Listen to the podcast episode
Or listen now on Apple Podcasts or Spotify.
Key Takeaways
Don’t build your strategy around ranking in LLM search results: Put referral traffic in a pie chart and ChatGPT is a sliver next to Google’s 80-90% — one client gets ten times more traffic from DuckDuckGo. Those 80-90% LLM growth numbers sound exciting until you realize they get you to 30 or 40 sessions. Nobody can point to a site winning LLM search while failing regular SEO; even if you gamed it, it would revert like old-school SEO tricks did.
Answer the questions guests actually ask instead of chasing the algorithm: The durable play is content that answers what travelers and owners type into these tools — ask the LLMs what questions people ask about lodging in your area and build real answers. The tangible LLM checklist is short and mostly overlaps SEO best practice anyway: don’t block the OpenAI crawler, keep clean heading structure, and name your location outright — in North Myrtle Beach, South Carolina — instead of a vague come visit us.
Use AI for first drafts and fact-checks, never publish-and-walk-away: ChatGPT is the best junior employee Conrad has ever had — twenty bucks a month, never sick, never takes a day off. Write your top-ten restaurant list, then ask what it would add; paste a finished post and ask what it would correct. Writer’s block shouldn’t exist anymore. But raw output gets a human edit before it touches your website.
Twelve great posts beat 240 garbage ones: AI makes volume cheap, and that’s the trap: Google notices when a site that never published suddenly floods itself with thin content, and rankings go backwards. Three paragraphs is not a blog post. Dress the good ones properly too — an image under each heading, internal links, a personal tip like the most-mentioned dish from the reviews — and the same text performs far better.
Let AI chew through data that would take you hours: Dump 2,000 Performance Max search terms into Gemini with your asset group’s theme and ask for negative keyword suggestions grouped by root term — a two-hour review becomes about a minute. Feed guest emails into NotebookLM, which only uses what you upload, and ask which local recommendations you give most. It’s also a solid educated guess for things like picking origin states for a brand-new ad account.
Structure content so humans and machines both find the answer fast: You get two to three minutes of attention. Proper H1, H2, and H3 tags, bullets where they help, key information surfaced instead of buried in prose. Then go deeper than the surface-level results Google itself serves — the where-to-park tip and the best-dish detail are what AI can’t fake and what real trip-planners remember you for.
Run plagiarism and fact checks on anything AI wrote: You can’t see where an LLM pulled from, and sometimes it’s your local competitor’s blog — Paul has watched it happen in the Destin-30A market. Free plagiarism checkers exist; use one before publishing any big AI-generated chunk. ChatGPT’s own disclaimer says it makes mistakes — believe it.
Prompt quality sets output quality — a D-minus prompt earns an F response: A generic write-me-a-fall-guide prompt returns milquetoast filler any competitor could generate. Add your own outline and bullet points, use a purpose-built writing tool like Cuppa, and the draft jumps to a B-plus. Spend 30-40 minutes on the outline to save two hours of writing, and ask for longer output than you need — trimming back beats padding out.
Keep a human hand on the wheel: Set reminders to audit what AI built — like rechecking state-level click data two weeks after Gemini picked your geo targets. Some functions deserve humans, period; guests and owners can tell the difference. There is no one-button solution, and the top companies in your market prove it — if anything, you need more people managing AI outputs, not fewer.
What We Cover In This Episode
Conrad and Paul lay out the do’s and don’ts of using AI in vacation rental marketing: why chasing LLM rankings is a trap, where AI genuinely earns its keep — first drafts, fact-checks, bulk data analysis — and the checks that keep it from hurting your SEO.
