Could AI Agents Turn the Time Between Trips Into a Revenue Channel?

According to a recent Skift Research survey of nearly 7,000 global travelers, the modern path to purchase has become exceptionally complex. The discovery phase is now highly fragmented, spanning traditional search engines, social media platforms, and emerging AI-powered planning tools. In fact, more than 60 percent of travelers who are aware of AI tools are already actively using them to research and plan their itineraries.

However, actual bookings remain concentrated on a select few channels where price competitiveness, brand trust, and payment flexibility are the deciding factors. For airlines and travel suppliers, this fragmentation presents a significant commercial challenge: how to capture a traveler’s attention during a highly disjointed discovery process and maintain a direct, meaningful relationship all the way through to booking, boarding, and beyond.

The Fragmentation of the Modern Booking Journey

The modern traveler’s journey is rarely linear. A consumer might spot a destination on Instagram, ask an AI search engine for a five-day itinerary, compare flight options on an Online Travel Agency (OTA), and eventually purchase a ticket directly on an airline’s website.

Throughout this fragmented sequence, the traveler may encounter the same airline across multiple touchpoints, yet these interactions are almost always treated as isolated events. When a customer moves from an OTA to an airline’s mobile app, or from a social media ad to a customer service chat, the context is typically lost. The customer is forced to start their search or service request from scratch.

This lack of continuity is more than a customer service annoyance; it is a major driver of customer churn and lost direct-booking revenue. When airlines fail to connect the dots between these touchpoints, they lose the ability to nurture the relationship, personalize offers, and build long-term loyalty. The opportunity for travel brands lies in bridging these disconnected experiences, transforming isolated transactions into a continuous, ongoing conversation.

Beyond the Chatbot: The Rise of Long-Running AI Agents

To solve this fragmentation, forward-thinking airlines are looking beyond the traditional, rigid chatbots of the past. Early-generation customer service bots were built on deterministic decision trees, capable of answering basic FAQs but notoriously poor at handling complex, multi-step scenarios. If a query deviated slightly from the pre-written script, the system failed, frustrating the customer and forcing a transfer to a human agent.

Enter "long-running" AI agents. Unlike their predecessors, these advanced agents are designed to retain context over days, weeks, or even months. They can operate across multiple channels and departments, acting as a persistent digital companion for the traveler.

"Long-running AI agents are designed to stay with both the customer and the task over time," says Erik Zahnlecker, agent product manager at Sierra, an enterprise AI company. "This kind of continuity can connect planning, booking, service, loyalty, and the months between trips into a more persistent customer relationship."

Rather than treating each interaction as a clean slate, a long-running agent remembers past conversations, understood preferences, and ongoing issues. If a traveler starts planning a family trip on a desktop computer on a Monday, asks a follow-up question via SMS on Wednesday, and finalizes the booking on a mobile app on Friday, the AI agent carries the thread of that conversation seamlessly across every device and channel.

The Power of Omnichannel Memory

This level of continuity requires a fundamental shift in how customer data is managed and utilized. Traditionally, an airline’s data is siloed across various legacy systems: the Passenger Service System (PSS), the Customer Relationship Management (CRM) platform, and the loyalty database.

AI agents act as an intelligent orchestration layer on top of these disparate systems. By retaining context, the agent knows who the customer is, remembers their prior interactions, and understands their current intent.

This capability also enables seamless channel-hopping. "The most advanced version has agents working across multiple channels, including chat, email, SMS, and voice," Zahnlecker notes. "A traveler could say, ‘I’m on the phone, but I need to go. Can you text me?’ and pick up the conversation there."

This level of flexibility reduces friction for the traveler, who no longer has to repeat their booking reference, personal details, and specific issues to multiple representatives across different platforms.

Managing Chaos: How AI Transforms Flight Disruption Recovery

Perhaps the most critical test case for long-running AI agents is irregular operations (IROPS). When severe weather, mechanical failures, or air traffic control delays ground flights, airlines face an operational and financial crisis. Contact centers are quickly overwhelmed, hold times spike to hours, and passenger frustration mounts rapidly.

In a traditional disruption scenario, the burden of resolution falls heavily on the passenger. They must monitor flight status apps, queue at airport customer service desks, or wait on hold to find an alternative flight.

A long-running AI agent completely flips this dynamic. Because it is integrated directly into the airline’s operational systems, the agent can detect a flight cancellation signal in real time and proactively reach out to affected passengers before they even realize their travel plans have been disrupted.

[Flight Cancellation Signal]
          │
          ▼
[AI Agent Proactively Reaches Out via SMS/App]
          │
          ├───────────────────────────────┐
          ▼                               ▼
[Passenger A: Needs Urgent Travel]  [Passenger B: Flexible Schedule]
  - Filters: Business Class only      - Filters: Economy is fine
  - Action: Books next available      - Action: Books next day + hotel voucher
          │                               │
          └──────────────┬────────────────┘
                         ▼
          [Context Carried to Claims/Refunds]
                         │
                         ▼
          [Post-Trip Loyalty Re-engagement]

From there, the agent can navigate complex rebooking scenarios by weighing the specific needs of the traveler against the airline’s real-time inventory and commercial rules.

"The agent can navigate both the traveler’s needs and the airline’s rules to find the best available flight and ultimately get them where they need to go," Zahnlecker explains.

For instance, the agent can distinguish between a high-value business traveler who must reach a meeting today at any cost, and a leisure traveler with a flexible schedule who might willingly accept a travel voucher and a flight the next morning. It can automate these highly nuanced negotiations at scale, resolving thousands of passenger disruptions simultaneously without human intervention.

Crucially, the agent’s job does not end when the new boarding pass is issued. If a passenger is entitled to a meal or hotel voucher, or needs to file a baggage claim, the agent carries that context forward. It can automatically initiate the reimbursement process or monitor seat inventory to move a displaced passenger back into their preferred cabin class if a seat opens up later.

By treating a disruption as a single, continuous journey rather than a series of disconnected headaches, airlines can preserve customer goodwill and protect their bottom line during operational crises.

Capitalizing on the Dormant "Between-Trip" Window

For most airlines, the period between a customer’s trips is a commercial dead zone. Once a traveler steps off the plane, active communication from the airline typically ceases, replaced by generic, automated marketing emails that suffer from low open and engagement rates. The airline essentially loses touch with the customer until they decide to book their next flight.

This post-trip window represents a massive, untapped revenue opportunity. Skift Research on destination loyalty highlights the strategic importance of this specific period, identifying the 30 days immediately following a trip as a critical window for customer engagement. In fact, 52 percent of surveyed travelers stated that receiving personalized recommendations for a future visit during this post-trip window would be highly valuable.

This is where proactive AI agents, such as Sierra’s Horizon agents, can fundamentally change the dynamics of customer retention. Instead of sending static blast emails, these agents can monitor subtle customer signals and initiate personalized, highly relevant conversations.

"The more personalized the interaction, the more likely someone is to engage and see the agent as a trusted concierge," says Zahnlecker.

An agent might note that a customer recently completed a family vacation to Hawaii. A few weeks later, the agent could proactively reach out via the customer’s preferred channel to highlight a localized fare drop to a similar beach destination, or suggest an itinerary for an upcoming anniversary based on the traveler’s historical preferences.

By transforming passive marketing into an active, tailored conversation, airlines can stimulate travel demand that might otherwise remain dormant, driving direct bookings and increasing customer lifetime value.

Personalization and the New Economics of Ancillary Sales

Beyond booking flights, airlines rely heavily on high-margin ancillary products—such as cabin upgrades, extra baggage, in-flight Wi-Fi, lounge access, and third-party partnerships like rental cars and hotel bookings.

Historically, airlines have struggled to market these add-ons effectively. Offers are often presented in a generic, one-size-fits-all format during the online checkout process, where travelers are highly price-sensitive and focused solely on the base fare.

An AI agent equipped with continuous customer context can dramatically improve the conversion rates of these ancillary offers by introducing them at the right moment in the travel timeline.

For example, an agent might wait until 24 hours before departure—when a traveler is preparing to pack—to offer a discounted checked bag. Alternatively, if the agent knows from a previous interaction that a traveler values productivity, it can offer a bundled Wi-Fi and lounge pass package during the check-in window.

Zahnlecker notes that these offers can be dynamically shaped by customer data and past outcomes. One customer may have a high propensity to purchase premium seat upgrades, while another might be a prime candidate for a co-branded credit card sign-up or loyalty program enrollment.

The financial impact of these optimized interactions can scale rapidly. Even a minor percentage-point increase in ancillary attachment rates and repeat bookings can translate into millions of dollars in incremental high-margin revenue for a major carrier.

This commercial potential is already being demonstrated in adjacent sectors of the travel industry. A leading travel platform integrated a Sierra AI agent across its website and mobile app to help members navigate their plans, understand benefits, and identify overlooked value.

According to Sierra, this context-aware agent helped drive a 5 percent increase in plan retention while maintaining a stellar customer satisfaction score of 4.7 out of 5. While this specific case study focused on membership retention rather than airline ancillaries, it underscores a universal truth: when customers receive highly personalized, context-rich support, both customer satisfaction and commercial outcomes improve.

Balancing Engagement and Noise: The Importance of Relevance

While persistent engagement offers immense promise, it carries a significant risk: the danger of over-communication. If an AI agent reaches out too frequently, or with irrelevant offers, it quickly becomes indistinguishable from spam, leading customers to mute the channel or uninstall the airline’s app.

To prevent this, AI agents must rely on precise operational and customer data signals. An agent should only initiate contact when it has a highly relevant, high-value reason to do so.

"Over time, those conversations can build a richer customer profile that the agent can act on," Zahnlecker explains.

For instance, if a traveler previously indicated during a chat conversation that they preferred extra legroom, but none was available at the time of booking, the agent can silently monitor seat inventory. If a premium economy seat opens up due to a last-minute cancellation, the agent has what Zahnlecker calls "a reasonable moment to reach out."

By ensuring that every proactive message is rooted in a specific customer preference and a real-time operational update, the airline can position its AI agent as an indispensable, personalized travel concierge rather than an annoying marketing channel.

A Unified Challenge Across the Travel Ecosystem

While the operational complexity of airlines makes them a natural test bed for these technologies, the need for continuous, context-aware customer relationships extends across the entire travel and hospitality ecosystem.

• Hotel Groups: Hospitality brands are looking to connect the disparate phases of the guest journey—linking central reservation systems and loyalty databases with pre-arrival upgrade offers, digital check-in, on-property concierge services, and post-stay feedback loops.
• Online Travel Agencies (OTAs): OTAs are deploying AI agents to handle complex customer service workflows, such as multi-provider cancellations, refunds, and rebookings, aiming to reduce the high cost of human customer support.
• Cruise Lines: Cruise operators have a unique opportunity to use AI agents to manage a highly complex customer lifecycle, guiding guests from initial cabin booking through pre-cruise shore excursion sales, onboard dining reservations, and post-cruise re-engagement campaigns.

Across all of these sectors, the core challenge remains the same: breaking down the internal organizational and technical silos that prevent customer data from flowing smoothly across service, operations, and sales channels.

A Strategic Roadmap: Starting Small to Scale Safely

For travel executives, the prospect of deploying a comprehensive, end-to-end AI agent can feel overwhelming. Integrating AI across legacy reservation systems, loyalty platforms, and customer service centers requires significant technical coordination and cross-functional alignment.

To mitigate risk, Zahnlecker recommends a phased, pragmatic approach to implementation. Rather than attempting a massive, all-at-once launch, travel brands should start with highly focused, high-volume use cases.

"Focus on a few use cases that are maybe not your most critical, but where there’s enough volume to learn quickly, build trust, and improve the experience," Zahnlecker advises. "Then scale from there."

Starting with a pilot project—such as deploying an agent to handle a small percentage of chat traffic for basic booking modifications—allows airlines to test the technology, refine the conversational design, and build organizational confidence. Once the foundation is proven, the airline can progressively integrate the agent with deeper systems, expanding its capabilities into disruption management, proactive sales, and long-term loyalty building.

The competitive window for establishing these capabilities is narrowing. As generative AI continues to mature, travelers will increasingly expect brands to understand their preferences, remember their history, and anticipate their needs. Airlines and travel suppliers that invest in building a strong foundation of context-aware, long-running AI agents today will be uniquely positioned to build deeper customer relationships, secure direct bookings, and unlock powerful new revenue streams in an increasingly fragmented digital world.