This insight summarises AI Can Find Your Hotel. It Just Won't Recommend It. by Markus Busch, published on Hospitality Net in June 2026.
Hotels won the structured-data race — and it didn't move the needle. Hotels now sit at 73% machine readability, ahead of car rentals, cruises, and airlines. Yet structured data investments don't determine which eligible property the assistant actually recommends. The recommendation signal lives in a different layer entirely.
What an analysis of 730,000 AI responses showed. Ahrefs analysed 730,000 AI responses across ChatGPT and Google AI surfaces and ranked the correlations with brand visibility:
- YouTube mentions: 0.737 (strongest signal)
- Branded web mentions: 0.664–0.71
- Backlinks: just 0.27 (weak)
The lesson: earned presence on external platforms outweighs owned-site optimisation. AI systems recommend brands that the broader internet talks about — video, editorial, organic mentions — not the brands with the cleanest schema markup.
Two other numbers worth holding.
- AI-referred visitors convert 28% less than non-AI sources, despite spending 70% longer on site. The intent is shallower because the AI citation was thin.
- Consumer trust in AI recommendations dropped from 82% to 54% in one year — travellers are increasingly cross-verifying before they book.
The independent hotel problem. Smaller properties don't have the distributed brand footprint — video content, third-party editorial, organic mention volume — that AI systems reward. Technical infrastructure won't close that gap on its own.
Where to actually spend:
- Treat structured data as table stakes — get to 100%, then stop.
- Build a real earned-media programme: video reviews, creator partnerships, editorial coverage on the publications AI models actually train on.
- Manage reputation across the platforms AI systems cite — TripAdvisor, Google, Reddit, YouTube — not just your own site.
- Plan for the conversion gap: tighten landing pages, retargeting, and proof points for AI-referred traffic that arrives less qualified.