Open data
AI Visibility Industry Benchmark
Aggregated statistics from real Geovory scans: how often AI assistants mention businesses when buyers ask, which domains they cite, and how that differs by industry. Only measured data — no estimates.
Brand mention rate by industry
Share of AI answers that mentioned the scanned business, per industry.
Industries marked “limited sample” have too little data for firm conclusions; treat them as directional.
Domains AI cites most
The domains AI answers cited most often across all scans.
| Domain | Type | Citations | Share |
|---|---|---|---|
| yelp.com | Directory | 1526 | 6.1% |
| google.com | Directory | 1518 | 6.0% |
| dianping.com | Directory | 1145 | 4.6% |
| meituan.com | Other (incl. official sites) | 595 | 2.4% |
| bbb.org | Directory | 468 | 1.9% |
| map.baidu.com | Other (incl. official sites) | 441 | 1.8% |
| capterra.com | Review platform | 416 | 1.7% |
| g2.com | Review platform | 389 | 1.5% |
| homeadvisor.com | Directory | 377 | 1.5% |
| cloud.google.com | Other (incl. official sites) | 348 | 1.4% |
Citation source types
What kind of sources AI answers cite: directories, review platforms, social media, media, and everything else (incl. official sites).
Industry deep dive
Per-industry sample size, mention rate, and the top cited domains.
AI Software Subscription Store
Restaurants
宠物医院
Plumbers
Auto Repair Shops
Coffee Roasters
Other industries
⚠ Limited sampleClínicas Odontológicas
Law Firms
Dentists
Landscaping Companies
⚠ Limited sampleMethodology
Sampling
Every data point comes from real Geovory scans: localized buyer questions (e.g. “best plumber near me”) asked against each AI channel. Answers, brand mentions, and cited URLs are stored verbatim and aggregated here.
AI channels
Direct model answers plus a search-grounded channel (Brave web search + model synthesis). Channels currently measured: gpt-5.6, claude-sonnet-4.6, claude-opus-4.6, gemini-3-flash, gemini-3.1-pro, deepseek-v3, kimi-k3, glm-5.2, gpt-5.6-grounded.
Sample size & freshness
The exact sample (scans, answers, citations) and the collection window are shown above. Industries below the sample floor are hidden; small ones are flagged “limited sample”. The snapshot refreshes weekly.
Honesty rules
Only measured data — no estimates, no synthetic or demo data. Aggregates are computed directly from stored scan records; small samples are flagged instead of extrapolated.