Where AI Belongs, and Where Guests Now Find You

The triad still holds

In her piece for HSMAI, Zoba frames hotel marketing around the familiar triad of owned, earned and paid media. Her argument is simple and right. Treating AI as a new channel only produces more emails, more posts and more noise. The better use is to make the decisions inside each channel sharper: pulling scattered property information together, reading reviews for patterns before replying, weighing where the next marketing dollar should go. And some things stay human. Brand promises, sensitive guest moments and budget calls are not jobs to hand over. As she puts it: "people first."

The shelf has moved

There is a second shift happening at the same time, and it sits at the edge of the triad rather than inside it. More travelers now ask an assistant where to stay. And those travelers are ready to book. According to Adobe data reported by Forbes, visitors arriving at travel sites from AI answers converted 47% less often than other visitors in July 2025. By July 2026, that gap was down to 1%.

The same data shows the average hotel homepage is only 71% readable by AI. More than a quarter of the page, often the most beautiful part, never reaches the assistant a guest is asking.

An AI answer is assembled from what is written about a hotel: its own pages, its OTA listings, its reviews, its press. In other words, the owned, earned and paid work of the triad, read by a machine before a guest ever sees it.

Being cited is the new earned media

Being named in an AI answer behaves a lot like earned media. You cannot buy it. It is granted on evidence. That evidence comes from three places: clear, specific facts on the hotel's own pages, consistent descriptions across every channel, and reviews that use the guest's own language.

When a hotel does not tell its own story clearly, the assistant learns it somewhere else: from OTA listings and from reviews. Positioning has quietly become a question of source.

A story problem before a technology problem

Zoba writes: "AI can help express a genuine story. It can't make an inauthentic story true." The same holds for discovery. An assistant can only repeat what a hotel says clearly and consistently.

A homepage built on "unforgettable experiences" and "unparalleled luxury" gives it nothing to cite. Specifics give it everything: the beach the rooms face, the walk to the village, the chef's name, the time it takes to get there from the airport. This is the same work good brand storytelling has always required. It simply has a second reader now.

Three questions for the next planning cycle

1. If a guest asked an assistant about our destination tonight, would we be named, and would the description be right?
2. Do our website, OTA listings and sales materials describe the property the same way?
3. Which specific facts make us who we are, and are they written down anywhere a reader, human or not, can find them?

Owned, earned and paid remain the foundation of hotel marketing. AI belongs inside that work, as Zoba says. It also now stands at the door, deciding which hotels a guest hears about first. The hotels that tell a true, specific story will be found by both readers.