Skip to content

Case Study

Lake Rockets to #1 in AI Visibility, Overtaking Airbnb

Lake needed to show up where travelers were already asking for vacation advice: ChatGPT, Perplexity, and AI Overviews. Growth Marshal rebuilt the retrieval surface so Lake could be cited as an answer, not just ranked as a result.

Customer
Lake
Engagement
Generative Engine Optimization
Window
3 weeks
Read time
3 minutes
Citation shareDay 21
Lake47.0%
Airbnb41.9%
Vrbo28.7%
Expedia12.4%
Unbranded lake-rental queriesSource: Profound
/0147%

AI answer visibility

Of unbranded lake-rental prompts returned Lake by day 21.

/02+33.5pt

Citation-share gain

A decisive move from near-invisible to category leader.

/0315x

Answer-share lift

U.S. AI answer share rose from 0.9% to 13.5%.

/0421d

Time to signal

No paid amplification. Just cleaner retrieval surfaces.

Result headline

Lake became the category leader inside AI search, reaching 47% citation share in three weeks.

/01

Travelers shifted from search boxes to answer engines.

Travelers stopped searching. They started asking.

Lake.com is a vacation-rental marketplace built around life by the water: cabins, cottages, and lake houses across North America. By early 2025, the discovery pattern had started to move. Trip-planning questions were showing up inside ChatGPT, Perplexity, and AI Overviews before buyers ever reached a results page.

The category leaders were already present in those answers. Lake was not. Ranking on a SERP mattered less when the traveler never saw the SERP.

We needed to be inside the AI answers themselves.
David CiccarelliCEO and Founder, Lake
/02

Retrieval-ready content, schema, entities, and crawler paths.

Lake managed content. Growth Marshal engineered the retrieval surface.

We mapped the conversational micro-moments travelers use when they ask for lake-house recommendations, family trip ideas, seasonal travel guidance, and comparisons between rental platforms.

Then we rebuilt the pages around answerability: tighter narratives, embedded FAQs, cleaner internal structure, entity reinforcement, JSON-LD, and an llms.txt endpoint that gave AI crawlers a direct map to Lake's strongest assets.

geo_pipeline.sh

$ map intents --engines=gpt,claude,perplexity

-> 412 micro-moments cataloged

$ benchmark --domain=lake.com

-> baseline: 0.9% answer share

$ rewrite --schema=jsonld --surface=lake

entities linked

llms.txt deployed

D+21 -> 47.0% citation share

Pipeline

From keyword ranking to retrieval as an answer source.

The work made Lake easier for models to parse, classify, cite, and recommend when a traveler asked for lake-specific vacation guidance.

/01 Intent

Conversational micro-moment mapping

Catalog the prompts buyers ask inside LLMs, then benchmark Lake against Airbnb and Vrbo across each answer surface.

/02 Narrative

Answer-shaped content

Rework destination, drive-to, seasonal, family, fishing, and event pages into dense answer candidates.

/03 Graph

Entity resolution and JSON-LD

Make Lake, its inventory, its differentiators, and its market position legible to retrieval pipelines.

/04 Crawl

llms.txt and bot invitations

Give GPTBot, ClaudeBot, and PerplexityBot a structured path into the pages that matter.

/03

Citation share, answer share, and thematic leadership moved.

Lake became the category leader inside AI search.

Lake's citation rate climbed to 47%, topping Airbnb at 41.9% and Vrbo at 28.7%. U.S. answer share grew from 0.9% to 13.5% in three weeks.

For unbranded lake-house searches, Lake started appearing in roughly half of AI-generated answers. The shift expanded beyond a single query class into brand, pricing and fees, comparisons, and integrations.

Citation share

Lake47.0%
Airbnb41.9%
Vrbo28.7%

Answer share climb

Answer share rose from 0.9% to 13.5%

0.9% to 13.5% in 21 days

Thematic lanes

Brand52%
Pricing and fees49%
Comparisons46%
Integrations41%
/04

The answer engines needed cleaner evidence, not louder claims.

The winning move was making Lake easier to retrieve, trust, and cite.

AI visibility did not come from stuffing pages with more keywords. It came from reducing ambiguity. The content explained what Lake is, where it is strongest, what it offers that large rental marketplaces do not, and how those facts connect to buyer questions.

Once the evidence became structured, specific, and crawlable, answer engines could place Lake in the category with confidence.

Ready when you are

Make your business answer-engine ready.

Build the structure, evidence, and automation foundation needed to compete in an AI-first market.