Not sure where to start? Take a guided tour.
AI Site Grader
AI Query Scanner
llms.txt GeneratorDoes AI rely more on reference sites than the open web?
The index cannot support that claim. Reference sources account for 1.48%, while 88.40% remains unclassified in “other”. The taxonomy is too partial to compare reference sites with the open web as a whole.
The key numbers.
- Reference category share
- Partial hand-maintained source classification.
- 1.48%
- Unclassified share
- Domains outside the current named source map remain in “other”.
- 88.40%
- Citation appearances
- 80,769,585 appearances in the verified October 14, 2026 snapshot.
- 80.8M
Reference sites are visible in the named taxonomy, but classification coverage is the binding limit on broader source-type conclusions.
The evidence.
Source
AI Citation Index
| Measure | Result |
|---|---|
| Reference category share Partial hand-maintained source classification. | 1.48% |
| Unclassified share Domains outside the current named source map remain in “other”. | 88.40% |
| Citation appearances 80,769,585 appearances in the verified October 14, 2026 snapshot. | 80.8M |
Population and method
Built from 80,769,585 citation appearances across 725,777 unique domains in the latest valid snapshot for 1,928 tracked brands, observed from October 3, 2025 to October 14, 2026.
Where the evidence stops
The index is a changing panel of tracked brands and prompts, not a random sample of all AI responses.
Domain share is observational. It does not establish causal influence or reveal the full retrieval set.
The hand-maintained source taxonomy is partial: 88.40% of citation appearances remain in “other”.
Model attribution, response-level citation denominators, geography, raw prompts, and raw responses are not retained in the public aggregate.
What to do with this
- Audit the domains that already dominate your category and decide which ones are realistic citation targets.
- Build source-worthy pages for the query shapes where AI is already citing third parties.
- Treat reference, review, and social sources as part of your AI visibility surface, not as noise.
This answer matters because it turns a study finding into an operating rule teams can use when they decide what to publish, refresh, or measure next.
Questions about this answer.
Cite this answer.
Source study: Citation Sources“The index cannot support that claim. Reference sources account for 1.48%, while 88.40% remains unclassified in “other”. The taxonomy is too partial to compare reference sites with the open web as a whole.”
Mack Grenfell · Does AI rely more on reference sites than the open web?
AI Citation Index · Trakkr Research · Updated 14 October 2026