AI referral traffic is growing, but the sector benchmark matters

Compare growth multipliers, not just headline percentages

Adobe's August 2026 report gives different year-on-year growth rates for AI referrals to U.S. travel and retail sites. The opening calculation converts each rate to an index relative to its own prior-year base before comparing them.

The ratio does not tell us that travel received more visits than retail. A smaller channel can grow faster while remaining smaller. It also does not say how much of either stream came from ChatGPT rather than other identified AI sources.

Build the comparable view for your site

Use human sessions attributed to a defined set of AI referrers. Keep that list fixed or document changes to it. Choose the same year-on-year window, then show the earlier and later session counts alongside the growth rate.

For a small base, display the counts prominently. Moving from 10 visits to 30 is 200% growth but still only 20 additional visits. This is illustrative arithmetic, not an observed brand result.

Do not substitute a crawler chart when human analytics is unavailable. OpenAI describes ChatGPT-User as a user-triggered agent for certain actions; its requests are not a count of all customers clicking through from ChatGPT.

Add quality and completed outcomes

Review landing-page fit, qualified actions and completed orders. Compare the same products or services where possible. A rise in traffic can reflect a wider, less-qualified audience; a smaller stream can still generate useful enquiries.

Keep retail conversion evidence separate from travel conversion evidence. Sector aggregates are useful context, not a promise that your own site should achieve the same conversion rate.

Do not turn a growth observation into a causal case study

A content update followed by referral growth does not by itself establish that the update caused it. Record other campaigns, platform changes and seasonal effects. Use a credible comparison or experiment before reporting incremental revenue.

This revision removes the previous page's unsupported anonymous brand-uplift examples. The reported external figures now link to their source, and the new ratio is labelled as Aiso's calculation rather than an Aiso client result.

Use the measurement guide to separate agent activity from human referrals, then inspect the queries and source pages behind the answers before choosing a content task.