| Research overview | The conclusions and the key data, shortened |
|---|---|
| Blog post with the full details | New AI Search Visibility & LLM Ranking Factors Report Data (July 2026) |
| Live report | AI Search Visibility report |
What the research found
For the July 2026 report, I measured 18 signals against the domains AI models recommend in 100 industries, using about ten times the data of the May 2026 report. Across all industries, search engine outbound links had the strongest correlation. But even that signal explains only about 11% of the variation. The leading signals change by industry, and the recommendations change based on who is asking. In airlines, for example, five buyer personas produced four different top recommendations.
Findings you can quote
- Across all industries, search engine outbound links have the strongest correlation at 0.331.
- Homepage keyword relevance was the weakest signal in May (0.072). In July, it was the second strongest (0.204).
- Since May, the correlation for Reddit comment mentions rose from 0.111 to 0.148. For Wikipedia presence, it fell from 0.077 to 0.055.
- In airlines, Reddit comments (0.551) and search engine outbound links (0.552) are essentially tied for the strongest signal. Both Reddit signals have stronger correlations than every backlink measure.
- For airlines, delta.com is the most recommended domain. Google.com ranks fourth because the models often suggest searching Google Flights.
- No domain leads for every persona. Delta leads for two. Southwest, United and studentuniverse.com each lead for one.
Quotes from Ben Wills
These quotes are from the published report post.
“A single aggregate score isn’t a meaningful thing to monitor. If you’re tracking AI visibility without specifying who’s asking, you’re measuring noise.”
“Search isn’t dead. It’s one layer of AI recommendations.”
“Treat this report as a map of what to test, not a guarantee of what to fix.”
Story angles by audience
AI search visibility
You need to know who is asking before an AI visibility score means much. In the report, I propose measuring persona coverage: how many distinct buyer contexts lead to a recommendation for your brand at all. One score for the whole brand hides those differences.
SEO
Search visibility and AI recommendations have a consistent relationship in this data. Models also search the web while answering. So authority, relevance, crawlability, citations across sources and useful content still contribute to what the model has to work with. You still need to measure AI visibility separately from search rankings.
Public relations
Look at the searches the models run. They name specific third-party authorities far more often than the brand’s own site. In airlines, those include Skytrax, OAG, The Points Guy, NerdWallet, J.D. Power and AARP. How you rank on the review sites and industry lists the model searches matters as much as your own homepage.
Reputation monitoring
The correlations for Reddit and Wikidata both grew stronger since May. In airlines, Reddit discussion is essentially tied with search citation as the strongest predictor. That gives reputation teams something specific to investigate in their own industry.
Competitive analysis
The same airline can look very different across four channels: direct model recommendations, the searches models run while answering, Google AI Overview citations and organic search. Alaska Air, for example, is at the 96th percentile for direct recommendation and the 26th for organic search. Measuring one of these won’t reliably tell you how a competitor is doing in the others.
What to keep in mind
- These figures are correlations. None proves why a model recommends a domain. Most of what determines the recommendation happens inside the model.
- The changes since May also reflect changes in the research. The dataset grew about ten times, and I re-scored two signals: a phrase now counts 3 times in a title, 2 times in a meta description and once in the body.
- May covered 145 industries; July covers 100. I removed 45 industries without adding replacements.
- The blog post walks through the numbers across all industries, then uses airlines as an example. The report includes the full breakdown for all 100 industries.
Interviews
For an interview with Ben Wills, get in touch. You can find his background and the other research in the media kit.