When the era comes where AI selects and recommends hotels instead of travelers, the hotel distribution structure will be fundamentally overturned. Hotels can no longer just focus on ranking high in search results and OTA listings; the actual guest experience and level of operation itself become their most powerful commercial strategy. Joe Pettigrew emphasizes that hotels do not need to be AI experts, but they must create great, differentiated products and experiences that AI will recommend.

1. Travelers already trust AI

Host Jason Noronha invites Joe Pettigrew, who has worked in hotel commerce for 20 years, to discuss how AI will change the way hotels operate, discoverability, and distribute. Jason says this may be the most exciting time in his travel industry career, but Joe immediately responds:

"It's scary. It's also scary."

AI has the potential to broadly change not only hotel operations but also the way travelers choose destinations, find hotels and restaurants, and plan their itineraries. Joe points to several studies showing that 40-80% of travelers are already using AI to plan their trips. The numbers vary depending on the research organization and time, but either way, it means that a significant number of people are already entrusting their travel plans to AI.

He believes that consumer trust in AI has formed faster than expected. Travel planning, in particular, is an area where people can ask AI questions relatively easily because it is not an extremely sensitive area like health or finances. Joe cites his experience with his three-year-old daughter as an example. When a child develops blisters on their head or sneezes continuously, they ask questions to AI before asking a doctor.

"An AI agent knows more about my three-year-old daughter than any doctor or pediatrician in the country."

Of course, this does not mean that AI will actually replace doctors, but rather that AI has become the first source of information people turn to in situations where they are curious. If you trust AI to that degree, travel planning becomes a question with a much lower psychological barrier.

"If I am already using AI for my daughter's health issues, travel is an area where trust can be lower."

In the end, the important question is not whether people trust AI, but which hotels the AI ​​recommends. AI can suggest to travelers where to go, what hotel to stay in, what activities to do, where to eat, and even how to plan their travel itinerary. For hotels, a new commercial challenge is "How can we make AI recommend our hotels, restaurants, and services when asked the right questions?"

2. Hotel distribution funnel is turned upside down

The current hotel distribution structure generally starts with being exposed to the list. This is done by entering Google search results or OTA lists such as Expedia and Booking.com, ranking high within them, and decorating photos and descriptions in attractive ways.

Joe explains the existing structure as follows:

1. Enter the search results or OTA hotel list.
2. It ranks high on the list.
3. Optimize photos and descriptions to attract clicks.
4. The customer stays at the hotel.
5. Customers leave reviews and ratings about their experiences.
6. The rating again affects search rankings.

Previously, awareness and exposure were at the front of the funnel, and customer experience was at the back of the funnel, occurring after the stay. Customer experience was summarized in numbers such as "8" or "9.5" and reflected as one of several ranking factors.

"As a commercial strategist for hotels, I've worked to get them on the list and get them ranked high on the list."

However, this structure changes when AI processes travelers' questions instead. Instead of allowing travelers to check out 1,000 hotels in London, AI first interprets the conditions and filters out candidates. We internally evaluate hundreds or thousands of hotels based on conditions such as location, price range, child-friendly availability, breakfast, parking, and room type, and ultimately recommend only a few.

"AI can't recommend 50 hotels to a traveler. In reality, it should only recommend about three hotels to consider."

AI first narrows down the hotels that meet the criteria to about 50, and then re-evaluates which hotels within them are actually differentiated. And finally, I recommend two or three places. At this time, rather than knowing and comparing all hotels from the beginning, the traveler only recognizes for the first time hotels that have already been verified and selected by AI.

Joe calls this phenomenon 'Inverse Distribution Theory'. Previously, customer experience was at the end of the distribution funnel, but in the future, customer experience will move to the front of the funnel.

"We used to think customer experience was at the end of the funnel, but now it's almost at the front of the funnel."

In other words, by the time AI recommends a hotel, many hotels have already been eliminated. A hotel's reputation is not achieved at the beginning of the search, but only occurs when AI determines that "this hotel is recommendable."

3. Human choices are replaced by AI qualifications

Jason points out the difference between existing human-centered search and AI-centered search. While looking at numerous hotel lists, people are influenced by the color of the photos, representative images, prices, first impressions, etc. If you're tired, you may roughly choose the first or third hotel on the list.

"People click if they like the color of the photo and the representative image looks good."

However, AI is not greatly influenced by simple visual impressions or fatigue like humans. It checks many more variables simultaneously and evaluates relatively consistently whether it meets the conditions of a specific customer.

Rather than choosing a hotel just by looking at the mood of the photo, AI can also look at the following factors:

-Customer's specific requirements

• Location and price range
• Whether accompanied by children
• Rooms and facilities
• Real customer reviews
• Evaluation of the press and travel media
• Differences compared to other hotels
• Whether the services promised by the hotel are actually provided

Therefore, the strategy of using loopholes in the algorithm to rise to the top of the list, as in the past, becomes increasingly difficult.

"Now you can't play the list game and fool the algorithm to get to the top."

Jason jokes about this, saying, "Shouldn't we be targeting AI now, not algorithms?" Joe replies that to fool an AI, you need actual content rather than just packaging.

"If you want to get to the top of AI, you have to provide substance, not gimmicks."

This change may be burdensome for hoteliers, but it could also make the industry as a whole more honest. Before the advent of the Internet and OTAs, a hotel's reputation was built on service and word of mouth. A good location, high-quality products, and excellent service attracted customers. Travel agencies, recommendations from friends and family, and reviews from travel guidebooks were the main means of distribution for hotels.

Joe recalls the days when being featured on Lonely Planet could determine the fate of a hotel.

"Whether I was buried or featured on Lonely Planet could have determined my business for three years."

Today, hotels are constantly evaluated through OTA reviews. In the AI ​​era, that evaluation can be made virtually at every moment.

"If a good hotelier provides good service, people will eventually come."

AI synthesizes signals such as customer reviews, media reports, and ratings from authoritative travel media to determine which hotels are actually good. If all hotels meet similar criteria, the final recommendation goes to the hotel that offers a better experience and actually stands out from the crowd.

4. Strengths of the owner/operator model

Jason notes that Joe's hotel business is structured as an owner, operator, and brand. In a franchise or general hotel operating structure, there are usually three main stakeholders.

1. Owner who owns hotel assets
2. Brand company that manages the brand
3. Management company in charge of actual operations

They may each have different goals. Owners focus on profitability and asset value, and brands try to maintain consistent brand standards across multiple regions and hotels. Operators may pay greater attention to sales or commission structures.

"There are three main stakeholders in a hotel: the owner, the brand and the operating company, who may have competing interests."

Conflicts of interest can slow decision-making, prolong arguments, and potentially lead to suboptimal compromises in technology adoption, organizational structure, and sales strategies.

On the other hand, if the owner, operator, and brand are combined as one, the experience can be designed according to the hotel's location, assets, customer base, and employee composition. The need to compromise to simultaneously satisfy the standards of other brands, the needs of the owner, and the profit structure of the operator is reduced.

"Because we are the owner, operator and brand, we can look at the customer experience and product quality from a blank slate and execute on what is best for the hotel."

In the AI ​​era, this flexibility becomes more important. If customer experience becomes a core criterion for AI recommendations, an organization that can quickly decide what experience a hotel will provide and reflect it in actual operations will have an advantage.

Joe explains that standardization among big brands doesn't necessarily mean it's a bad thing. However, applying the same standards to all hotels can make it difficult to create unique experiences tailored to specific regions and customer groups. The owner-operator model has the advantage of being able to design services according to the "framework" of the hotel, local conditions, and customer characteristics.

5. Which department should be responsible for AI strategy?

Jason asks who should be responsible for a hotel's AI strategy: marketing, IT, commercial, or operations. This could be for marketing because AI impacts searchability and discoverability, or it could be for IT because it has to deal with website data and technical structures. However, the actual customer experience is created by the operations team.

Joe responds that AI should be viewed as a collaborative performance improvement function rather than as a standalone project for one department. As the focus is on leveraging AI to increase hotel sales and asset profitability, the initiative must lie with those responsible for commercial performance.

"I see AI and AI discoverability as a function of maximizing performance. Ultimately, it is a commercial function."

However, the role of the operation team is also very important. This does not mean that the operations team must have a deep understanding of AI technology. Instead, in the AI ​​era, the quality of products and services provided by operations teams becomes much more important than before.

"It's not that operations teams need to understand AI, it's that what they provide is more important than ever before."

Using AI, tasks such as answering phone calls, receiving customer requests, and delivering requests to the appropriate system and person in charge can be automated. However, the core of operations remains improving the customer experience. For example, increasing the repeat guest rate by 5% may be an important goal for the operations team.

In the end, the structure is as follows.

• The commercial and revenue management team is responsible for how AI increases sales and NOI.
• Operations Team creates exceptional customer experiences that can be supported by AI.
• IT Team manages data, security, systems, governance and control.
• Marketing Team ensures that the hotel's reputation and content spreads to external information that AI can refer to.

AI is not a function monopolized by one department, but becomes a joint task to improve the performance of the entire hotel.

6. You can start AI work even without technical knowledge

Joe and Jason say one of the biggest changes in AI adoption is the diminishing need for technical expertise. In the past, linking systems required understanding programming languages, APIs, and complex integration structures. However, modern AI tools can organize tasks by giving instructions in natural language.

"You don't have to write a single line of code."

By explaining the desired results and rules to the AI ​​model, the AI ​​can ask back what information it needs and adjust its work process. Business rules that had to be written in complex code in the past can now be managed in the form of simple text documents.

"Deploying an agent doesn't require any secret code behind it. Just create a file and write instructions in it."

Joe recommends that hotel employees first install the AI ​​app themselves and have them perform actual tasks beyond simple questions. For example, it's not just about asking a simple chatbot a question, it's also about having the computer perform repetitive tasks.

Even if the AI ​​can't find the API document, it can open a browser, log in to the website, and perform the task. This function means that tasks that were previously thought to be "not compatible" or "not supported by the vendor" can be automated.

Jason introduces real-life examples. I had to check my ranking on a hotel's small OTA list every day, a task I thought would be difficult to automate due to the lack of an API. He scheduled the following tasks in his AI desktop tool:

"Go to the relevant OTA every day at 9 am to check the ranking of our hotel and let us know where it is exposed."

Now, employees no longer have to check the list themselves while drinking coffee every day. You only need to receive notifications and respond when your ranking drops.

There are many repetitive tasks like this in hotel operations. For example, you can automate the transfer of virtual credit card information included in OTA reservations from one window of the system to another.

"You can automate that task, moving credit cards from one window to another and even leaving a note saying it's done."

Therefore, the first step in introducing AI is not planning a huge project. First, employees find repetitive and annoying tasks in their work and leave them to AI.

7. The financial impact of AI first appears in operational efficiency

Jason asks which metrics will be most effective when starting an AI project. Several indicators can be candidates, including market share, channel mix, proportion of direct reservations, and sales.

Joe says that, surprisingly, effects are likely to be revealed in terms of operating efficiency and costs before sales.

"I think the effects of AI will first appear in operational efficiency."

That's because AI can consolidate customer requests, automate repetitive tasks, and better understand demand and staffing. For example, even if a customer sends a request through multiple channels such as phone, message, or front desk, AI can gather them all in one place and connect them to the appropriate workflow.

This reduces the need for customers to wait for the front desk to answer the phone. Even if you order a burger late at night, or if a hotel employee makes a request in an unfamiliar language, AI can immediately understand the content and connect the next steps.

"Customers no longer have to wait for the front desk to answer the phone."

AI can also reveal blind spots in operations. Hotels usually know how many phone calls they receive and how many room service orders are placed per day, but they often cannot systematically analyze what exactly the calls were about or what requests are made at specific times.

When AI analyzes this data, the following decisions become possible.

• How many people should be assigned to the night reservation department?
• When are late-night room service orders really concentrated?
• Do you need a room service delivery person at night?
• Is there a possibility of a surge in order volume due to a specific event or change in customer composition?
• Are you unnecessarily securing excessive inventory?
• Can repetitive phone calls or administrative tasks be automated?

As AI improves the customer experience while deploying personnel and costs more precisely, the effects can first be seen in income statement cost items and NOI.

However, the more important indicator in the long run is not cost reduction. The ultimate goal is for AI to inform more customers of the hotel's good experiences, bring in new customers through recommendations, and increase existing customers' return visits and recommendations to their acquaintances.

"More important in the long term is increased referrals from new guests, increased repeat visits, and digital word-of-mouth as guests recommend the hotel to their family and friends."

8. Customer experience becomes the most powerful marketing

Jason concludes that the key ultimately lies in actually improving the hotel's products and services. Because AI makes the connection between hotels and customers more transparent, it can easily reveal the hotel's weaknesses.

"AI will very easily expose the hotel's lack of experience."

In the past, you could entice customers to click through great photos, descriptions, high rankings, and advertisements. However, when AI synthesizes multiple sources and real-world experiences, simple promotional statements are not enough. If there is a significant difference between customer expectations and actual experience, it is immediately reflected in reviews and external evaluations, and also affects the likelihood of AI recommendations.

There are several ways for AI to notice when a hotel improves the experience.

Customer Reviews

Customers leave their actual experiences on external platforms such as OTAs, Google, and TripAdvisor. These reviews are more reliable than those written or controlled directly by the hotel.

"What other people say about a hotel will become more important than what hotels say about themselves."

Internal surveys or recommendations posted on the hotel's own website may gradually become less influential. This is because AI is likely to rate the hotel's praise for itself relatively low.

Authoritative media and expert evaluations

Joe says not all reviews will be treated the same. How famous travel media, professional reviewers, and authoritative media evaluate a hotel can have a greater impact than a typical 5-star review.

"AI can evaluate an opinion from an authoritative media outlet differently than a typical 5-star review on Booking.com."

Therefore, hotel commercial teams need to go beyond simply collecting their own reviews, help actual guests leave their experiences on external platforms, and promote the hotel's experiences to trusted travel media and content creators.

Virtuous cycle of operational improvement

Good operations increase customer satisfaction and generate positive reviews and media coverage. This information becomes the basis for AI to recommend hotels. New customers visit with AI recommendations, and when they have a good experience again, additional reviews and word of mouth accumulate.

In this way, a virtuous cycle of operational improvement → improved customer experience → increased external reputation → AI recommendation → influx of new customers is created.

9. YouTube, the channel that hotels are missing

During the conversation, the two point out that while social media like TikTok and Instagram are often talked about in the hotel industry, YouTube may actually be the most underrated channel.

When it comes to social media in hotel marketing, Instagram usually comes to mind first, followed by TikTok. But Joe says the channel where he saw real, tangible, verifiable effects was YouTube.

"The most important and most underrated social media channel in our industry is YouTube."

YouTube is the world's second largest search engine, where videos and text can be searched and analyzed. Given that the AI ​​model can reference YouTube content and is connected to ecosystems such as Google and Gemini, it can play an important role in hotel discoverability in the AI ​​era.

"YouTube is the second largest search engine in the world and is fully indexed and analyzable."

Hotels can also use content posted on Instagram on YouTube. You don't necessarily need a long video or a review from a famous influencer. Short formats such as YouTube Shorts are also possible, and content produced by the hotel itself can also be used.

However, the reason the hotel industry does not actively utilize YouTube is due to the inertia of marketing agencies and organizations. Many agencies are built around Instagram, and owners and executives don't fully recognize the value of YouTube.

"If a hotel is creating content for its own social media, why should it only be posted on Instagram? It should also be posted on YouTube."

If AI's ability to understand videos and images continues to improve, hotel-related content accumulated on YouTube could become important data for AI to understand the actual appearance of the hotel and customer experience.

10. Practical methods to create an AI-ready hotel

Jason asks what it takes to start to make a hotel truly AI-ready.

Joe recommends first going beyond the typical ChatGPT chat window and directly using AI tools that can perform tasks. For example, by installing Claude or a modern desktop-based AI tool, you can have your computer do real work, rather than just get answers.

"Most people still think of AI as a chat window that provides answers to questions."

However, AI can now perform tasks such as:

• Log in to the website and check information
• Execute repetitive tasks at set times
• Reading and organizing multiple documents and materials
• Automate employee work procedures
• Classify customer requests and forward them to the person in charge
• Enter data into the system and record completed results

The first step is for employees to find repetitive and boring tasks in their work and leave them to AI. You can then have more specific conversations with specialized vendors or technology companies to find out what tasks your hotel actually needs.

"First you have to understand what's possible with AI, and then finding an expert to address the needs of your specific hotel becomes a much more high-level conversation."

In other words, hotels do not need to order large-scale AI projects from the beginning. It is important for employees to feel the potential of AI, starting with small automation.

11. Standardization and hotel individuality can go together

There are also concerns that standardizing AI work when operating multiple hotels may result in the individuality of each hotel being lost. If the data and processes of all hotels are made the same, the unique characteristics and regional charm of each hotel may be diluted.

Joe answers that a certain level of data and process standardization is virtually essential to perform meaningful AI work. In order for AI to process information from multiple hotels, the way rooms, requests, tasks, and systems are called must be somewhat unified.

"Meaningful AI work almost requires standardization."

The purpose of standardization is not to force the same experience on customers. The key is to organize data structures and internal processes to prevent AI from becoming confused. For example, all hotels should be able to process "towel requests" with the same data items and business procedures.

However, each hotel is free to decide how to deliver the towel to the customer, how to respond, and how to reveal the hotel's atmosphere and personality.

"Standardization should be used to organize data and stabilize internal processes, not to limit the customer experience."

Therefore, internal data and technology must be standardized, but the experience delivered to customers must be kept as flexible as possible. This is because the hotel's unique atmosphere and service method are the factors that allow AI to determine the hotel as a differentiated place.

"The things that make each hotel unique and provide a differentiated experience are what makes AI recommend that hotel more."

Standardization is not a tool to eliminate a hotel's individuality, but rather can be a foundation for more stable implementation of individuality.

12. Personal information, security, and speed of AI adoption

Jason presents privacy, security, transparency, and fairness as risk factors to watch out for when introducing AI. Joe says he tends to look at technology with a relatively liberal attitude.

"I'm pretty liberal about these things."

Rather than overly suppressing new technologies, he prefers an approach of trying them out and solving problems as they arise. However, it is recognized that within the organization, the IT department and other responsible persons must control him and establish appropriate governance and control mechanisms.

Even if AI is introduced quickly, all tasks should not be entrusted indefinitely. In hotels, the following controls are necessary, especially since customer information, payment information, reservation information, and personal information are handled.

• Setting the scope of data that AI can access
• Protection of personal information and payment information
• Approval and recording of automated tasks
• Review procedures when errors or hallucinations occur
• Distinguish between tasks that require employees to check the results and tasks that can be fully automated
• Ensuring transparency of answers provided by AI to customers

Joe wants to leverage the full potential of AI, but he also recognizes that he needs to be kept in check by his organization's IT leaders.

"The IT director needs to hold me accountable and make sure I don't cross the line."

13. What are the criteria for successful AI introduction?

Lastly, Jason asks what indicators will be available to judge the success of AI introduction three years from now. Joe says the pace of change in AI is so rapid that it's difficult to come up with one specific number as a benchmark for success.

Over the past year and a half, it has taken the hotel industry a considerable amount of time to get employees used to reviewing emails with ChatGPT once before sending them. However, around the time people adapted to simple searches and writing assistance, AI quickly developed to the level of directly manipulating computers, writing code, and performing tasks.

"The moment we feel we have caught up with AI, another change has already occurred."

Therefore, there is no need for hotels to feel the burden of predicting every technological change or being at the forefront of AI adoption. Joe is honest about how he doesn't always keep up with the pace of technological change.

What's important, rather than the AI ​​technology itself, is understanding the downstream effects that AI will have on the hotel industry - how it will change discoverability and demand for hotels. Even if you don't become an AI expert, you should be prepared to know that AI will change the way customers and hotels connect.

The conclusion Joe returns to again and again is a very practical one.

"Once you have a differentiated product and a great customer experience, the rest will more or less follow."

He says hoteliers don't need to fully understand all of AI's capabilities. Instead, you must provide good service, create great products, and provide an experience that sets you apart from competing hotels.

"You just need to focus on great service, amazing products, and differentiated experiences."

Conclusion: AI raises the floor of operations, and great hoteliers raise the ceiling

The core of this conversation is that AI does not eliminate the essence of hotels, but rather reveals the essence of hotels more clearly. AI reduces the list of hotels, recommends only a few that fit the customer's criteria, and evaluates hotels based on actual experience and reputation.

In the future, it will not be enough for hotels to simply appear at the top of search results. You must provide an experience that will keep customers coming back, recommending it to others, and leaving positive reviews on external media and platforms. Operational efficiency and automation reduce costs and increase service consistency, but the ultimate competitive advantage still lies in the unique experiences a hotel creates.

"Technology can raise the bottom line of operations, but great hoteliers will continue to push the top line."