Most "Pet-Friendly" Hotels Are Dogs-Only. I Built a Site to Say So Out Loud.
Most "pet-friendly" hotel directories list every property that allows a dog and call it a day. Cats get a footnote. The reader is left to call the front desk themselves to find out whether the boutique they just bookmarked actually accepts a cat or quietly turns them away at check-in.
I just spent four days building The Burrow — an editorial guide that does the calling and publishes the answers. Five cities live, fifteen verified picks, thirty-plus ruled-outs with verbatim policy citations.
This is the build log.
The Phase 0 prompt
It started with my friend Midchat. He runs SEO at Moovit and noticed something in his Ahrefs data: pet-friendly accommodation is a 413K-monthly-search market in the US alone, sitting at an average keyword difficulty under 10, dominated by a single pure-play vertical incumbent (BringFido) with a thin product moat — they're a directory.
He sent me a 27MB file. 5,323 keywords, 36,892 keyword-by-URL ranking pairs. I dropped it into DuckDB and ran the analysis.
The data said three things:
- The market is long-tail. 71% of total volume sits at KD ≤ 10. The fight is on the long tail, not the head terms.
- BringFido is the wall, but the wall has cracks. A SerpApi sample of 16 representative queries showed BringFido in the top 3 on 11 of them — entrenched. But on cat-specific queries, the top 3 was a backpack-product blog, PetsWelcome, and Reddit. On property-type queries (pet-friendly resorts, ski lodges), tourism boards were inexplicably winning. The wall isn't load-bearing on the edges.
- AI Overview pressure is real. People-Also-Ask on 16/16 queries. AI Overview on 8/16. Editorial structure with explicit FAQs and verified policy data is what gets cited; directory pages don't.
A clone of BringFido is a 3-5 year fight against an entrenched incumbent. A wedge — cat-first, with property-type and tourism-board cities as flanking moves — is a 6-12 month fight against scattered weak competitors. We took the wedge.
The thesis
Cats are the most under-served segment of the pet-travel niche. Almost every dog-friendly hotel page treats cats as an afterthought when it mentions them at all. Half the time, the published "pet-friendly" policy is dogs-only when you read the actual fine print.
The Burrow's pitch:
Three properties per city, verified end-to-end. We rank, review, and explain. We do not list everything.
Three is the number because three is honest. A list of fifteen hotels is a list. A list of three is a recommendation.
Building it
The stack is built for performance and editorial cadence. Next.js 16 App Router with RSC and full static generation. Tailwind v4 with @theme tokens. System fonts only — zero web font load. Supabase (Tokyo region) for the structured data. MDX for editorial prose, TSX for templates. Cloudflare Workers via @opennextjs/cloudflare. GitHub Actions auto-deploys on push to main.
Schema is six tables: cities, brands, hotels, city_picks, policy_verifications, faqs. Hotel facts in Postgres, voice in markdown. The seam matters: editorial copy stays in version control where it diffs cleanly in PRs, and structured data (fees, weight limits, addresses) stays queryable for cross-page comparison surfaces.
Five cities shipped: Charleston, Asheville, Savannah, Portland, Austin. Each one followed the same workflow:
- Survey 14-18 candidate hotels (every property that claims to allow pets, sourced from BringFido, Booking, brand pet pages, and Google Maps).
- Read the published pet policy on every single one.
- Rule out the dogs-only majority. Capture verbatim quotes for the rejection list.
- Pick three. Write 700-1000 words of editorial prose explaining who each is for and what we ruled out.
- Cross-link policy data into a brand page (Kimpton, Hyatt, Hilton, Marriott) and a comparison table.
The result: 16 substantive pages — five city guides, four brand pages, a cross-brand comparison, an FAQ with FAQPage schema, hub, cities directory, about, author, privacy, terms, custom 404. Sitemap covers everything. Structured data ships Article, Hotel ItemList, FAQPage, BreadcrumbList, Person, Organization, and WebSite. Per-city and per-brand 1200×630 OG images so every shared link previews correctly.
The moat
Three things separate The Burrow from the directory layer above it.
The phone call. For each pick, the editorial methodology calls the front desk and asks four questions:
- Do you accept cats specifically — not just dogs?
- What's the fee, and is that per stay or per night?
- Can a cat be left alone in the room — with a crate, or at all?
- What's the most cats you've had in a single room recently?
The fourth question is the gold one. It separates "policy says yes" from "we actually do this." A hotel whose policy permits cats but whose staff has never hosted one is a worse experience than a Hampton Inn that gets three cats a week. No directory asks. We do.
The published verification block. Every directory says "call ahead." We show the questions and the answers. The pick card has a collapsible section — "What we asked the front desk" — that renders the verifier's name, the date, and verbatim answers pulled from the database. Trust mechanic, surfaced.
The honest ruled-out list. Most pet-friendly roundups quietly skip the properties that turned them down. We publish ours, with reasons. The Vendue in Charleston says it on the record: "We love dogs here, but just dogs." Hyatt Place Charleston: "Sorry, no cats." Hotel Saint Cecilia in Austin: dogs only. Those properties show up in every "best Austin/Charleston pet-friendly hotels" article on the internet, miscategorized. We name names.
The honest caveat
The five cities are currently web-verified — we've read the published policies and have verbatim citations, but we haven't completed the phone pass yet. Every page discloses this on the methodology bar:
This guide was verified from published pet policies. Phone re-verification will follow.
Web-verified data is enough to publish responsibly because the editorial pivot — naming the dogs-only properties out loud, refusing to pad the list — is itself a value-add over directory aggregators. But the phone calls are the moat that compounds. They get done in the next 90 days.
What I learned
A few things crystallized that I'll take into the next build:
Editorial structure is the AI-search moat. Articles with explicit FAQ schemas, verbatim policy quotes, and clear ranked verdicts get cited by ChatGPT and Perplexity. Directory pages don't. The structured data we ship — Article, Hotel ItemList, FAQPage — is the same shape that AI engines extract entities from.
Three-pick editorial beats fifteen-pick directory. A reader doesn't want fifteen hotels. They want the answer. Picking three forces the writer to think about archetypes — flagship, character, value — and forces ruled-outs to do explanatory work the directory format never asks them to do.
The wedge is a real thing. I almost built a BringFido clone. The data told me to, on the surface. Looking deeper at SERP composition revealed where the incumbent was actually weak (cats, property-type, tourism-board cities) and where the editorial format had structural advantages over the directory format. A wedge is not a smaller version of the same thing; it's a different shape of thing in the same market.
What's next
Pending domain bind on theburrow.co. Once that registers and the Worker is bound to it, the next moves are GSC + Bing Webmaster verification (already wired via env vars), Awin and CJ affiliate applications (Booking, Expedia, Hotels.com — the approval cycle is 2-6 weeks), and the phone re-verification pass.
Then a sixth city. The 5-city pattern is repeatable; the next picks come from the same Phase 0 dataset.
If you're curious about the build, the live preview is here. The Hyatt brand page is the one I'd point you at first — two cat-friendly properties, ten ruled-out, with the verbatim "no cats" quotes underneath each. That single page makes the editorial value prop legible in one screen.
