Why OTAs Are Winning The AI Flight Search Race

The Rise of AI in Flight Search

Travelers are increasingly using ChatGPT's generative AI to search for flights, a trend that is rapidly growing and reshaping the digital landscape for the airline industry. According to analysis from PROS, an airline retailing and revenue management company, ChatGPT accounted for 2.3% of referral traffic to airline websites from search engines in September. While this number might seem small, its growth is explosive. It's an increase from just 1.8% in August and represents a staggering 230-fold jump since November of last year when OpenAI first launched its service. Currently, ChatGPT is the undisputed leader in this space, driving 95% of all AI engine referral traffic to airlines. However, this dominance is expected to face challenges as other tech giants develop their own AI search competitors.

The Imbalance Favoring OTAs

Despite this new wave of traffic, a closer look reveals a significant problem for airlines. Within ChatGPT's conversational responses to flight queries, the AI overwhelmingly favors Online Travel Agencies (OTAs) and metasearch sites like Kayak and Expedia. An analysis of referral links showed that by a wide margin, ChatGPT sends more traffic to these aggregators than directly to airline websites. The data is stark: 41.8% of brands mentioned in flight-related chats did not receive an associated link. For airlines specifically, this unlinked rate was a massive 74.6%. Instead of linking to the carrier, ChatGPT redirected users to OTAs, metasearch engines, and in 5% of cases, even to Wikipedia. In contrast, OTAs and metasearch sites saw unlinked rates of just 8.8% and 8.6%, respectively.

A Shifting Landscape for All

Enmanuel Tirado, PROS's head of SEO and digital analytics, warns that this trend is an emerging problem for airlines hoping to maintain their market share against aggregators. Vik Krishnan, a partner at consulting firm McKinsey & Co., notes that this is one of many unintended consequences that will arise as conversational AI encroaches on the traditional link-based results of search engines. But airlines are not the only ones at risk. OTAs and metasearch sites, despite their current advantage on ChatGPT, face their own threats. Krishnan points out that many websites reliant on click-through traffic have already seen visibility drop by 20% to 30% as Google incorporates AI-powered overviews above its standard search results. This shift threatens any business, from a large OTA to a small Mexican hotel, that relies on traditional SEO and Google ads for traffic. Even Google is feeling the pressure, exploring ways to replace lost ad revenue, such as by integrating paid advertising into its AI overviews. The script for how AI will monetize user attention is still being written.

How Airlines Can Compete in the AI Era

So, what can airlines do to level the playing field? One key reason ChatGPT favors aggregators is their wealth of comparative flight and fare data. Tirado suggests that airlines can counter this by creating dedicated, detailed webpages for each individual flight. These pages should include comprehensive fare information across different cabins, product details, flight durations, and schedules. Furthermore, airlines need to adapt their technology stacks to be more accommodating to AI, ensuring that ChatGPT's bots can effectively access and crawl their webpages for information.

The Future of AI Travel Search MCP

Looking ahead, a new technology standard may hold the key. Mike Coletta, a senior manager at Phocuswright, explains that AI engines currently struggle with accessing the most up-to-date pricing for flights and hotels, which is typically held within Global Distribution Systems (GDSs), OTAs, and airline websites. However, a technology called Model Context Protocol (MCP), released by AI research firm Anthropic, allows AI engines to pull this live data by sending structured queries directly to travel suppliers and aggregators. Pioneering companies like Turkish Airlines, Kiwi, and Sabre are already building out MCP servers to grant this access to AI engines. As the battle for attention intensifies in this new AI-driven search environment, becoming a leader in MCP adoption could provide a crucial competitive advantage. Coletta suggests it is a strong possibility that every travel company will eventually need an MCP server to stay relevant.