Each year during the summer travel period, customer experience operations face their biggest challenge yet. From thousands of passengers being rebooked and the pressure to move quickly within narrow time frames to high spikes in contact center volume, summer travel season is one of the most stressful times for CX organizations.
This summer, there is no debate over the presence of artificial intelligence in the contact center. It's been established. The debate, however, centers on whether companies have built AI-enabled systems capable of handling the challenges posed during this most stressful time of year, with frustrated passengers, constrained inventory, and highly complex circumstances.
Last year, discussions about the use of AI in CX centered on incorporating the emerging technology into CX processes to provide agents with assistive capabilities. However, AI has progressed far beyond that point. Modern-day agentic AI systems now resolve requests end to end using knowledgebases and organizational processes, resulting in complete resolution and closure without handoffs.
With the rapid advancement of AI, customer behavior has adapted as well. Customers used to avoid interacting with AI systems due to previous negative experiences and preferred to get help from human agents. Now that AI interactions yield positive experiences, customers opt for them to receive faster service.
In terms of large volumes that arise during travel disruptions, agentic AI resolution systems offer several advantages. For example, AI systems can handle many requests simultaneously, deliver fast responses, and ensure consistent, accurate solutions.
At the same time, though, the travel industry is a complex sector characterized by limited inventory and changing conditions, in which airlines or travel operators must consider multiple factors simultaneously. Human agents add value here by understanding customers' concerns, assessing the situation, and providing some flexibility by deviating from the usual workflow.
AI systems differ significantly in CX performance. Their main advantage is consistency. Customers interact with a consistent system that enforces all relevant data and policies, regardless of volume. This consistency becomes especially valuable during travel disruptions.
Expectations Are Rising Too Fast
The rise of advanced conversational AI solutions raises customers' expectations for interactions. Unlike older self-service or voice menu solutions that were associated with negative experiences, advanced AI interactions create the expectation that issues will be solved quickly, automatically, and with a high degree of personalization. Customers expect systems to recognize the issue and provide a solution right away.
However, the best conversation in the world won't solve customers' problems if a particular problem is unsolvable in conversation. In situations where tens of thousands of people are vying for rebooking, neither machines nor humans can create seats where none exist. Here, efficiency means solving a logistical problem.
Until recently, many CX operations evaluated their success based on deflection rates: how many times customers' problems were resolved without human agent involvement. The idea sounded sensible in the previous era of CX technology, when the main focus was on volume management and minimization of human agent interactions. Nowadays, the right question to ask is whether the problem was really solved and how well it was done.
Agentic AI improves overall handling effectiveness by providing necessary information about context, surfacing data, narrowing down the choices, and prepping interactions beforehand. The remaining tasks of human agents become more focused on assisting customers with complicated or emotionally charged issues.
In times of increased volume, the key problem is typically not in the AI solution itself but in its ecosystem.For any travel booking application, all other parts of the infrastructure are essential for a successful interaction: third-party booking platforms, APIs, inventory management systems, and more. As soon as one of the links breaks down or runs into trouble, it takes only milliseconds for the customer experience to noticeably deteriorate.
And the second challenge is unpredictability: In real life, customers don't behave the way they do in testing environments. They frequently shift their destinations during conversations, make several requests at once, and raise new use cases that highlight the limitations of the current workflow and integration solutions.
Those organizations that will survive the summer travel peak season can adapt quickly, analyze the issues, and respond fast.
As much as it might seem like a decisive moment, the summer travel season won't answer the question of whether AI is a valid CX tool. Instead, it'll reveal how well different companies prepared their solutions for actual operation in a busy environment. In terms of travel, the difference between resilience, reliability, and adaptiveness matters a lot more than the quality of demos or ideal test scenarios.
Sara Hanley is senior vice president of marketing at ASAPP.