Three years ago, flights were grounded, the economy was frozen, and we were experiencing unprecedented global lockdowns. Today, we are seeing a surge in international travel and the reinvention of the travel industry at large. In 2019, you couldn’t imagine such a dramatic series of events.

Another thing you likely couldn’t imagine? Generative artificial intelligence (AI).

Online travel agencies (OTAs) have started dipping their toes in the waters of generative AI. But in an industry rife with opportunity for customer service challenges (very few businesses consistently face unpredictable factors like tornadoes or blizzards), what role can generative AI play across the value chain for OTAs?

Let's dive a little deeper and explore some areas of implementation for generative AI along the customer journey and discuss some of the challenges and limitations of the technology in the travel industry.

Reducing friction and expanding possibilities

Through natural language processing, generative AI allows customers to express their needs, wants and desires in a creative and illustrative way, rather than checking boxes in an online search. The adaptability of generative AI allows it to surface the answers, information, and resources to meet a customer’s exact needs. Removing rigid parameters often found in search criteria opens up a world of possibilities for the customer.

Travel is not your typical service. You spend significant time planning for it, it is expensive and you do it for all sorts of reasons: for work, to visit family, to expand one’s cultural horizons or simply to disconnect. It is highly emotional and important. At the same time, planning a trip can become a tricky process. “Which activities should I go for, and do I need to reserve them now or once I am there? Should I keep comparing flight prices? Should I get travel insurance? Am I choosing the right hotel, or should I go for a different location?” In most cases, travel is meant to be a mental release, yet planning it can be quite stressful.

If we are able to rethink that experience such that it becomes more natural, fluid, and more adapted to the customer’s specific desires and interests – and on top of that, we keep it as easy as a person-to-person conversation – generative AI could become an essential component of the travel booking experience. By strategically utilizing generative AI, booking platforms will be able to offer more customized results and reduce friction in the customer experience.

Allocating human resources where they are needed most

The key to that streamlined experience is determining which customer requests can be successfully handled leveraging generative AI and which ones would require an agent.

We start by looking at repetitive, routine tasks; there is no real reason for agents to be handling these. In these situations, the agent does not bring any additional value because the customer will likely get a satisfactory outcome regardless of whether they are working with an agent or not. Ultimately, we want the customer to get what they want quickly and transparently. If there is no additional benefit in having an agent intervening, we would favor streamlining that interaction.

There are clear situations however where you do want an agent to be involved. When you are traveling, there are several atypical situations that can occur. You're stuck in the middle of your trip, there’s a visa issue, there’s a hurricane... you need the more complex issues to be handled by an individual. You need a person because human empathy is required.

Assuming we correctly perform triage between interactions that can be handled by generative AI from the more complex situations that require an agent, we will unlock an unexpected benefit: rather than making humans obsolete, we make agents’ jobs more fulfilling. Suddenly, you need more qualified people in these roles.

It's a win-win situation: the customer is satisfied, and the organization successfully delivered on its customer service promise efficiently and effectively.

Challenges and limitations

OTAs will need to develop the tools to assess and triage customer requests appropriately. It’s one thing to implement generative AI; it’s another to determine where and how to use it. This is a complex challenge that needs to be driven by an in-depth understanding of the nature of the customer request, customer behavior, and their needs.

The other significant limitation of generative AI is that currently, the information base does not go beyond 2021. Data that is two years late in the travel industry has the potential to be extremely dated and unreliable. However, once generative AI has access to the most up-to-date information, travel itineraries could be altered and updated in real time.

Another challenge is that today, the software is much like we are – it doesn’t consistently know the difference between reliable and unreliable information. This means that generative AI can present information as factual that is untrue, which could turn into a problem quickly when planning travel itineraries.

Regardless of these current challenges, if there is one innovation that is more than a fading trend, I believe it’s generative AI. That’s because the potential of what generative AI can do for the travel industry is nearly unlimited: natural language search, responsive and creative itineraries, customer service perfectly tailored for each situation, and up-to-the-minute adjustments and changes. The key to bringing this vision to life is, ironically, using our uniquely human ability to understand when and where humans need help from humans.