Why is Data So Important For AI to Work?

There is no question that AI in general, and generative AI (GenAI) specifically, are having a moment. Ironically, while most of the current excitement is focused on generative AI, the overall field of AI has made significant inroads into the travel industry for quite some time. For example, Amadeus began implementing AI solutions in the enterprise in 2011, and Expedia processes over 600 billion AI predictions each year.  

AI adoption has reached a tipping point and is accelerating rapidly because it excels at managing uncertainty and data, two things the travel industry has in abundance. However, unlike traditional data and risk models, which rely on predefined rules and operate within strict parameters, AI adapts and refines its models without business rules or supervision after an initial training period. This adaptability enables AI to tackle complex problems in the travel industry that traditional models may struggle with, such as:

  • Predicting flight delays with dynamic factors like weather and mechanical issues.
  • Offering personalized travel recommendations that evolve based on user preferences and changing trends.
  • Managing customer service interactions with flexibility and nuance, understanding the intricacies of human language and sentiment.

The importance of data as the foundation and enabler for AI cannot be underestimated. Even the most sophisticated AI models may falter without accurate, comprehensive, and timely data, leading to inaccurate predictions or suboptimal recommendations. Data integrity is a technical necessity and a strategic imperative in leveraging AI for the travel industry.

The ability of AI models to continuously learn and adapt has led to significant improvements in their performance and business impact. AI is impacting all parts of the travel industry value chain, including booking and personalization, back office and reconciliation, in-trip traveler support, and post-trip with tasks like expense management integrations to ERP platforms. To better understand how AI currently impacts travel, let’s explore these areas in more detail.

Booking and Personalization

The travel booking process is where AI excels in personalizing and optimizing experiences. AI analyzes past user interactions and preferences through machine learning and reinforcement learning to tailor search results and recommendations. Reinforcement learning, a type of AI that learns optimal actions through trial and error to maximize a specific reward, continually improves the relevance of travel options presented to the user in real-time.

Data, again, is a crucial enabler for AI-driven personalization.  Diverse and extensive datasets on user behavior, supplier preferences, loyalty programs, booking patterns, and feedback are indispensable for effective personalization. These datasets enable AI to understand nuanced customer needs and adapt recommendations in real time, ensuring each interaction is as relevant and personalized as possible. The emphasis on data quality directly translates to the effectiveness of AI in delivering tailored travel experiences.

Navan’s AI personal assistant, Ava, is an example of this trend in personalization. It can offer individual travelers tailored recommendations for hotel and air bookings based on individual preferences and booking patterns, saving time and increasing efficiency for travelers. Ava can also analyze real-time spend data, provide insights into travel and expense questions, and proactively find opportunities for savings.

Back Office and Data Reconciliation

Once a booking is completed, machine learning aids in the back-office reconciliation process, efficiently handling data from various sources and normalizing it for storage in data lakes. This data becomes a valuable resource for reporting and analytics. In corporate travel, it allows businesses to monitor travel expenses and compliance as well as purchasing patterns to aid in contract negotiations, for example. Customer facing companies like Hilton aggregate and analyze guest booking data and use AI to:

  • Segment guests into different categories based on preferences, past stays, and spending habits.
  • Develop targeted promotions and offers for each segment, increasing the likelihood of conversion and guest satisfaction.
  • Predict guest behavior and anticipate their needs, creating a more personalized experience during their stay.
  • Optimize pricing strategies based on real-time demand and guest characteristics, maximizing revenue without sacrificing bookings.

In back-office and reconciliation processes, high data quality is paramount for trust and accuracy. Machine learning algorithms rely on accurate and normalized data to efficiently manage and reconcile vast amounts of booking information. High-quality data ensures these algorithms can accurately categorize, analyze, and store information, reducing errors and enhancing decision-making. This foundational data quality directly contributes to operational efficiency and strategic insights within the travel industry.

AI Throughout the Travel Experience

As travelers progress through their trip, they encounter AI throughout their journey, often unbeknownst to them. From flight and schedule optimization and customer service to fraud prevention by credit card companies, AI ensures a smooth and secure travel experience. Uber, for example, uses AI to predict demand for rides in specific areas and dynamically allocate drivers, to ensure cars are always a few minutes away. Google Translate is another example of AI enhancing the travel experience; it can translate text, images, and audio in real time.

Customer service is one of the most significant areas where AI impacts the travel industry, potentially increasing support staff productivity by 20% to 50%. Airlines like Cathay Pacific already handle 50% of their customer care chats with AI assistants, allowing human agents to focus on more complex tasks.

Expense Management

Upon trip completion, AI facilitates expense reporting by extracting and categorizing information from receipts and invoices. This automation streamlines expense management, reducing manual effort and increasing accuracy. The ExpenseIT feature within the Concur Mobile app is a good example; it automatically extracts information from uploaded receipts, such as amount, date, and expense type, using machine learning algorithms, significantly reducing the time and effort required for users to create an expense report, as it minimizes manual entry and improves accuracy.

Cost Saving and Analytics

AI’s ability to identify patterns and analyze vast amounts of data enables companies to uncover cost-saving opportunities, whether through optimizing travel policies, identifying preferred vendor compliance, or negotiating better rates based on volume and travel patterns, among other applications. As AI becomes ingrained in the enterprise, companies are rapidly adding prescriptive capabilities to the predictive models in broad use today, leveraging their massive data sets in new and innovative ways. Some examples of how data can be combined with predictive AI include:

  • Descriptive Analytics: dives into history, using data to answer “what, where, and when” through dashboards and visuals. Think flight delays: it analyzes past reports and then presents key insights on a clear dashboard.
  • Diagnostic Analytics: beyond the “what, where, and when,” Diagnostic Analytics delves deeper, unearthing the “why.” Imagine analyzing weather data and aircraft logs to diagnose the root cause of flight delays. These insights shed light on the past and empower AI models to predict future delays with improved accuracy.
  • Prescriptive Analytics: beyond just predicting the future, Prescriptive Analytics recommends how to thrive in it. It combines forecasts with insights, recommending the best actions. Imagine airlines using it to anticipate delays and automatically reallocate crews, minimizing disruptions before they even happen. It’s proactive planning powered by data and AI.

Efficiency and Data

Artificial intelligence – particularly machine learning, reinforcement learning, and generative AI – continues to reshape the travel industry rapidly. Large travel companies that invested in AI and data management early are beginning to separate from those waiting for or needing more resources and knowledge. With the investment in data management and the robust data sets to analyze, AI’s predictive and prescriptive capabilities would be significantly improved, underscoring the critical role of high-quality data and data management solutions in driving forward-looking insights and actions.

Because of the scale and costs involved in building AI capabilities, the rest of the travel sector is lagging behind other major consumer-facing industries regarding AI adoption and implementation. A digital AI divide is beginning to emerge, driven by a number of reasons.

  1. Shortage of AI skilled workers
  2. Limited AI infrastructure and data networks (such as cloud computing)
  3. Lack of a formal ‘AI strategy’ alongside an overall business plan
  4. Reactive to the use of AI and waiting for others to move first versus being proactive and experimental

Smaller companies, in particular, need more capital and knowledge to invest in AI and the data to build and refine AI models at scale. The good news, however, is that we are still in the early stages of AI and that companies that embrace the change now will be positioned for success in the future. Traditional technology solutions providers in the travel industry are stepping up to fill the AI knowledge and technology gap, providing technical expertise and solutions to companies looking to integrate AI and data into their organizations.

AI starts and ends with data, and Cornerstone has been providing award-winning data solutions for the travel industry for over 30 years. Learn how our platforms, iQCX and iBank, can help your organization.  Contact us here.