“The tourism industry of the future will not consist of moving people from point A to point B, unveiling new destinations or organizing trips, instead focusing on a 360-degree perspective around the traveler, which implies creating special, unique and memorable experiences for him.”
Defining the Future of Travel through Intelligence,
discussion paper by Amadeus Travel Intelligence
Do you remember your last trip? Did you collect your tickets directly at the counter? I doubt it. In our accelerated day to day, finding time to go to an agency, station or desk for tickets is a luxury that few can afford. Also, why bother if you can get them with just a couple of clicks through your laptop or smartphone?
Effectively, travel sales through the digital channel have grown exponentially in recent years, with a total of 564.87 billion dollars in 2016. And the figure is expected to reach 817.54 billion dollars in 2020. This explosive growth is being driven by recent technological advances, including Data Science.
If you are looking for new ideas on how to make good use of the huge data your business activity yields, here are 12 scenarios of Machine Learning applications and data analysis in the travel industry.
- Personalization of the UX (user experience) in airlines
According to the McKinsey 2016 report, travel companies (and airlines, in particular) if they use Data Science well, have a probability of multiplying their total customers by twenty-six, their customer retention rate by six and their profitability by nineteen.
Going back to the airlines, although in most of the cases where Data Science is applied, it is in the field of contingency management and predictions of delay, some are geared towards UX personalization. United Airlines, for example, had a “collection and analysis” approach to its data, but has redirected this concept and since 2014, it is applying “collect, detect and act” methods already in its landing pages.
The company tracks customer behavior, collecting data with more than 150 variables, including individual data (for example, previous purchases and search destinations) and general historical data. This huge amount of data is used to perform a detailed segmentation of the customer to adapt the UX in real time, according to the category to which a user belongs.
According to the assigned segment, the airline’s website performs calculations in 200 milliseconds to adjust the designs on the screen, the text and other elements of the website to increase the likelihood of conversion. This dynamic personalization has already increased complementary revenues by 15% in recent years, according to United.
- Recommendation Platforms
Possibly, it is the most common application case of Data Science. Currently, 99% of successful products have a recommendation solution incorporated. Like suggestions for personalized content on Netflix or the “Featured Recommendations” box on Amazon, online travel suppliers often offer personalized suggestions, based on your recent searches and bookings.
For example, when searching on Expedia for flights to London, you will be offered several accommodation options for your trip. Similarly, Booking.com offers alternative destinations that may interest you in your next getaway, also based on previous searches.
These are just two of the ways to use data-based recommendation engines in the travel industry. Following this pattern, online travel agencies can offer car rental offers, alternative travel dates (such as Fareboom.com travel agency) or routes, new travel destinations according to user preferences or even some recommended local attractions.
With a sufficient minimum of data on typical searches or preferred offers, a powerful recommendation algorithm can be built, which learns even more from a user’s personal browsing behavior to offer more personalized and more valuable suggestions.
For example, 8 out of 10 families may book a trip to Disneyland in July. This offer is of little interest to a business traveler traveling alone in January, but if you show that user a day trip to Las Vegas, the chances of success are quite high.
Consequently, investing in Data Science consulting could not only have a positive impact on your income through additional sales, but could also improve user participation by contributing with a personalized and efficient UX (user experience).
- Estimates of flight and hotel rates
Flight rates and hotel prices vary constantly according to the supplier and the anticipation of the purchase. Almost no one has time to track all these changes manually, so today the smart tools that monitor and send timely alerts with interesting offers are in great demand in the travel industry.
AltexSoft Data Science team has created an innovative rate prediction tool for a global online travel agency. By working on their main product, a digital travel booking website, they were able to access and collect historical data on millions of rate searches for several years. Thanks to such information, they created a self-learning algorithm, able to predict the future movements of prices based on a number of factors, such as seasonal trends, demand growth, airlines special offers and promotions.
With an average confidence rate of 75%, the tool can make forecasts in the short term (several days) and in the long term (a couple of months).
For example, if during the last holiday seasons all the prices of flights from San Francisco to Dallas grew significantly a couple of weeks before Christmas, the same trend could occur this year as well. In this case, the algorithm will say that waiting can be too risky and will ask you to book your flight now.
Similarly, if the price of flights to Las Vegas generally falls below the average one week before Christmas, it will be offered to wait and book flights at a closer date.
Another prominent startup that uses Data Science to help people book the cheapest flights using applied predictive analytics has created an intuitive mobile app for predicting air fares.
The Hopper tool analyzes billions of data sets daily to provide accurate predictions of price oscillation. There is a detailed description of how the system works on the company’s website.
Therefore, forecasting tools based on analysis, such as those mentioned above, have proven to be a valuable addition to a travel booking portal. Although they mainly apply to flight bookings, such tools could be used in other specific areas of the travel and accommodation industry.
They can be used to predict changes in hotel prices, inform when all rooms in a certain hotel are predicted to be full or even suggest alternative itineraries for trips according to weather forecasts or the projected airport load in a determined day, to help users not only save money, but to improve their customer experience.
- Smart travel assistants
People seek above all comfort in their day to day. Thus, intelligent concierge services, driven by artificial intelligence (AI) are gaining momentum in several industries. The booking of trips is just one of the highly automated areas using algorithms.
Smart programs, trained to perform a specific task at the user’s request, are often referred to as “bots” or “chatbots”. Some large companies have adopted instant messaging platforms as an excellent way to reach customers and build better relationships with them.
For example, Hyatt, a world leader in hospitality, has been using social platforms to connect with its customers since 2009. Recently, the company expanded its set of customer support tools with Facebook Messenger. However, 24/7 mobile support requires considerable resources, both human and financial. Hence the usefulness of virtual assistants with artificial intelligence.
In most cases, virtual travel assistants are integrated with popular instant messaging applications, such as Facebook Messenger, Slack, Telegram or Skype, and are trained to perform various tasks: from searching for the cheapest offers, booking flights and making hotel bookings, to planning complete trips and improving your general customer experience through useful information, tips and suggestions on popular tourist destinations, places to eat or local attractions: these are the most popular ways to use AI Bots.
Other popular travel booking services allow you to plan trips directly from the Facebook Messenger application through more “humanized” chatbots. They do not need special commands, they understand simple questions and respond in an informal and colloquial style.
The list of travel chatbots does not stop growing, just like the cases of possible use of this technology. The application of these AI travel assistants does not end with the search and booking. They can take on the role of mobile travel companions / advisers, solving problems on the fly such as:
What is the baggage policy of my flight?
Where is the nearest waiting room?
What is my boarding gate number?
How long will it take me to get to the airport?
Whereas chatbots have not yet reached massive popularity among travelers, smart assistants are becoming increasingly popular when combined with voice-activated devices. The estimated number of Amazon Echo owners was approximately 8.2 million people in July 2017. Powered by Amazon Alexa, a smart assistant, the Echo device provides a fully voice-activated interface. Similar products are available on Google, with Google Assistant technology, and Apple, with Siri technology.
Amazon and Apple already offer open development kits to integrate voice assistants with third-party services. For example, Alexa from Amazon has around 15,000 integrated service skills. When users enable search engines or aggregators in Alexa, they can track flights in real time, explore travel options and book hotel rooms.
The best-known smart assistants already have enough public and integration with them is the shortest way to reach travelers who are used to voice-activated interfaces.
- Tailored offers for MVC (most valuable customers)
The importance of loyalty programs for the travel and accommodation industry continues to grow. In 2016, the number of members of the loyalty program of the main hotel chains increased by 13.1% (344 million members). As this figure will not stop growing, travel agencies and hotels will have enough stored data to effectively apply artificial intelligence-based personalization.
The members coming from these loyalty programs, which means, the most valuable customers, are the users on which the industry should focus first to avoid abandonment. And this is the application that machine learning has at hand.
Providing personalized offers for new or unregistered users using behavior tracking techniques, metadata and purchasing history will not always work well. Some predictions can be made based on the general data history, but they will not be as accurate for newcomers. On the other hand, combining historical data and real-time data from frequent buyers is much easier. OTAs and hotels can make personalized suggestions for MVCs, reducing the abandonment rate and strengthening ties between brands and their loyal customers.
Tailored suggestions for MVCs also include complementary sales opportunities. For example, OTAs can use their data to suggest certain hotel rooms or vehicle rental options, in addition to airline ticket purchases, since they use supplier booking APIs. This action can be done via mailing or directly from a website.
- Optimization of contingency management
Whereas the previous case focuses mainly on planning trips and helping users in the most common situations during their trips, automated crisis management is somewhat different. Its objective is to solve the real problems that a traveler could face on his way to a destination.
Applied mainly to business and corporate trips, the management of contingencies is always a time-sensitive task, since it requires an instantaneous response. While the chances of being affected by a storm or a volcanic eruption are very small, the risk of interrupting a trip is still quite high: there are thousands of delays and several hundred flights canceled a day.
Regardless of the reasons, being stranded somewhere in Europe late at night when you must be in Tokyo before noon tomorrow can cause huge inconvenience. In addition, in business trips, this could cause significant losses and serious implications for your company.
With recent advances in technology, it has been possible to predict such interruptions and efficiently mitigate the loss for both the traveler and the operator. The 4site tool, built by Cornerstone Information Systems, aims to improve the efficiency of business travel. The product caters to travelers, travel management companies and corporate clients, providing a unique set of features for the management of travel interruptions in real time.
The advantage that the Data Science brings here lies in predicting travel interruptions based on the information available about the weather, current delays and other data from airport services. A trained algorithm for monitoring this data can send timely notifications, alert users and automatically implement a contingency plan.
For example, if there is a heavy snowfall at your destination and all flights are redirected to another airport, a smart assistant can check if there are available hotels there or book a transfer from your actual arrival location to your initial destination.
Not only passengers are affected by unexpected events; Airlines also suffer significant losses each time a flight is canceled or delayed. Therefore, Amadeus, one of the main global distribution systems (GDS), has introduced the Schedule Recovery system, with the aim of helping airlines to mitigate the risks of cancellation of the trip. The tool, a recommendation engine powered by Data Science, helps airlines efficiently address and manage any threat or disruption to their operations.
It also guides operators to make well-founded decisions aimed at optimizing operations for better efficiency. Qantas, the largest airline in Australia, was the first to apply the system to improve its operations. As a result, the company stated that “The Amadeus solution helps reduce the number and duration of delays, whether due to excessive traffic, operational delays or weather conditions, leading to a better customer experience for travelers.”
The system was subjected to a real hard test in 2016. During heavy storms that caused delays along the entire east coast of Australia, only 15 of the 436 Qantas flights (approximately 3.4%) were canceled, compared to 70 of 320 (22%) operated by Virgin Australia, which uses an outdated manual system to manage incidents. The company’s punctual performance also remained high: 86% on Saturday and 62% on Sunday, while Virgin’s performance was 74% and 48% respectively.
- Customer Service
Similar to personal travel assistants and intelligent contingency management, airlines and other travel companies can use the power of artificial intelligence to streamline the customer service process. Especially now, when 50% of consumers say that the speed of response to a query is the most valued in good customer service.
Based on the experiment conducted by Qantas to evaluate the effectiveness of its crisis management system, it takes between 15 and 20 minutes to an experienced professional to solve a situation whereas it can be done in less than a minute using an algorithm.
That said, Gartner predicts that 25% of customer service and support operations will depend on the virtual assistant technology in 2019.
Combining virtual assistants with humans cannot only help you increase your brand loyalty, but also optimize your business performance. For example, if a passenger’s luggage is lost, reporting the loss or even performing an automatic search through a virtual assistant could significantly speed up the search process. This approach eliminates bureaucracy and paperwork, which is an excellent way to rehabilitate yourself in terms of customer experience. Also, offering a free bonus for any inconvenience caused is an even better way to retain your customers.
- Sentiment analysis in social networks
90% of American travelers with smartphones share their experience, photos and opinions about services on social networks, according to Amadeus. TripAdvisor has 390 million unique visitors and 435 million reviews. Every minute, approximately 280 traveler reviews are sent to the site.
This large pool of valuable data can be analyzed by the industry to improve its services. In addition to the conventional statistical analysis of subsets of opinions, the calculation power and the underlying techniques of machine learning allow analyzing all the reviews related to a brand.
The analysis of emotions aims to explore the text data to define and classify their emotional and real qualities. For example, the Google Cloud Natural Language API is a standard application programming interface that can be adjusted and integrated with analytical tools to provide real-time analysis of all brand-related revisions.
The maturity of natural language processing and sentiment analysis allows the adoption of precise analytical tools avoiding the time consuming data collection. The range of applications is vast. It could be the dynamic monitoring of the brand image in general, or the ad hoc analysis of the reaction of social networks after an update or novelty in the product or service.
- Dynamic pricing in the tourism industry
The dynamic pricing consists in changing the prices of rooms or tourism products depending on the various circumstances of the market. Said price optimization is nothing new; Hilton and Marriott have been modifying their room rates once or twice a day since 2004. As machine learning applications become popular, dynamic prices are based on predictive analytics of the best price that includes more variables.
In 2015, Starwood Hotels began to develop a predictive analytics tool that took into account hundreds of factors to show the most efficient price at all times. These include data on competitive prices, weather, user’s booking pattern, occupation data, room types, daily rates and other variables. While the system can operate fully automatically, it also allows human operators to view the dashboard and adjust rates manually if necessary.
Dynamic pricing fixes the establishment of efficient yield management based on the understanding of relevant data from customers and markets. While online travel agencies (OTAs) dominate the field of room booking, the dynamic pricing based on data allows hotels to manage room reservations directly avoiding intermediaries.
- Specific offers to deal with travel contingencies
Trip interruptions, such as flight cancellations, may offer unexpected opportunities for hotel owners. In the terrible winter of 2013, 500 daily flights were canceled in the USA. The following year, the Red Roof Inn hotel chain launched an event-based campaign.
The hotel used public data sets on cross-flight cancellations with weather information. The data was collected through an API and processed through a conditional algorithm. With the result, the hotel chain predicted the cancellation of flights and directed its advertising to travelers from areas that would likely be affected. The campaign offered information about their available rooms and the distances from the airports to potential guests looking to make last-minute reservations.
The results of the campaign were impressive: the number of Red Roof Inn reservations increased by 266% the reservations via mobile devices.
- Fraud detection
According to several reports, airlines and the travel industry suffer a high level of e-commerce fraud. They lose billions of dollars each year by returning stolen money to customers. Payment fraud is one of the most popular types of scams in this industry. It involves the use of a stolen credit card to book flights or accommodations. Another popular type of fraud is a “friendly” fraud, when a customer pays for a purchase, and then claims that the card was stolen, demanding a refund of the charge.
Analyzing customer behavior through the use of machine learning profiles and technologies can help prevent and detect illegal transactions. Applying an Artificial Intelligence technology to detect fraud improves the average final cost per transaction.
- The experience during the stay
AI solutions can help travelers not only on their way to a destination, but also during their stay at the hotel. With virtual voice assistants in the rooms, guests can feel more comfortable. For example, they can set a temperature in a room, adjust the light or turn the TV on and off. With facial recognition, hotels can speed up registrations and make hosting more secure.
Facial recognition and voice services are already working in many hotels around the world. More and more establishments use chatbots, or even have janitor-robots. Voice virtual assistants are increasingly frequent. According to the Oracle Hotel 2025 report, 78% of the hotels will improve the rooms incorporating voice-controlled devices, and 68% will use robots for check-in and check-out in 2025.
Wynn Las Vegas has equipped all its rooms with Amazon Echo, Safeco Field Suits is not only used in the rooms, but also suggests the guests what to do in the city during their stay. Radisson Blu Edwardian Hotel, in London, uses a chatbot named Edward, and Las Vegas Hotel Cosmopolitan has Rose, another virtual concierge who answers any questions and helps 24 hours a day, 7 days a week. Clarion Hotel Amaranten, in Stockholm, also uses a chatbot based on Alexa.
With facial recognition technology, the experience at the hotel becomes much safer. For example, Lemon Tree Hotel, in Deli, installed a facial recognition system to advance security. This system captures facial images from the CCTV camera and compares them with existing images in the database. And a Japanese hotel, Henn na, is completely managed by robots. In Henn na Hotel the receptionists are robots, the concierge is also a robot and, instead of issuing the electronic keys, the guests must register their facial images during the registration.
Data Science is showing the way of how we will travel in the future. The fields listed above for your application are just the tip of the iceberg.
Find the reference information in Altexsoft