Travel Discovery Is Shifting to Generative AI

The New Front Door to Travel Is Conversational

A traveler planning a long weekend no longer solely opens a browser and scrolls through listicles. A growing share opens ChatGPT, Gemini, or Perplexity and asks: “Best destination for a fall long weekend within four hours of Boston?”  The response they receive is not ten blue links — it is a synthesized shortlist of two or three options with a rationale, a suggested itinerary, and accommodation recommendations.

Nearly four in ten U.S. travelers (38%) used generative AI to research trips in the past 12 months,  up 11 points year-over-year.¹ For DMOs and tourism operators, this shift changes which destinations get considered, which attractions get booked, and where in-destination spending flows. The marketing implications are direct and measurable.

The 2025 Snapshot: What Changed in a Single Year

Adoption is no longer niche

Phocuswright’s 2025 data marks a turning point: generative AI moved from early adopter territory into mainstream travel research behavior. The 11-point year-over-year increase in U.S. traveler AI usage is not incremental — it reflects a behavioral shift across demographics.¹  Importantly, AI users skew toward the higher-value travel segments destinations work hardest to attract: longer trips, higher spend, and more complex itineraries.

The research channel mix is shifting

At the same time, Phocuswright’s forward-looking data documents something that search-focused marketers need to take seriously: traditional search engine share among travel research channels fell between late 2024 and the second half of 2025, while generative AI platform usage rose.² This is not the end of search — but it is the beginning of a channel mix that destinations must actively manage.

GenAI is now a measurable traffic source

Adobe’s analysis of U.S. travel site traffic quantifies the scale of this change: AI-driven referrals to travel websites increased 3,500% year-over-year as of July 2025. ³ That number warrants a second read. It reflects a channel that barely registered in analytics tools 18 months ago and now drives a significantly growing portion of new visitor traffic to destination, hotel, and attraction websites.

Behavior Shift #1: From “Search and Browse” to “Ask and Shortlist”

The structural change in how travelers research is best described as the shortlist effect. Where a search engine returns ten or more results for a traveler to evaluate, a generative AI model synthesizes available information and returns two to four options with reasoning. The traveler’s consideration set narrows before they ever reach a brand’s website.

For DMOs and operators, this changes what drives selection. AI models favor destinations and businesses with clear, consistent, factual information they can cite: verified operating hours, structured itinerary content, credible third-party mentions, and accurate listings across platforms. Destinations that show up in AI recommendations are not the ones with the most website traffic — they are the ones whose information is most reliable and most citable.

Examples of prompts DMOs and operators should expect travelers to use:

Destinations: ”What’s the best time to visit [destination]?” / “3-day itinerary for [city]” / “Hidden gems near [major metro]”

Attractions and tours: ”What outdoor activities fit a two-hour window with kids under 8?”

Restaurants: “Where should I eat near [neighborhood] tonight — outdoor seating, good wine, no reservation required?”

Each of these prompts surfaces in AI responses as a shortlist. If a destination, attraction, or restaurant is not consistently represented in the data AI models train on and reference — official websites, review platforms, travel publications, maps — it will not appear in those responses.

Behavior Shift #2: AI-Referred Visitors Are Higher Intent

One of the most actionable findings from Adobe’s 2025 analysis is that visitors arriving from generative AI platforms behave differently from those arriving via traditional search. AI-referred traffic showed longer time on site, more pages per session, and a lower bounce rate.³  These are not casual browsers — they are travelers who have already received a recommendation and are arriving to validate a decision, not begin one.

The conversion data also shifted significantly. In October 2024, AI-referred visitors were 86% less likely to convert than other traffic sources. By July 2025, that gap had narrowed to 47%.³ The channel is maturing, and the pages AI sends travelers to matter more than ever.

“AI-referred visitors arrive having already received a recommendation. They are validating, not exploring — which means the landing page experience must meet a higher standard of specificity and trust.”

Practically, this means destination websites need to be built for intent confirmation, not just awareness. Itinerary pages, point-of-interest detail pages, direct ticketing links, and reservation pathways must be optimized for visitors who arrive informed and ready to act. Destinations that still rely on homepage traffic as their primary conversion surface will under-serve this growing segment.

Behavior Shift #3: GenAI Shapes Where Travelers Spend In-Destination

The influence of generative AI does not stop at trip planning. Adobe’s consumer survey data shows that 53% of AI travel users are asking about attractions, restaurants, and hidden gems, 45% use it for transportation planning, and 39% ask for local food recommendations.³ These are in-destination decisions that directly affect where visitor spending flows.

Deloitte’s 2025 Holiday Travel Survey adds category-level conversion data that DMOs and their stakeholder partners should bookmark: among travelers who used AI for planning, 67% acted on activity and attraction recommendations, 56% acted on destination recommendations, and 54% acted on accommodation recommendations.⁴  Restaurants show the highest conversion rate of any category — 55% of AI-generated restaurant recommendations resulted in an actual visit.⁴

The strategic implication for DMOs is clear: visitor economic impact is increasingly shaped by what AI recommends in-destination, not only what got a traveler to the area. DMOs that curate and syndicate experiences and dining — not only lodging — are better positioned to capture the full value of AI-influenced travel behavior.

2026+ Trends: Two Parallel Shifts DMOs and Operators Must Plan For

A) Marketing-channel shift: from SEO-only to AI discovery management

The ongoing pressure on traditional search as a travel discovery channel , documented by Phocuswright’s channel-mix data², calls for an evolution in how destination marketers think about organic visibility. The emerging practice is called Answer Engine Optimization (AEO) or Generative Engine Optimization (GEO): building content and data infrastructure so that AI models consistently, accurately surface a destination’s offerings when travelers ask relevant questions.

The key levers are not fundamentally new, but they require more discipline: structured entity data, authoritative long-form content, consistent information across all listings and maps, a healthy review ecosystem, and credible third-party citations in travel media and publications. The destination that wins in AI discovery is the one that has made itself easy for a language model to understand, verify, and recommend.

B) Visitor-experience shift: AI as both planning layer and in-destination concierge

Phocuswright’s forward-looking survey found that 78% of travelers who have used generative AI say it improves their trip planning experience.⁵ A quarter to a third are interested in booking directly within a generative AI interface or letting an AI assistant handle bookings.⁵ The practical implication: accuracy and frictionless handoffs matter more than ever. A traveler who receives a recommendation with outdated hours, incomplete availability, or no direct booking link will encounter friction at exactly the moment they are ready to convert.

The 2026 Readiness Playbook

• Build an AI-ready destination knowledge base. Create a single source of truth for points of interest, like hours, seasons, fees, accessibility, parking, transit, and booking links, and keep that information consistent across your website, Google Maps, partner platforms, and OTAs.
• Publish prompt-aligned itinerary and decision content. Create themed itineraries (food, outdoor, arts, family, accessibility) and seasonal guides. Link directly to bookable inventory wherever possible, including tours, timed-entry tickets, and reservations.
• Strengthen trust signals. Operationalize review generation and responses across Google, TripAdvisor, and category-specific platforms. Surface credible third-party coverage, awards, and editorial mentions that AI models can reference.
• Instrument measurement for AI referrals. Adobe’s engagement data provides justification to segment AI referral traffic in Google Analytics 4 and attribution platforms. Track micro-conversions: “Reserve” clicks, “Get Directions” actions, itinerary saves, and partner referrals; not only final bookings.
• Prepare inventory for emerging AI booking layers. Standardize product descriptions, availability, pricing, and booking policies for tours and attractions. For restaurants, ensure menus, dietary attributes, and reservation pathways are accurate and current across all platforms.

A 90-Day Plan to Get Ahead of 2026

The 2025 data is unambiguous: generative AI now influences both travel discovery and in-destination decisions on a measurable scale. The destinations and operators that perform well in 2026 will not be the ones who simply waited to see how the technology matures; they will be the ones who made themselves recommendable, accurate, and easy to book before AI discovery became table stakes.

A practical 90-day starting point:

1. Audit your top AI prompts. Run the 10 to 15 most likely traveler questions about your destination through ChatGPT, Gemini, and Perplexity. Document where you appear, where competitors appear, and what information is cited or missing.
2. Fix factual consistency. Cross-check hours, fees, descriptions, and booking links across your website, Google Maps, and all partner platforms. Eliminate contradictions.
3. Publish 5–10 itinerary and guide assets. Prioritize the formats AI models surface most: themed itineraries, seasonal guides, and “best for” decision content.
4. Add measurement. Segment AI referral traffic in your analytics platform and establish a baseline for micro-conversion tracking.

The window to get ahead of this shift, rather than respond to it, is now.

Frequently Asked Questions

References

1. Phocuswright — Search Slips, AI Surges (U.S. traveler genAI usage, +11 points YoY). phocuswright.com
2. Phocuswright — Travel Forward: Data Insights and Trends for 2026 (resource mix shift; search vs. genAI platforms). phocuswright.com
3. Adobe — Consumers Embrace Generative AI for Trip Planning (U.S. travel sites: +3,500% YoY genAI traffic; engagement and conversion gap trends). business.adobe.com
4. Deloitte — 2025 Holiday Travel Survey (genAI usage trend; category usage and conversion patterns). deloitte.com
5. Phocuswright — From Hype to Habit: 78% of Travelers Say GenAI Improves Trip Planning. phocuswright.com