As AI search takes over, airline websites are an interesting battle ground

We’ll get to travel soon – but first, it’s worth taking stock of the enormous changes being wrought to smart web strategy by AI.

For one thing, users are already turning to their favorite GPT instead of traditional search engines, often finding richer contextual answers. For another – and partially in response to this shift in search behavior – search engines are already using AI to anticipate intent, playing the word match game.

The upshot: keyword stuffing is already a thing of the past, and the search engine ranking game is seeing its rules change before our very eyes. Take Google’s AI overview, previously known as the Search Generative Experience (SGE), already available to limited users in the US with plans to expand to more countries soon. The feature answers user queries with a results box synthesised from top results, without the need for a click beyond Google.

Interestingly, websites that rank number-one are not always cited in these overviews. That’s another seismic shift.

Designing for the robots

Since the birth of the internet, a website has always been a collection of pages a brand created so that a human could use, essentially self-serving the content they want to see. With AI this ‘traditional’ structured approach is becoming less relevant. Just like in a high-street store, it may be easier to ask a question than to search through the store yourself.

Brands need to start thinking about search differently. Brands have to be ready to go to market where their number-one customer profile is an AI robot. Brands’ product information and data needs to be optimized for ‘robots’ to understand it, if they want the learning language model to ‘think’ of them as the best answer to the question.

Airline websites: A unique test case

These shifting dynamics have particularly interesting ramifications in the travel industry.

When we’re booking flights or accommodation – for either leisure or business – we rarely have a fixed date, time or even location that cannot be changed if the price or package suits. We might happily fly out the day before or stay at a different hotel if shown a benefit to doing so. Even when choosing a sunny holiday destination, we might start research looking for holidays in Europe but end up in the Caribbean.

AI search, and its different approach to surfacing information, will have a huge influence on that travel buying process.

Travel companies have long relied on global distribution systems (GDS) to access and manage travel-related inventory, such as flights, hotels, and car rentals. Over the last decade or so, online travel agents (like Expedia and Booking.com), ‘metasearch engines’ (like Skyscanner and Trivago) and many other travel APIs and aggregators have created a vast global ecosystem of channels through which a travel company’s ‘product’ reaches its audience.

The outward-facing interfaces of these channels are how both humans and crawlers see them. In our new AI-search world, brands need to understand how the key LLMs (like ChatGPT, Gemini, and Claude) will interpret their product information and use it to ‘corroborate’ an AI generated answer to a question.

Most travel brands would rather sell direct to the consumer than via a third party. Equally, the user experience of the airline’s website, compared other channels, needs to provide a friction free experience for both human users and AI.

If a travel brand wants to win the fight to be the ‘best answer’, it needs to learn how to structure and share its product information – but also ensure that AI understands its context and why it is the best answer to the question. Having a clear data schema and a good content strategy is arguably not a new concept for marketers, but now might be the time we all need to think differently about what that means in the age of AI.

For more on travel, tourism and the auto industry, head over to The Drum’s dedicated focus week hub.