Is travel booking ready to fully embrace artificial intelligence?
By Sam Hilgendorf, CIO for Fox World Travel
In the 1980s, I vividly remember McGruff the crime dog telling me how dangerous it was to get in a car with a stranger. Then, in the early 2000s, everyone knew that you should never meet a stranger you spoke to on the internet in person. But, in 2008, it was suddenly okay to find a stranger on the internet and now pay to get in your car?
What was previously unknown has become such a common activity that it has spawned its own viral meme. Of course, I’m talking about Uber. Founded during the Great Recession of 2008, Uber sparked a cross-societal shift with its industry-defining ground transportation solution. What would previously have been greeted with cynicism is now a globally recognized brand.
Uber wasn’t the only company to create behavior change in times of crisis. During the Great Depression, Disney brought Snow White and the Seven Dwarfs at theaters. Conventional wisdom at the time questioned the idea of adults sitting in front of a 90-minute cartoon that included singing dwarves. And yet today Pixar’s animated films are among the highest grossing films at the box office every year.
As we have seen time and time again, adversity inspires innovation, and innovation can change the norms of behavior.
So how do these examples apply to the travel booking experience?
Conventional wisdom has been that travel planning, whether for leisure or business, is still too personal and confusing for artificial intelligence (AI) to replace the travel shopping experience. While the stereotypical road warrior may know all the flight times and hotels within a five mile radius of his destination, the average traveler starts from scratch every time. Searching for the right price, at the right time, with the right services is often tedious and time consuming.
There are a lot of questions and challenges in trying to implement AI for travel. If it takes me so long to find exactly the right options at the right price, how could a bot do better? Would AI really understand the interdependencies between airline, hotel, and ground transportation? What if you included all of my personal preferences? An AI engine can’t guarantee that I have a great travel experience, can it?
These are all questions that have been asked of me over the past year as AI has started to advance in other industries. The unspoken rule of travel booking is simple: “To ensure a good experience, either myself or someone I trust has to plan the details. “
Well, chaos has a way of just rewriting the rules.
These are the darkest times the travel industry has ever seen. There continues to be tremendous pressure across the entire travel supply chain to reduce operating costs and labor as demand remains extremely low. Even with the positive progress on vaccines and stimulus measures, this pressure to control costs will not abate for the foreseeable future. There is an industry-wide need to provide an easier trip planning experience that is not labor intensive.
The answers can be found in natural language processing (NLP). NLP is a subfield of artificial intelligence focused on the processing and analysis of spoken and written languages. An NLP application could provide the mechanism for planning and booking a complete travel experience. Only with exceptions and errors sent to an experienced travel agent for analysis and discussion.
NLP technology is not new. Grammar checkers, autocorrect engines, and spam filters using NLP have been mainstays for years. Voice assistants, such as Alexa and Siri, are also examples of NLP coupled with voice recognition technology. Like Uber and Disney, each of these solutions changed our standards of behavior, many of which are now taken for granted. NLP provides both strategic direction and a means to permanently change the booking experience of travelers.
In 2020, important steps were taken in NLP. Microsoft, Google, and OpenAI all made significant strides in the last year in understanding and interpreting conversational language. Natural language processing has recently shown its ability to correctly complete sentences, summarize documents, and provide straightforward answers to complex questions. This fall, OpenAI’s GPT-3 language generation platform wrote a 500 word editorial to convince us all that the robots really come in peace.
Here are some areas where NLP could dramatically change the travel planning experience.
Travel agent assistance for complex itineraries
For more difficult trip planning, NLP could help experienced travel agents, reducing the time, errors and costs associated with planning.
Take, for example, a European business trip to multiple countries that includes air, train, ground and accommodation, each with negotiated corporate rates. NLP could help the agent by listening in to the conversation and recognizing key search parameters, such as destination, date, and preferences. The agent would receive relevant answers and search options, with any corporate travel policy applied. This cuts down on time and eliminates all data entry errors. It also allows the agent to focus their energy on making recommendations and recognizing risks or better alternatives within a route.
Conversational chatbots for travel planning
Conversational chatbots have been used across many customer service industries, but they’re not ubiquitous on travel planning websites.
There are already great travel research and review websites, such as TripAdvisor and Yelp, to determine what experiences are available in a specific city. But they don’t include a conversational aspect. With the prediction modeling capabilities of NLP, the conversational chatbot could inspire a unique experience during the trip planning process.
For this example, let’s take a leisure traveler who is thinking about a Caribbean vacation. In the initial dialogue, an NLP-enabled chatbot might come up with the typical questions a travel agent would ask about that part of the world, such as interests in activities like scuba diving or snorkeling, culinary interests or cultural attractions. By asking just one or two probing questions, the chatbot could inspire new ideas or new interests.
Let’s say I’m interested in snorkeling, but not scuba diving. The chatbot could ask a simple follow-up question: “Why not SCUBA?” My answer may be that I have never been trained. With that entry, the chatbot could then suggest that I easily get certified on the same vacation with only a few hours of training on my first two days, including my first dive.
Take this single example and multiply it now for each choice you make when planning a vacation. While many decisions can lead to the desire initially expressed, the ability to spark reflection on new and unexpected experiences can make vacation planning more inspiring.
A voice assistant for booking the entire route
Voice assistants have become a mainstream household appliance, but there has been limited adoption of voice assistants for purchasing travel. Today, very few applications allow you to book trips. And they’re limited to a single plane ticket or a single hotel reservation. Buying a complete itinerary combining air, hotel and ground transportation always presents complexities in terms of logistics and dependencies. As NLP better understands the requirements and dependencies of travelers, a solution to book a complete trip with your voice assistant will become available.
Taking a voice assistant and combining them with conversational chatbot capabilities creates a whole new travel planning experience. Suddenly I could speak through my desires and interests, being pushed by new ideas and opportunities that I previously had to find on my own over the course of my research time. If my itinerary becomes too complex, then the voice assistant can seamlessly transfer from me to a knowledgeable travel agent for further assistance. These capabilities could both reduce time and effort and lead to a better travel experience.
These potential innovations could spark Disney-like cynicism during the Depression or if Uber had been suggested in the 1980s. However, the devastating impact the pandemic has created on travel has rewritten all the rules. We need to be creative in leveraging AI and NLP to book trips. It’s time to challenge conventional wisdom.
20 years from now, if internet memes still exist, we’ll see some fun stuff about how travel “used to”.
Sam Hilgendorf, CIO of Fox World Travel
Sam Hilgendorf is the Chief Information Officer for Fox World Travel. As a global leader in travel management, Fox has over 60 years of success serving business and leisure travelers. Sam is responsible for the vision and execution of Fox’s technology portfolio, which includes IT, business intelligence, product development and technology professional services. Sam’s 20-year history with leading IT consulting organizations brings a unique perspective to travel management services. Sam is particularly inspired by the challenges that arise from the fragmented nature of travel. Under Sam’s leadership, Fox implemented a new customer contact center and enterprise data management platform specifically focused on consolidating information and presenting travelers with a simpler travel experience.