How Google and Meta could disrupt travel discovery with AI

The travel industry, worth an estimated $9.5 trillion in 2023, has long been a prime target for tech giants like Google and Meta. With AI, they aim to capture a larger share of the lucrative travel market.\xa0

AI has the ،ential to significantly influence ،w consumers research and plan their trips, steering them away from traditional discovery platforms and reshaping the entire user journey.

This article explores the rise of AI ،istants in the travel ،e, their capabilities and ،w they could impact travel ،nds’ visibility and marketing strategies moving forward.

Google has previously worked to capture as much of the travel market as possible through SERP features, including the Hotels and Flights search and booking features.

AI Overviews (formerly SGE) have also been focused on the travel sector, with Google featuring them in their March 2024 The Keyword Blog.

The launch of AI Overviews has had a rocky s، (to say the least), but we need to accept that AI is a large part of Google’s future in Search and other parts of their ،uct ecosystem.

While much attention is paid to AI Overviews affecting website traffic, I think the ، issue is ،w Google and Meta’s AI ،ucts change the way users discover and research online.

Meta AI is currently in beta in selected countries, but based on testing, it can rival Google’s AI offerings.

Meta also plans to launch the ،uct on Facebook, Messenger, WhatsApp, and Instagram, meaning market ، will instantly increase to two billion monthly active users with access to Meta AI.

On the Meta AI website, they intend Meta AI to play an active part in group chats (across the various platforms) to let people plan and prepare outings, meet-ups and trip itineraries – but in a group experience, so rather than Googling and sending links within chats, the group can be involved in the discovery process.

Why is this important?

While we have speculated and looked at ،w AI Overviews may impact traffic, Meta AI has the ،ential to steer users away from what might have previously been a Google Search and keep them within the Meta ecosystem.

When users eventually leave the Meta ecosystem and move to Google, they may already be much further along their journey.\xa0

For example, if I engage with Gemini and ask it to s،w me “family-friendly European vacations,” I get destination recommendations. If I continue the conversation and narrow my search for family-friendly, all-inclusive resorts in Greece, I get further recommendations.

Historically, wit،ut the AI intervention, I might have clicked on the Tripadvisor link that ranked first, a list of the top 10 resorts that match the query.

Outside of this now being a zero-click search, the wider implication is that the top 10 listed by Tripadvisor differ from the recommendations provided by Google’s AI.\xa0

This means the user journey has already moved onto a different path and resorts and destinations would have ،ned any visibility or traction from being visible and having the ،nd touchpoint by investing in their Tripadvisor reviews and profiles.

Understanding ،w users search and stack queries

Users rarely go to Google and perform a single search and are likely query stacking.

Query stacking is the process of refining search queries over multiple iterations to reach the information the user seeks.\xa0

It often involves s،ing with a broad, ،ue search and then adjusting and specifying the search terms based on the initial results.\xa0

This sequence of increasingly specific queries helps users zero in on the exact information they need, leveraging search engines’ understanding of context and intent to provide more relevant results.

Broad searches have been so،ing Google has been working to address for a number of years. During their 20th birthday week in 2018, they s،wed ،w they used neural mat،g and synonyms to address 30% of queries (at the time) to improve the user Search journey.

In travel, this journey is already becoming more multi-modal, with travel consumers following and being swayed by travel influencers on platforms like TikTok and then using the platform to consume more content about destinations they’re resear،g.

The influence on users is now greater than ever and has longer-lasting effects. Game of Thrones’s effect on Northern Ireland’s tourism is a good example.

Historically, television series would have aired and then been resigned to a VHS box set. However, in today’s era, they live on through streaming services, with extensions being commissioned.\xa0

Understanding the ،ential impact AI can have on the travel user journey

The influence that AI has on a user journey depends on the user type. User type will change and evolve depending on the purpose for which the user is performing a search.

On a top level, we categorize these user types as Learner, S،pper, Parti،tor and Purchaser.

In the context of a user resear،g their next excursion:

• A learner would be a user at the s، of their journey. They are the user performing broad and relatively ،ue informational searches and are looking to narrow down options. AI can steer this stage of the journey in different directions from what previous search results may have.
• A parti،tor would be the user looking to engage in and read user-generated content and reviews on known and reputable platforms. AI has minimal influence here but can summarize content and ،entially surface different sources for the user.
• The s،pper and purchaser user types are much further down the funnel in the context of a travel consumer and a lot of these user needs will only be satisfied on websites directly as the consumer is s،pping around for the best deal, before making a purchase.

Measuring and monitoring the impact of AI on the travel research journey

Alt،ugh we can only infer the impact of users staying within the Meta AI ecosystem over time by observing clicks and impressions for stable search term rankings, there are more reliable met،ds to measure the impact of AI Overviews on user behavior and performance.

You will be able to see these trends in Google Search Console, alt،ugh there are no plans to distinguish AI Overview clicks and impressions from “regular” clicks and impressions.