You’re reading an issue of “The AI Economy,” my newsletter exploring the forces shaping the AI era—tracking how AI is rewriting business, work, technology, and culture. Subscribe to get expert insights and curated updates delivered straight to your inbox.
When Jesse Middleton, a venture capitalist at Flybridge Capital, was looking to arrange travel from New York City to Cancun to attend a wedding, he didn’t scour Expedia, Travelocity, Kayak, or even individual airline websites—at least not on his own. Rather, it was handled by an AI agent. He wrote on LinkedIn that it searched flights at every NYC airport, found unused JetBlue credits for him, and made the reservation for his family, even saving him money.
Scenarios like Middleton’s are becoming more common. New research from Adobe shows AI traffic to travel sites is still climbing—it grew 119 percent in July 2026 over the same month last year. It’s a smaller increase than in May 2026, when AI traffic grew 194 percent year over year. Travelers are using large language models mainly for research and inspiration, cited by 43 percent of respondents each, followed by transportation planning at 41 percent and itinerary creation at 34 percent. Eighty-four percent said AI improved the travel experience.
And the conversion rate gap between AI and non-AI traffic continues to shrink. In May, AI traffic converted 28 percent below non-AI sources. By July, the gap was down to one percent. A visit from an LLM is now worth roughly as much as a visit from any other source, such as search results or emails.
However, there is one significant catch: Adobe said “significant portions” of U.S. travel and retail sites still cannot be read by machines. This limits a brand’s visibility across AI search results, whether it’s in Microsoft Copilot, OpenAI’s ChatGPT, Anthropic’s Claude, Perplexity, or Google’s Gemini chatbot.
To create a benchmark, Adobe analyzed webpages for U.S.-based hotels, cruise lines, car rental services, and airlines using the AI Content Visibility Checker diagnostic tool it launched in 2025. Each site was evaluated to determine how much of its content an LLM could read, with scores ranging from zero percent to 100 percent—anything below 50 percent means at least half of its content cannot be read by machines. Adobe reported that hotel homepages had an average score of 71 percent, cruise lines 70 percent, car rental services 64 percent, and airlines 42 percent.
Digging deeper, the data showed that product pages fared better, though with one exception. Cruise line pages showing itineraries and ship options scored 76 percent, followed by car rental vehicle option pages at 73 percent, and airline pages showing routes and fare prices at 59 percent. Hotels were the only category scoring lower on product pages, at 68 percent, than on the homepage.
Organizations have spent years investing in SEO to land among the ten blue links. But 95 percent of people who use AI tools now tell Adobe they trust those outputs as much as traditional search results—and Adobe argued that confidence is translating into bookings on brand sites. If travelers trust the answer, they have less reason to click through and check it. Content a model can’t read doesn’t get corrected downstream. It just doesn’t get surfaced.
U.S. retail sites are also seeing the impact of LLMs. AI referral traffic was up 62 percent in July over the same month last year, and it converted 60 percent better than non-AI traffic—the 11th straight month AI has come out ahead. Adobe’s data showed that when shoppers are referred by AI, they’re 14 percent more likely to engage, spend 59 percent more time on the site, and are 33 percent less likely to leave. These shoppers have a 28 percent greater chance of adding items to their shopping cart.
Yet, just like with the travel industry, retailers have an AI visibility problem. Adobe’s analysis found that homepage visibility scores averaged 61 percent, indicating that nearly 40 percent of homepage content was unreadable by machines.
So which retail categories had the highest score spanning homepages, product pages, corporate blogs, and more, according to Adobe’s AI Content Visibility Checker?
- Apparel: 76 percent
- Electronics: 70 percent
- Cosmetics: 68 percent
- Sporting Goods: 67 percent
- Furniture and Home: 64 percent
- General Merchandise: 63 percent
- Grocery: 59 percent
Adobe’s research merely provides a snapshot of how American consumers use AI to plan travel and shop. What it doesn’t provide is guidance on what brands must do to improve their AEO/GEO standing. Adobe markets visibility tools, so it hopes customers will take advantage of offerings such as its LLM Optimizer to monitor AI traffic and improve discoverability. It also has Enterprise AIO, tailored for organizations, that’s powered by Adobe’s $1.9 billion acquisition of SEO platform Semrush.
