Is AI Being Overhyped Or Misapplied In Hotels?

“The answer here isn’t black and white — and that’s exactly where the industry keeps getting it wrong. There’s a category of hotels where full digital interaction is the product. Budget, extended-stay, student housing concepts — guests chose that property precisely because they don’t want or expect a traditional service layer. AI handling the entire guest journey there makes complete sense. It’s not a compromise; it’s the promise.

But most hotels aren’t that. Most are trying to use AI to improve service — and that’s where the misapplication creeps in. The mistake isn’t using AI in emotional moments. The mistake is letting it be the final stop. A guest messaging at midnight because the room next door is loud doesn’t need an AI that understands the problem. They need an AI that recognizes this is the moment to hand off — fast, warm, without friction — to someone who can actually fix it.

The properties getting this right aren’t asking “can AI handle this?” They’re asking “at what point in this conversation does a human need to take over, and how do we make that transition invisible to the guest?” That’s a design question, not a technology question. Build the escalation path first. Then layer the AI on top.
The ones getting it wrong have escalation as an afterthought — or worse, no clear escalation at all. And guests notice. Not immediately, but in the review, three days after checkout.”

“The biggest misapplication is AI pointed at the guest when it should be pointed at the operation. Hotels buy guest-facing chatbots and call it transformation. A bot answering checkout questions solves a problem that the FAQ page solved years ago, while the back office loses hours every day to manual reporting and rate research. That is where agents create real return.

The overhype starts at the vendor booth. Press most products claiming AI and you find a logic tree firing canned responses. A CDP claiming it knows your customer better than ever is often running a complex if-statement.

Ask four questions:

1. Is a model reasoning or a logic tree firing?
2. Can it pursue a goal without constant prompts?
3. What happens when it is wrong?
4. Show me a live deployment at a hotel like mine.

Most pitches collapse by question two. The most expensive lesson the industry has not admitted is that data infrastructure determines AI outcomes. Hotels buy tools to sit on fragmented systems and undocumented processes, then wonder why pilots stall. AI amplifies whatever sits underneath it. The honest sequence is clean the data, document the processes, map the integrations, then buy. The industry keeps running it backward and paying twice.”

“I believe AI is sometimes applied simply because it is the latest trend, rather than because it solves a real problem for either the guest or the hotel. We should use it for our benefit, to simplify tasks, create productivity, but not fully rely on it in all areas.

One area where I think AI can be used too much is guest communication. While AI can be very effective in answering simple questions or handle routine requests, depending on the type of hotel and thus guests, you still need properly trained staff to handle the human interaction. So, while staff training in communication, empathy, problem solving skills, and how to handle unexpected situations is already very important, it might be even more important now since it can be too easy to fully rely on systems and become a bit more automated in service as well.

I believe we should understand our hotels, guests, and our strategy in attracting those guests, and then see how we can use AI tools to make that purpose easier. There are hotels with full automation where it fits the type of guests, for example airport / highway locations where time is of the essence, and guests just simply need a room and a bed.

In my opinion, the hotels that will benefit most from AI are not necessarily the ones investing the most money in technology, but the ones that understand where technology adds value and where human judgment remains essential.”

“Certainly an area where there has been excessive hype or at least a risk of misapplication is that of the personalisation of stay experiences.

The border is very blurred. On the one hand, AI can help personalise suggestions of experiences or services at a hotel based on certain segmentation and profiling characteristics (nationality, age, segment: couple, family, etc.), if it is a new customer, or preferences expressed during previous stays if it is a returning customer (provided that the hotel uses a system capable of tracking the guest’s preferences).

On the other hand, there is also the risk that the guest may perceive an invasion of privacy, you have to move very carefully. Then it must be said that each person is different and their tastes and needs can change over time, so personalisation based only on experiences or preferences expressed in the past could be misleading.

Gastronomic and leisure experiences in particular are likely to be very subject to variability, so in this case the ideal thing would be to propose something similar to what the guest has expressed in the past but at the same time propose alternatives that inspire the guest to explore something different. On the other hand, on more habitual aspects (types of pillows, room amenities, types of curtains, temperature, etc.) predictive personalisation can generate the best results in terms of guest satisfaction.”

tbc

“The hospitality industry is currently repeating its historical mistake of mistaking novelty for true business value. The most egregious misapplication of AI today is the reckless race to inject generative AI, avatar concierges, and front-facing guest-journey automation into luxury, resort, and five-star properties. Forcing a guest to navigate an algorithmic chat interface to request a late checkout or order room service solves a problem that did not exist, while actively damaging the premium brand promise. In the high-end tier, technology must remain an invisible backend architecture. Its core purpose is to connect data silos and orchestrate workflows, freeing human staff to deliver personalised, face-to-face engagement. Complete front-of-house automation belongs strictly in select-service and extended-stay environments where customers explicitly prioritise transaction velocity over high-touch hospitality.

The most expensive lesson the industry refuses to admit is the crushing capital drain of deploying fragmented, poorly integrated AI tools on top of legacy property systems. Operators are burning multi-million-dollar tech budgets buying shiny, isolated applications that create massive structural friction and dirty data reservoirs. True ROI requires an integrated technology change management strategy that treats platforms as long-term enterprise assets. Until brands stop chasing overhyped front-facing novelties and focus on backend data integration and infrastructure reality, they will continue to realise negative returns on their capital investments.”

“One of the biggest mistakes is assuming that more AI automatically creates a better guest experience.

Many hotels are replacing meaningful interactions with generic AI experiences that lack empathy and context. AI can answer questions quickly, but it cannot fully replace human hospitality.

Another area is content creation. Publishing large volumes of AI-generated content without unique expertise or local insights rarely creates a competitive advantage.

The most expensive lesson the industry hasn’t fully admitted yet is that AI cannot fix poor strategy. If data, processes, and guest experiences are broken, AI simply scales those problems faster.

Hotels should focus less on where they can use AI and more on which guest or business problem they are trying to solve.”

“In my opinion, AI is most often misapplied when hotels adopt it because it appears innovative rather than because it solves a meaningful problem. The industry sometimes focuses heavily on visible guest-facing applications, such as AI concierges, virtual avatars, or highly personalised marketing messages, without first asking whether guests actually value those features.

One area that may be overhyped is the assumption that all AI applications are good and useful. While AI can improve efficiency, many hospitality experiences still depend on empathy, judgment, and relationship-building. Automating these interactions may seem appealing at first, but it could hurt the business if it does not improve guest satisfaction. For example, replacing front-desk agents with robots may generate attention, but not necessarily better guest experiences. In fact, some hotels that experimented with robot receptionists later reintroduced more human interaction into the check-in process.

The most expensive lesson the industry has not fully acknowledged is that AI adoption needs careful planning and is more than just a technology issue. Poorly implemented AI can introduce new costs, including employee resistance, guest frustration, data quality issues, and reputational risks. In some cases, hotels invest in AI tools without redesigning workflows, training employees, or establishing accountability structures.

Overall, organisations must align AI initiatives with specific operational goals and use technology to augment and empower employees. AI adoption and innovation should be measured by business outcomes and value creation, not simply by the number of AI tools deployed.”

“One of the most common misapplications is the assumption that guests want fully automated interactions throughout their stay. While guests appreciate speed and convenience for routine tasks, hospitality remains a people-centric industry. The majority of the guests want empathy, ownership, and a human solution. In these moments, technology cannot replicate emotional intelligence and relationship-building.”

“The biggest misapplication I see is using AI as a guest-facing mask — deploying chatbots and automated responses to look innovative, when what guests actually want is to feel looked after. When chatbots try to sound human and fall short by giving inaccurate information or behaving inconsistently, users don’t just get annoyed — they feel unsettled, and they disengage. In luxury, that’s not just a conversion problem. It’s a brand problem.

There’s also a lot of noise around AI “personalisation” that is really just slightly smarter email segmentation. Calling that personalisation is a stretch. Genuine personalisation means a returning guest feels that the property already knows them, not that they’ve been placed in the right demographic bucket.

The expensive lesson the industry hasn’t fully admitted: hotels are deploying visible AI features without redesigning the underlying operations — what some are calling “AI Theatre” — creating the impression of transformation without changing the actual guest experience or business metrics. The technology becomes the story, when the story should always be the place, the people, and the stay itself. AI should be working quietly in the background. If your guests can see it, ask yourself why.”

“I believe agentic AI and AI platforms as a booking channel have been overhyped. Yet, Expedia’s latest survey shows that only 8% of travelers are comfortable booking a hotel on an AI platform.

I am sure this percentage will go up, the question is when are we going to see that? Not in 12 months, that’s for sure. Don’t forget travel consumers’ inertia and loyalty for doing things they are sure to work based on their past experiences. It took the Internet how many years to be adopted as a booking channel? Yet, today less than 60% of hotel bookings are booked online. 99% researched online, only 60% booked online.”

“One area where AI is being misapplied is the push toward fully contactless check-in. In the drive for efficiency, many hotels have removed one of the most valuable moments in the guest journey: the arrival conversation. Check-in isn’t just administrative, it’s a key opportunity to understand why a guest is travelling, identify upsell opportunities, and start building a relationship.

By automating this completely, hotels are often missing the chance to turn first-time visitors into repeat customers. For example, without a quick interaction, a hotel may never learn that a guest is celebrating a special occasion or is open to recommendations, which could otherwise lead to upgrades, dining bookings, or curated experiences.

RevenYou is already seeing a shift away from fully contactless models toward hybrid check-in. These approaches combine automation for efficiency with targeted human interaction to capture insight and drive revenue. The industry is realising that removing all friction can actually remove value.

The most expensive lesson the industry has not fully acknowledged is that in chasing labour savings, many hotels have unintentionally sacrificed revenue and loyalty. AI works best when it removes low-value tasks, not when it replaces high-value human moments that drive connection, spend, and long-term guest relationships.”

“Personally, I see guest-facing AI as one of the biggest ones. A guest who wants extra towels doesn’t want to chat with a bot, they want towels. These tools get funded because they demo well, not because guests asked for them. The most overhyped is fully automated pricing. Hotels quietly abandon these systems within months because nobody can explain to the owner why the rate dropped 18 percent on a Saturday. I believe the industry built a decade of automation before building the explanation. That’s a trust problem, not a change of management problem.”