Google AI Overview
youtube.com
6.2%
google.com
5.2%
reddit.com
1.9%
facebook.com
0.9%
linkedin.com
0.6%
instagram.com
0.3%
gartner.com
0.2%
forbes.com
0.2%
g2.com
0.1%
AI Visibility Stats
AI Visibility Stats: Sep 7, 2026 – Sep 13, 2026
Benchmarks based on 348,539 citations analyzed across major AI models.
Most Cited Domains
Global platforms that make up at least 0.1% of each AI model's citations.
AI Mode
google.com
31.5%
youtube.com
8.3%
reddit.com
0.7%
facebook.com
0.4%
linkedin.com
0.3%
gartner.com
0.2%
instagram.com
0.1%
forbes.com
0.1%
Claude
g2.com
0.6%
gartner.com
0.5%
linkedin.com
0.4%
Gemini
google.com
9.6%
reddit.com
0.7%
youtube.com
0.5%
forbes.com
0.3%
gartner.com
0.3%
techradar.com
0.1%
ChatGPT
google.com
0.7%
techradar.com
0.3%
reddit.com
0.2%
g2.com
0.2%
gartner.com
0.2%
linkedin.com
0.1%
forbes.com
0.1%
Perplexity
reddit.com
1.0%
linkedin.com
0.6%
forbes.com
0.3%
g2.com
0.1%
youtube.com
0.1%
techradar.com
0.1%
Citation Properties
Content types, formats, length, tone, and structural signals found in pages that AI models cite.
What structural and content signals appear in pages AI models cite?
54.8%
Have opinions
10.4%
Have an FAQ section
48.7%
Have visual data
83.5%
Use bullet points
15.7%
Original research
66.1%
Author credentials shown
60.9%
Empirical evidence
115
Sources analyzed
Content length of analyzed citation sources
4,896
Avg length (chars)
1,494
Median length (chars)
66.8
Avg paragraphs
What types of websites do AI models cite?
corporate website31.3%
academic journal21.7%
government website16.5%
non-profit organization12.2%
news outlet9.6%
blog7.8%
encyclopedia0.9%
What content formats do AI models prefer to cite?
article73.9%
research paper13%
listicle8.7%
multimedia3.5%
homepage0.9%
What tone of voice do cited sources use?
informative68.7%
formal21.7%
persuasive5.2%
technical2.6%
conversational1.7%
What types of content do AI models cite?
Homepage24.39%
Blog18.45%
Guide/Tutorial14.92%
Article7.75%
Product page7.38%
Marketplace listing5.12%
Wiki/Knowledge base4.86%
Listicle4.06%
Other13%
LLM Preferences
How each AI model differs in citation density, coverage rates, and brands mentioned per response.
Average citations per response (among cited responses), by LLM
| LLM | Avg citations / response |
|---|---|
| Perplexity | 9.76 |
| Claude | 7.8 |
| Google AI Overview | 7.29 |
| ChatGPT | 6.54 |
| AI Mode | 6.06 |
| Gemini | 3.97 |
Share of responses with at least one citation, by LLM
| LLM | % responses with a citation |
|---|---|
| ChatGPT | 98.5% |
| Perplexity | 96.9% |
| Google AI Overview | 88.1% |
| AI Mode | 84.8% |
| Gemini | 76.7% |
| Claude | 52.6% |
Average number of brands named per response, by LLM
| LLM | Avg brands / response | Median |
|---|---|---|
| Claude | 5.27 | 5 |
| Google AI Overview | 5.05 | 5 |
| Gemini | 4.91 | 5 |
| ChatGPT | 4.9 | 5 |
| AI Mode | 4.66 | 5 |
| Perplexity | 3.89 | 4 |
Distribution of number of brands named per response
| Brands named | % of responses |
|---|---|
| 0 | 17.2% |
| 1-3 | 19.8% |
| 4-9 | 54.1% |
| 10+ | 8.9% |
Notes on this report
Reflects Spotlight's own prompt-crawl schedule/mix across customer brands, not the LLMs' real-world usage share. Do not present as market share.