Because travel is an inherently human-centric experience, tourism and hospitality companies have been somewhat skeptical of AI in travel impact. But sentiment is shifting.
Europe is and will be a fantastic place to invent new AI applications, especially for tourism. We are the first tourism destination of the world, so we are the one generating the largest, biggest amount of data in tourism on this planet.
As domestic and international travel rebounds, companies are looking for new ways to capitalize on the growth momentum, and AI appears to be one of the vehicles for getting faster traction.
In this post, we analyze the scale of AI in travel, zooming in on benefits, commercial opportunities, and feasible use cases.
Key benefits of AI in travel
AI models are much better than humans at analyzing data — and the travel industry has deep data troves. By using algorithms for advanced data analytics, industry players can reach more customers, elevate service levels, tap into new revenue channels, and increase operating efficiencies.
Deeper customer insights
Machine learning (ML) and deep learning (DL) algorithms can trawl millions of data points in provided datasets to uncover new correlations, trends, and similarities. In short, AI and ML both enable advanced customer segmentation, sentiment analysis, and behavior forecasting.
Hostelworld, for example, successfully uses machine learning for sentiment analysis and marketing campaign optimization. By combining the analytical and predictive powers of ML, Hostelworld managed to increase its click-through rate (CTR) for email campaigns by 86% and its email open rate by 12%.
Better customer service
Thanks to natural language processing (NLP), algorithms can easily understand text-based commands and different contextual clues to better deal with incoming customer requests. At the most basic level, AI can help classify and prioritize customer support cases or look up relevant information for agents. More advanced AI use cases include end-to-end customer issue resolution and voice-based customer support.
Artificial intelligence in tourism can increase support staff productivity by 20% to 50% or more. Airlines like Cathay Pacific already handle 50% of their customer care chats with AI assistants, allowing human agents to focus on more complex tasks.
New revenue channels
AI models are good at needle-in-the-hystack types of problems. By applying classification, regression, or interference, algorithms can locate new revenue-generating and/or cost-saving opportunities within a presented dataset. An average AI model can make over 100 million sales-related decisions each day.
Model outputs can vary from hyper-personalized cross-sells or upsells to dynamic price optimization. Finnair, for example, increased revenue by 3% by optimizing prices across 70 origin and destination (O&D) segments with AI.
Streamlined operations
Apart from supplying teams with business intelligence, algorithms can also handle low-value menial work, ranging from data entry and reconciliation to data modeling and reporting. When integrated with other travel technology — booking engines, property management systems, revenue management software — AI algorithms can also complete more complex workflows: automatically check in guests, design better route schedules, or optimize staffing levels based on demand trends.
citizenM, for example, uses Mist AI by Juniper — an intelligent IT operations and support platform — to streamline the deployment and provisioning of IT services across its portfolio of properties. Thanks to AI, the hotel’s team can create automated workflows operated via the cloud to support exceptional guest experiences, ranging from guest self-checkout to in-room technology. With AI-driven IT infrastructure and network monitoring, citizenM can also have fewer technical staff on-site for troubleshooting.
7 real-world use cases of AI in tourism
AI models may be great with analysis, but what can they know about real-world adventures? A lot, actually.
ChatGPT has turned out to be a capable travel agent. Whether you’re looking for “things to do on a budget in Rome” or a “10-day itinerary for a backpacking trip in Peru,” ChatGPT has an unlimited roster of options. Thanks to plugins, travelers can also compare flights, research car rentals, and handle hotel bookings straight from the GPT app.
But conversational AI in travel is just one use case. Almost every major player in the tourism and travel industry is sizing up the technology’s cross-functional potential and looking into building foundational AI models in-house.
Among travel executives surveyed by Euromonitor, 97.8% agree that AI will have a major impact on the industry over the next five years. The travel booking industry is expected to be among the first to be disrupted by AI, though other players can also see substantial dividends from AI investments.
If you want to be among the next generation of leaders, we recommend looking at the use cases of AI for travel.
Digital concierge services
An outstanding guest experience is a major revenue factor for the hospitality industry. So is impeccable customer service. But with ongoing staff shortages, both are hard to deliver.
Eighty-two percent of US hotel managers are experiencing a staffing shortage, 26% severely so – meaning the shortage is impacting the hotel’s ability to operate.
Unlike the early generation of chatbots, which were mostly driven by pre-programmed rules, an AI travel agency or AI travel assistants have more wits and can perform a greater repertoire of tasks. Thanks to natural language processing (NLP) and large language models (LLMs), chatbots can analyze and summarize content from a wide variety of sources to reply to different user queries.
On the back end, conversational systems can also interact with other tech systems: exchange data, look up information, update records, etc. Thanks to such integrations, an AI concierge can automatically handle a wide range of tasks, from guest self check-in to ordering late-night munchies and upselling some neat services in between.
At Virgin Hotels, guests are greeted by Lucy — an in-app virtual assistant. Lucy functions as a contactless mobile key to access the room and can automatically adjust the lights, thermostat, and TV. She’s also the one to ring up for room service or ask about any details regarding the stay.
Lucy is a capable concierge because it integrates directly with Virgin’s property management system (PMS), which contains data about guest bookings; a point of sale (POS) system used for managing food and beverage operations; a smart system for controlling every appliance in the room; and guest management software, which automatically generates checklists for staff based on guest requests. Thanks to such deep integrations, Lucy can perform a wide range of tasks across all Virgin properties and retain guest preferences for better experience personalization.
AI-driven revenue management
Profit margins in the travel industry went from 21.93% in Q3 2021 to 14.22% in Q2 2022. Energy price hikes, rising fuel costs, ongoing supply chain kinks, and accelerating inflation are behind that squeeze. Moreover, many travel companies are still rebuilding cash reserves depleted during the pandemic.
In the hotel industry, room rates have historically been the main source of revenue. But as external costs continue to rise and guest expectations evolve, hoteliers need to focus more on total profitability.
The problem? Most revenue management software typically produces spreadsheet-based reports, which managers then need to interpret and transform into better strategies. AI solutions can do more advanced number crunching when it comes to analyzing data, predicting trends, and prescribing strategic actions.
By analyzing historical data and operating trends, AI-powered revenue management solutions can suggest better:
- Channel strategies to maximize revenue per available room (RevPAR) and average occupancy rates
- Upsell and cross-sell opportunities to increase total revenue per available room (TrevPAR) and guest satisfaction scores
- Reward and loyalty schemes to increase customer lifetime value and strengthen retention mechanisms
Park Royal Hotels & Resorts used Duetto’s open pricing algorithms to determine new customer segments and adjust their rates according to dynamic demand trends. As a result, the chain saw an 8% increase in the average daily rate (ADR) and a 2% increase in occupancy rates within just six months of adoption.
Dynamic offer optimization
Profitability is a tough act in the airline industry, especially as airlines aim to recover from the record $183.3 billion of net losses for the 2020–2022 period. With passenger numbers going up, the airline industry is expected to post a net profit of $9.8 billion (a 1.2% net profit margin) by the end of 2023.
To secure higher profit margins, airlines have traditionally relied on two strategies:
- Dynamic pricing, which has been around since the 1980s
- Ancillary product sales (baggage, advance seat reservations, etc.)
Both revenue management and merchandising of ancillary products can be improved with AI. Using dynamic, personalized offer creation instead of traditional fares and ancillary filings, airlines can determine the optimal price for each seat and route based on customers’ willingness to pay while increasing sales conversions.
Source: International Air Transport Association (IATA)
The US carrier JetBlue, for example, uses a deep learning-based revenue optimization system from FLYR to get real-time intelligence on demand trends, leg-level opportunity costs, and itinerary-level traveler willingness to pay, among other factors. The system automates most pricing decisions with high precision both for seats and ancillary services.
Air Azul, in turn, uses Sabre Air Price IQ to dynamically optimize airfare and generate more relevant offers in real time based on available shopping and revenue management data. On average, airlines using Sabre generate up to 3% more revenue.
Route optimization
Transportation is a key sector in the tourism and travel industry and also the one prone to the most disruption. In the first half of 2023, almost 80% of American travelers experienced at least one travel-related problem.
For operators, disruptions mean an influx of annoyed customers (and often reputational damage), plus higher operating costs (and thus lower profitability). One minute of flight delay can cost the European economy up to €100.
Beyond that, poor operational practices also result in higher fuel burn rates, which are bad both for the environment and profitability. Forced cancellations and low flight occupancies also make operations more expensive.
The algorithms of AI in travel industry have already proven capable of predicting flight delays, modeling service costs for new destinations, and optimizing on-the-ground operations. Swiss’s Operations Control in Zurich recently adopted an AI platform for optimizing operations. The algorithm analyzes passenger itineraries, aircraft assignments, crew rostering, and aircraft maintenance dates and proposes optimal operating scenarios. The platform has successfully optimized over half of the flights in Swiss’s network, saving the carrier over $5.4 million in operating costs.
Lufthansa, in turn, uses AI to forecast the wind direction at Zurich Airport, where it frequently results in flight delays and cancellations. Thanks to AI, Lufthansa can predict wind direction with 40% accuracy, resulting in more reliable operations.
Ground transportation companies can also benefit from AI a lot when it comes to:
Intellias, for example, developed a cloud transport management platform for one of our clients that helps optimize vehicle routing in cities to ensure high operational efficiency.
Process automation
Many travel companies are short-staffed. With the rising cost of living and inflation, hiring staff has become more expensive. According to Skift, US accommodation providers now employ 12% fewer workers compared to before the pandemic, while hiring new staff now takes five to six weeks.
Intelligent process automation (IPA) can automate manual and menial business processes, ranging from payment reconciliation to shift planning and housekeeping management.
For example, Choice Hotels used IPA to automate the loading of negotiated rates with corporate clients. The bot closely mimics the work human agents used to do, but at a higher speed and with fewer errors. The new system decreased process completion time by 85%, and the team is now looking into using IPA to streamline responses to RFPs from franchisees.
We’re combining IPA with AI for better pricing, and looking at using it in other areas of the company to shift away from manual processes to help improve efficiency by 10 to 20%.
Airports are also looking to exploit the predictive capabilities of AI for optimizing operations. For example, Assaia offers Predicted Off Block Time (POBT) technology to optimize on-time arrival gate availability and alert ground handlers in real time for faster turnaround. The model factors in the aircraft model, landing site, weather conditions, and passenger count to provide teams with situational data.
JFK Airport in New York City recently adopted this AI solution and reported some amazing results: a five-minute average reduction in ground delays and a 25% increase in the percentage of occasions when the gate is clear upon aircraft arrival. Both of these improvements translate to cost savings of $40 million per year.
Hyper-personalization
An AI travel agent may have never been on a real trip but still knows how to delight travelers in a new way. Thanks to robust classification capabilities, AI travel apps can identify new patterns in data representing customers’ purchase intentions, stay preferences, and general travel interests. Similar to Netflix, such algorithms can churn out hyper-personalized travel offers based on customers’ past trips, known room preferences, and data from third-party sources.
Accor, for example, uses an AI-powered customer-relationship management (CRM) suite with some 400 customer attributes to run deep segmentation and personalize customer communication. According to the company’s CDO Alix Boulnois, the solution generates a significant uplift in sales, with revenue from emails increasing threefold.
Navan, a corporate travel and expense management platform, released a conversational AI assistant app that creates personalized travel itineraries. Nicknamed Ava, the app provides recommendations based on a user’s loyalty program membership, preferred hotel type, favored amenities, room size, and distance from the company office. The app also cross-checks reservation requests against corporate policies.
Customer care
Travel planning has a lot of moving steps, from figuring out the itinerary to finding a room with the right amenities, booking group transportation, and choosing add-on activities. In other words, there’s a lot of back and forth between customers and hospitality service providers.
The best travel businesses take pride in processing all customer inquiries and service requests swiftly and with a personal touch. And today, this personal touch is often delivered by AI.
Juliet, an AI voice bot from WestJet, can resolve up to 74% of customer service requests without human intervention. When the airline saw a 45x spike in customer service calls during the pandemic, Juliet successfully assisted human agents. Since the chatbot’s launch, WestJet’s customer satisfaction (CSAT) rating has increased by 24%.
In a recent interview, Brian Chesky of Airbnb mentioned the company’s plans to build a fully AI-augmented customer service platform so agents don’t have to go through 72 user policies to get an answer to a customer’s query. His goal for the next 12 months is to make AI an invisible force behind the sentiment Oh my, their customer service is amazing!
AI in travel and hospitality: Emerging trends
Though adoption rates have been growing, AI in the travel industry is still at the nascent stage. Only a fraction of travel companies have reached AI maturity. Yet AI transformations will likely happen faster than digital transformations. In a five-year perspective, we expect the following AI trends to enter the mainstream.
Generative AI as the main interface for vacation planning
The first generation of online travel agencies (OTAs) digitized travel planning and disrupted travel marketing. But with travel apps, initial success depends on users choosing the right search filters and keywords.
Generative AI can take this by inspiring travelers with hyper-personalized choices based on their personal preferences and real-time social media trends.
Source: Skift
A new cohort of generative AI travel startups help customers plan every leg of their journey via text-based exchanges. Troupe launched a beta AI-powered mobile app for looking up suitable group accommodation from Vrbo and Expedia. Based on a text prompt such as “a cottage with 4 bedrooms in New England,” the app provides a list of options on the map and booking details from the associated websites. CharterGPT by Jet.AI, in turn, helps users find and book private charter flights based on input criteria — origin and destination airports, dates, number of passengers, etc.
Gen AI can help bring fundamentally human characteristics such as aesthetic judgment or creativity to new IT solutions for discovering, booking, and enjoying travel to new destinations.
Algorithms can help fight climate change
The travel sector is responsible for 8% of the world’s greenhouse gas emissions. Mass tourism has also proved damaging to local ecosystems, increasing pollution and environmental degradation worldwide.
At the same time, the consequences of climate change — heatwaves, droughts, heavy rainfall, forest fires — are already disrupting vacation plans.
More than half of travelers are now considering how climate change will impact the way they plan their holidays in 2024.
Algorithms can help mitigate the physical damage humans are doing to the environment.
Alberta Tourism Agency, together with the Alberta Machine Intelligence Institute, has been exploring how algorithms can help predict travel patterns, accommodation occupancy rates, and weather conditions. According to James Jackson, CEO of Tourism Jasper, the team is very close to “providing insight on everything from dispersion — as to how vehicles travel within the park — to human-wildlife conflict, to climate change and sustainability around glaciation.” By using machine learning to process all available data, the team can get near real-time insights on what’s happening in every destination and how it affects the climate.
American Airlines and Breakthrough Energy have partnered with the Google Research team on clearing up the skies — literally. Clouds created by contrails (the clouds of condensation that are sometimes formed behind airplanes) account for 35% of aviation’s global warming impact. By combining data around weather and flight paths with satellite images, the three companies train AI to create real-time contrail forecast maps, which pilots can then use to optimize flight routes and minimize pollution.
Big data, combined with AI algorithms, also has strong potential to decarbonize the ground and maritime transportation industries through optimized route building, reduced fuel consumption, and more sustainable transportation network planning.
Intellias, for example, recently helped build an automated CO2 Emission Classifier platform for trucks. The tool automatically collects data on vehicle class, engine class, axle, and chassis configuration, among other parameters, to dynamically estimate CO2 emissions in grams per ton-kilometer. It also matches emissions levels to the five EU-established fuel emission grades for trucks, making carbon reporting easy and effective.
Smart sensing technology will enable new use cases of artificial intelligence in travel industry
Internet of Things (IoT) devices are ubiquitous. Smart traffic lights govern the streets. Temperature sensors collect environmental population data outdoors and measure energy efficiency indoors.
A growing number of hotels are experimenting with integrating AI into their rooms to offer continuous, personalized guest experiences. ING recently presented an AI bedroom concept. Almost everything in the room is controlled by a privacy-friendly hotel voice bot — light systems, shades, in-room speakers, and a smart TV. The AI concierge can also respond to different customer requests and provide essential information about the property.
Park Hyatt New York launched five AI-powered wellness suites last year. Equipped with a king-size Restorative Bed by Bride, the suites lure travelers with the promise of the best night’s sleep. The Restorative Bed monitors heart rate, respiratory rate, and motion, adjusting the temperature and firmness for the sleeper’s comfort. Thanks to AI, the bed also learns from each experience and automatically creates a personal bed profile for each user.
IoT devices are also the key technology for creating digital twins — virtual, data-driven replicas of physical assets such as a production facility or a hotel property. Intellias has recently launched a digital twin solution designed for the travel and hospitality industry. With IntelliTwin, businesses can remotely control and optimize property performance based on recommendations from the ML system.
IntelliTwin for travel and hospitality
Specifically, InteliTwin can measure a building’s exact power and water consumption using data from utility companies. When combined with smart on-premises technology, InteliTwin can also suggest scenarios for optimizing resource use. By integrating InteliTwin with a property management system, users can forecast energy consumption under different occupancy scenarios: for example, pre-heating rooms based on expected occupancy, switching on appliances just before check-in, and dynamically regulating air conditioning across the property based on the weather and occupancy rates.
The use of AI sensors
On the streets, AI is also helping popular tourist destinations delight visitors with better experiences. The city of Florence, visited by over 20 million people per year, has launched a Feel Florence app. Using data from street-mounted sensors and CCTV cameras, the app guides travelers to less crowded streets and experiences. Users can check how crowded different sights are in real time and choose an itinerary through a quieter neighborhood.
A German tourism agency, in turn, is using Lidar sensors and analytics to reduce overcrowding at North Sea beaches during summer months. Weatherproof sensors perform anonymous visitor counts and visualize occupancy rates at tourist hotspots to inform other visitors.
By combining sensor technology with AI, travel companies can conceive and deliver innovative guest experiences, ranging from hyper-personalized stays to more sustainable, crowd-free experiences on the streets.
The potential of artificial intelligence in the travel industry is massive
In 2024 and beyond, success in the travel and hospitality industry will be largely shaped by a business’s ability to apply the right technologies to the right operating pressures, whether to improve workforce efficiency, lower carbon emissions, or slowly grow revenue per room.
AI offers travel players a unique opportunity to reinvent and improve almost every aspect of travel planning, from creating personalized itineraries to booking unique stays and receiving unparalleled assistance during the trip.
In fact, AI can become a new channel partner, promoting the creation of new industry partnerships among travel suppliers, distributors, large global brands, and emerging startups, offering new customer-facing products for smarter travel. By sourcing more data and embedding products from partners, travel companies can reach new customer demographics, grow their ancillary revenues, and continuously adapt to evolving customer behavior.
Intellias helps global companies stand at the forefront of the AI revolution. From intelligent chatbots to advanced NLP solutions, our engineers help bring the boldest tech ideas to life. Contact us to learn more about our services for the travel industry.
