“AI and machine learning are paving a new direction for all of us in our daily lives – how we live, how we work and how we travel,” says Darrin Grafton, CEO of Serko. “The business travel industry has adopted AI and machine learning tools for virtually every aspect of the travel process, from researching and planning a trip, to traveler support and post-trip reconciliation.”
While artificial intelligence has been in the media for years, it has recently matured to a level where it’s being used more prevalently, according to J.B. Walsh, chief technology officer at Christopherson Business Travel. In the area of deep learning, advances in both hardware and cloud services have allowed real-time analysis of much larger data sets than ever before, the engine that drives AI.
With machine learning, a growing marketplace of off-the-shelf tools provides new avenues for experimentation. And advances in natural language processing engines have led developers to apply AI services in a variety of new platforms. “Ultimately, all of these advances mean it’s much less expensive today to invest in AI than it was five years ago,” Walsh says. “So the adoption of AI is growing rapidly.”
No More Tradeoffs
An ongoing challenge may be agreeing just what constitutes artificial intelligence. “True AI is different than machine learning, which is an emerging technology that is still in early phases of development, deployment and adoption,” says Michael Kubasik, executive vice president and chief information officer at Travel and Transport. “The maturity curve will evolve with the intent that automation will improve the traveler experience and create efficiencies for travel companies.”
When the combined benefits of the two are considered, it’s hard to overestimate their potential impact. “AI and machine learning are the lynchpins which allow travel managers to get away from having to make a tradeoff between tightening rules to increase compliance, and employee satisfaction,” says A.G. Lambert, SAP Concur senior vice president, spend, data and analytics.
Employing these advanced technologies, travel buyers can make business travel easier by unlocking greater compliance, security and savings while also improving employee productivity and satisfaction. “There’s currently a lot of experimenting on where it can best make an impact,” says Sam Hilgendorf, chief information officer for Fox World Travel.
Hilgendorf says there is clear investment and development in the traveler booking experience through the use of chatbots and predictive analytics. He also points to strong momentum with advancements in virtual assistants using personalization for intelligent recommendation services.
Such developments enhance opportunities for the travel industry at large to better deliver a hassle-free, end-to-end travel experience, according to Jay Richmond, head of business travel group, North America, Amadeus. He notes that bots using natural language processing can facilitate any real-time interaction or human support required by travelers.
“What started out as an automated tool to answer FAQs is now being used by many travel firms to help customers book and manage travel, as well as for cross-selling and upselling,” he says.
Richmond notes that machine learning technologies are being leveraged to fine-tune search results based on implicit and explicit traveler preferences and past trips, displaying tailored options on favorite travel websites and apps. Business travelers can use technologies powered by AI to capture receipts with their smartphones, while machine learning algorithms categorize expenses and spending patterns.
Currently, disruption management is a key pain point for travelers that invariably means ditching the online channels and turning to an airline counter or call center for personal help, Grafton says. “This aspect of travel represents an enormous opportunity for AI to intelligently review all possible options and automatically rebook a traveler for a seamless, stress-free experience.”
He notes that travel booking platforms can autonomously create personalized itineraries which are not only time-saving for travelers, but also help companies perform better in managing their travel budgets. “These technologies are also helping companies make improvements within their travel program through a well-defined analysis of their data, combined with automated recommendations,” he says.
The capacity for self-servicing offers increasing potential for efficient travel management, Walsh adds. With traditional processes, servicing individual traveler needs can be time consuming and the work load can fluctuate dramatically. Using AI in bots and other applications allows business travelers to self-service their own bookings, saving travel managers time and helping to mitigate variations in workload. Cost savings are also possible through measures such as uncovering seasonal or other less obvious trends.
Moving Ahead
So just where is artificial intelligence headed next?
Brian Beard, president of DVI, points out that considering AI’s place near the top of the overused buzzword list, it’s easy to have inflated expectations. But it seems reasonable to expect that a form of AI will replace the way that both leisure and business travel is booked.
“We can also reasonably expect to see AI being used to monitor duty of care functions and business travel policy compliance functions,” he says. “An AI that continually monitors systems and makes decisions based on situations is vastly more efficient than doing it manually.”
With AI creating all this improved efficiency in business travel programs, presumably traveler satisfaction will increase without the need for additional manpower. It will also provide more readily available data. “Having an intelligence updating you on various aspects of your travel program in real time will have significant advantages to the current model, where we wait for information to be compiled at the end of the month then need to analyze it for actionable information after the fact,” Beard says.
While much attention is being placed on front-end traveler tools, Hilgendorf notes that creating opportunities to tie the travel program back to actual business results will be more impactful. “Better understanding the influence the travel program and policy has on sales and operations, employee retention and customer satisfaction will create greater long-term value for the organization,” he explains.
In its 2020 Travel Trends, Amadeus identified artificial intelligence as a game changer for customer service and communication. “I certainly see the prevalence of AI increasing in the next year, but not at the expense of human interaction,” Richmond says. “AI-enabled service will complement human interactions that are the hallmark of the travel industry.”
Lambert predicts that in the coming months, machine learning will move from a novelty to a routine function as it proliferates under the hood of technology services everywhere. “In a way, it will fade into near invisibility, even while making a huge impact,” he says. “Technology services will increasingly anticipate your needs, whether they be related to expense reports, scheduling or other processes.”
Eventually this means faster turnaround on tasks such as creating a preferred travel itinerary based on previous trips, and automatically combining the employer’s preferred vendor commitments. Similarly, predicting the location of expenses based on cities previously visited will expedite completion of expense reports.
Of course challenges remain, including the likelihood that AI will create issues not previously experienced. “We tend to think of AI as a single entity, but in fact there will probably be many independent AI systems,” Beard notes.
For example, a travel agency may have developed an AI to aid in the booking process, an airline might have another AI, and hoteliers might have their own AI systems. How these different AI systems communicate with each other and then communicate to travelers, travel managers and others is still unclear. Among other questions is how conflicting messages will be handled.
“It gets a little crazy when we start thinking about it,” Beard says.
At the same time, some current limitations are likely to be overcome in the near future. “Today, AI requires large sets of data in order to find meaningful patterns,” Walsh says. “Furthermore, the data still needs some level of intervention by people in order to identify what data is important and what data isn’t.”
He anticipates the focus over the next five to ten years will likely be to improve in these key areas, with advances allowing AI applications and services to run on less powerful devices and make real-time decisions with little historical information. That should help alleviate growing issues around storing large amounts of personal data. It may also open the doors for more advanced self-service bots, innovative disruption management services and new fraud detection techniques.
Keying In
It’s important to realize that AI projects can be costly if not planned out properly, Walsh points out. “Start by understanding what data you have available to your organization and how it could potentially be used to drive additional revenues or lower expenditures,” he says. “Perform some experiments using the data you have available, so that you can estimate your potential yield should you invest further.”
Working with AI also requires patience and an adjustment in self-service, Kubasik says. “However, travelers should expect to see continued contextual information delivered to them and not have to dig to see if a hotel has a gym or restaurant, or attractions they like are nearby,” he predicts.
“A solid strategy for the present is pushing partners for greater insights and intelligence,” Hilgendorf says. “Challenge them to tie travel data together with your HR data, CRM data, and other relevant data sources.” He advises staying aware of consumer travel tools and asking partners how they intend to provide these types of services in the future. Take a close look at traveler needs and existing policies, and keep these emerging solutions in mind.
“Don’t sit on the sidelines waiting,” Lambert advises. “Look at how AI can improve your travel program today, and at how your peers are using it to their advantage.”
