening, or make a tired person feel quietly looked after. That work was always human, and automating the transaction doesn’t touch it. So the panic about AI hollowing out hospitality is aimed at the wrong target. The transaction was the commodity. The relationship is the product, and it’s as human as it ever was.
The delivery point is where it gets interesting, because agentic commerce quietly rewrites the requirements for every hospitality technology project on the roadmap.
For years, hotel tech projects have optimised for one thing: the human booker looking at a screen. The website, the booking engine, the app, all designed around a person browsing, comparing, deciding. If a meaningful share of bookings now arrives via an AI agent acting for the guest, that design assumption is out of date, and a project scoped on the old assumption will deliver a system fit for a world that’s receding.
I’ve watched technology projects deliver beautifully against a requirement that had shifted underneath them while they built. It’s one of the more painful ways to fail, because everyone did their job well; the job just stopped being the right one. The defence is the same every time: interrogate the assumption behind the requirement before you commit, and keep checking whether it still holds.
So for anyone delivering hospitality tech right now, agentic commerce lands as a set of concrete questions about scope.
Is the system built to be read by machines, not just humans? Structured data, clean APIs, machine-readable availability and pricing. If an AI agent can’t parse your inventory, it can’t book it, and it’ll book the competitor it can parse instead. Optimising only for the human eye is now optimising for a shrinking channel.
Does the distribution strategy account for a new intermediary? Agents become a channel, like the online travel agencies before them, with their own logic about who gets surfaced and who doesn’t. A project that ignores how agents discover and rank properties is building a front door that a growing share of traffic will never use.
And crucially, does the technology free humans for the part that matters, or replace them in the part that doesn’t? The right target for automation in hospitality is the transactional grind, the check-in queue, the repetitive query, so that human attention is spent where it actually creates loyalty. A tech project that automates the wrong layer, stripping out the human touch to save cost, delivers efficiency and destroys the product.
That last one is a scoping decision, not a technology one. It’s a choice about what the system is for. Built well, hospitality automation handles the commodity so people can deliver the relationship. Built badly, it automates the relationship and leaves the guest talking to a machine at exactly the moment they wanted a person.
Agentic AI booking a room doesn’t worry me. What it should do is send every hospitality delivery lead back to their requirements to ask a plain question: was this scoped for the world that’s arriving, or the one that’s leaving? Because the technology has moved, and a project that hasn’t noticed is already building the wrong thing well.
Related: When a sustainability pledge becomes a requirement you can build to and Local sourcing is a supply chain with a conscience.
Ben Webb is an award-winning project leader and delivery strategist. He was named Australian Institute of Project Management Project Manager of the Year in 2022 and nominated for the IPMA World Project Manager of the Year in 2023.
Leave a Reply