This insight summarizes the September 2026 PhocusWire opinion piece "AI search is changing the marketing playbook: How can travel survive?" by Mario Gavira, CMO of Travelier.
AI search is not killing marketing — it's rewriting where the work happens. Traditional SEO traffic will gradually decline, but branded and high-intent keyword searches are still the best bet to capture ready-to-convert users today. The bigger shift is how AI models decide who to recommend, and travel brands that keep spending only on their own website will lose share to those investing in third-party footprint.
The pivotal data point: between 51% and 85% of what AI cites about brands is third-party content. AI models trust community consensus far more than a brand's own marketing copy. If external sites — publications, review platforms, forums, industry directories — don't talk about you, AI has no signal to work with, regardless of how polished your own pages are.
Three principles for travel marketers navigating AI search:
- Double down on branded and high-intent keyword capture. Traffic on informational queries will keep falling as AI answers them directly. Users who type your brand name, or a booking-intent query, are still worth every dollar — and easier to convert.
- Invest in offsite presence as core marketing infrastructure. Reviews, PR mentions, partnership content, industry citations, community discussion — these are now the raw material AI ranks on. Treat offsite footprint the way SEO teams treated backlinks a decade ago.
- Optimize for machine comprehension, not just human navigation. Structured content, consistent data across platforms, and clear factual density help AI extract and cite you accurately. Marketing copy written to persuade humans often fails machines.
What to do now. (1) Audit how often your brand is mentioned versus cited in AI answers — mentions without citations signal an accessibility or trust problem. (2) Reallocate a share of paid budget from broad SEO toward PR, review-generation, and third-party content programs. (3) Add schema markup and structured data as table stakes, then measure presence across AI engines quarterly. (4) Stop measuring success only in click-through rate — AI can recommend you without a click.