How We Grew a Global Hospitality Brand's AI Search Audience by 450% in a Single Quarter
In one quarter, a global serviced residence and extended-stay hospitality client went from being a quiet challenger inside ChatGPT, Gemini, Google AI Mode, and Perplexity to a brand that AI engines actively recommend. Their generative search audience grew from 7.4 million to 34.6 million, a 450% jump. Mentions across AI platforms nearly doubled. Favorable sentiment climbed to 77%, sitting more than 20 points above industry peers. This is the story of how a structured Generative Engine Optimization (GEO) program rewrote the brand's visibility inside the AI layer that increasingly sits between travelers and their booking decisions.
The Challenge: Visibility in a Channel That Didn't Exist Two Years Ago
For decades, hospitality marketing has been a battle between organic search, paid search, and OTAs (online travel agencies). That fight is being reshaped. Travelers now ask conversational questions like "Where should I stay in Tokyo for a month with my family?" or "Which hotels in Jakarta have full kitchens for extended stays?" The answer they get comes from an AI engine that synthesizes thousands of sources and recommends a handful of brands.
When we onboarded this client, their position inside that emerging layer was modest. Compared with the largest global hotel groups, who held between 8% and 12% share of voice in AI generated responses, our client held 4%. The market itself was highly fragmented: 57% of generative responses cited "other" entities, generic mentions, listicles, and review aggregators rather than any named brand. There was room to move, but only for a brand willing to engineer for how AI engines actually retrieve and rank information.
Three problems were stacked on top of each other:
The brand's owned content was indexed for traditional SEO, not for generative retrieval. AI engines look for clearly structured answers, authoritative entities, and dense topical clusters. Most property and destination pages were written for human readers and Google's classic ranking signals, neither of which translates cleanly into a generative response.
Competitors with deeper pockets were already showing up in the same AI conversations. Larger industry peers had a head start on share of voice simply because their brand names were referenced more often across the open web.
And finally, the brand had a genuine differentiation story, apartment style living, sustainability commitments, extended stay expertise, that wasn't being surfaced by AI engines in the categories where it mattered most.
The Approach: A GEO Methodology Built for the AI Layer
We treat GEO as a distinct discipline, not a rebrand of SEO. Search engines rank documents. Generative engines synthesize answers. The work required to win the second is different from the work required to win the first.
Our methodology rested on four pillars:
Property optimization at scale. Each property page is a discrete entity in the AI knowledge graph. We rewrote 162 of them with the structural elements generative engines prioritize: clearly named amenities, hyper local neighborhood context, structured FAQs that act as direct "answer nodes" for conversational queries, and consistent entity references that make it easy for an AI to associate the property with the parent brand.
Prompt driven content creation. Instead of starting from a keyword list, we started from a prompt list. We mapped 21 specific traveler questions that AI engines were either answering generically or not answering at all, queries about sustainable travel, holiday planning, culinary destinations, digital nomad living, family stays, then built content designed to be the authoritative response. The goal isn't to rank for the question; it's to be the citation the AI engine reaches for when generating the answer.
Topical clustering for AI trust. Generative engines reward density. When a brand publishes ten well structured articles about extended stay living in Tokyo, AI models begin to associate the brand with that topic at a level deeper than any single page can achieve. We built 40 GEO specific articles across two layers: unbranded authority pieces ("Sustainable travel tips for eco conscious tourists," "Best child friendly hotels in London") to win top of funnel, and branded pieces reinforcing the brand's specific value propositions and loyalty program to convert mid funnel intent.
External authority signals. AI engines, particularly Perplexity and Gemini, weight third party citations heavily. We launched a targeted media outreach program to secure high authority backlinks across business, travel, and news outlets, with a goal of 20+ placements in reputable publications. These act as "votes of confidence" that compound the on page work.
Inside the Playbook: What 90 Days of Execution Looked Like
The work breaks down cleanly because GEO rewards consistency more than cleverness.
On the technical side, every optimized property page received a structured FAQ block targeting the specific questions that high intent travelers ask before booking. Where AI engines used to hallucinate or default to generic chain hotels, they now had a clean, structured source to cite.
On the content side, the 21 prompt driven articles were sequenced against both seasonal demand cycles (holiday travel, year end planning) and evergreen topics that generate steady AI traffic year round (wellness, business productivity, digital nomad living). This created a content base that doesn't just spike, it accumulates.
On the off page side, the media program targeted publications with strong domain authority and broad topical relevance, ensuring that when an AI engine looks for third party validation of the brand, it finds it. Coverage in major business and regional news outlets compounded the on page entity work and accelerated the brand's appearance in real time AI responses.
The Results: Visibility, Sentiment, and Share of Voice
Four months in, the numbers told a clear story.
Audience reach grew 450%. From a starting point of 7,462,248 in November 2025, the brand's AI search audience climbed to 34,644,410 by February 2026. This is the audience exposed to the brand inside AI generated responses across ChatGPT, Gemini, Google AI Overview, Google AI Mode, and Perplexity. The growth wasn't linear, it accelerated through January and February as the optimized pages and content clusters reached the depth needed for AI engines to treat them as authoritative.
4,300 cited pages. AI engines cited the client's owned properties and content across 4,300+ unique URLs. The top performing single page, a non branded article on essential amenities for extended stay guests, was cited in 48 distinct prompts. This is a critical insight: generative engines reach for high utility, non branded content first, then traverse to specific booking pages. Brands that publish only product pages miss the upstream citation.
Top tier visibility on priority topics. Across the brand's top ten performing topic and property clusters, visibility scores ranged from 94 to 99 out of 100. The highest performing geographic cluster was Tokyo, where 50% of top ten topics were concentrated, an AI moat that's now structurally difficult for competitors to displace. The single highest volume property cluster (a flagship Jakarta location) attracted AI volume of 13,600, the brand's most valuable generative asset.
Favorable sentiment at 77%, with a 6 point quarterly gain. AI engines don't just count mentions; they weight them by sentiment. The brand sustained 77% favorable sentiment across the reporting period, with a 6% increase quarter over quarter. The closest industry peer sat at 55%. Several peers ranged between 45% and 52%. Inside the categories AI engines use to surface travel recommendations, sentiment is a multiplier on mention volume.
Category leadership across eight of nine feature categories. AI engines classify hospitality brands across feature categories like extended stay, sustainability, technology, loyalty, and corporate travel. Our client led every feature category against the comparison set with one exception (global scale, where the largest industry peer holds an inherent footprint advantage). The strongest categories were Loyalty, Rewards, and Value (133 mentions), Extended Stay Apartment Living (101 mentions), and Sustainability, Wellness, and Safety (71 mentions, more than 10x the closest peer).
Why This Matters for Brands Operating Inside AI Search
A few themes ran through the data that we think generalize beyond hospitality.
First, generative engines prioritize utility content over brand content. The single most cited page in this engagement was a non branded amenities guide. Brands that publish only product or booking pages are invisible at the top of the AI funnel, no matter how strong their domain authority looks in traditional SEO tools.
Second, localization compounds. Pages addressing real traveler friction (how to use a specific payment method in China, how to navigate a public transit system in Australia) generated outsized AI citations. AI engines reward content that solves concrete problems for specific user contexts.
Third, sentiment is a competitive moat. Larger competitors with bigger backlink profiles and bigger PR budgets still lost ground in AI responses when their sentiment scores trailed. AI engines have learned to favor brands users actually like, which means guest experience and the digital narrative around it now feed directly into AI search performance.
Fourth, GEO and traditional SEO are complementary, not competitive. The optimized property pages still perform in classic search; the GEO work added a second retrieval surface on top. The brands that treat these as one program, not two, get the compounding benefit.
What's Next
The next quarter of work focuses on three priorities: defending the AI moat we've built in the highest performing geographic clusters before competitors catch on, expanding the proven playbook into the brand's secondary sub brands in less saturated markets, and closing the small remaining gaps in categories like multi brand booking and flexible cancellation where larger industry peers still hold a narrative advantage.
We treat GEO as an ongoing program because the underlying AI engines aren't static. Each model release reshapes the ranking surface. Brands that build a continuous program, not a one off project, are the ones that hold their ground.
Working With Search Agency
We help brands win the AI search audience the same way we've always helped them win Google: by understanding how the ranking system actually works, engineering toward it at scale, and measuring the right outcomes. If you'd like to see what a GEO program could do for your brand, we'd love to talk.
