Earned Media Drives 84% of AI Citations; llms.txt Adoption Stays Thin
New data confirms AI engines pull overwhelmingly from earned media, not brand-owned pages, while fewer than 9 in 100 top sites have published llms.txt. For businesses chasing AI visibility, the content distribution strategy matters more than the technical signal.
This Week's Signals
84% of AI Citations Come from Earned Media; Your Own Site Accounts for 5-10%
Why it matters for your score
If AI engines are drawing on earned media for the vast majority of citations and brand-owned pages for only 5-10%, your content presence in third-party publications is a more direct path to AI visibility than site-side technical tweaks. An Ahrefs study found no statistically significant citation lift from adding JSON-LD schema, which means schema alone should not be the centerpiece of your AEO strategy. Listicles and well-structured articles on credible external outlets are what the data points to.
Only 8.7% of Top 1,000 Sites Publish llms.txt as of June 2026
Why it matters for your score
Across a fixed sample of 1,000 top-ranked sites, just 87 publish llms.txt, a rate that rises to 15.8% among the 549 reachable domains in the crawl. The file format is not yet a standard practice even among high-authority domains, so publishing one puts you ahead of the field without being table stakes. Some sites publish only llms.txt, others add llms-full.txt, and a few publish both, suggesting there is still no settled convention on which to prioritize.
AI Overviews Appeared on 25.11% of Queries in Q1 2026, Peaking at 47% in January
Why it matters for your score
Conductor's benchmark across 10 industries and 21.9 million Q1 2026 queries found AI Overviews at 25.11% query coverage with average market coverage of 34.5%, peaking at 47% in January. The report frames the shift plainly: AI has replaced your website as the first touchpoint, not search itself. Industry-level benchmarks now exist, so generic AEO advice is no longer sufficient and vertical-specific positioning is the next step.
What to do this week
- 1
Audit your off-site content footprint this week: if earned media drives 84% of AI citations and your own site accounts for only 5-10%, map which third-party publications in your vertical are already being cited by AI engines, then prioritize pitching structured, list-format content to those outlets rather than adding more schema to pages that the data shows have no proven citation lift.
- 2
Publish or update your llms.txt file using Faro's /tools/llms-txt tool — with adoption sitting at 8.7% across the top 1,000 sites, doing it now puts you ahead of roughly 91% of high-authority domains before the format becomes a baseline expectation.
- 3
Run your site through Faro's /tools/ai-readiness-scan to see how your current structure and content hold up against the verticals in Conductor's benchmark, which found AI Overview coverage averaging 34.5% in Q1 2026 — knowing your vertical's baseline is the starting point for any credible gap analysis.