We audited 56 company domains on 6 November 2026 across eight dimensions of AI retrievability. The median score is 31.5 out of 100. Most sites are crawlable — and almost none of them carry the machine-readable entity, freshness and authority signals that decide whether an AI engine cites them.
36 of 56 domains land between 20 and 39. Two reach 60. None crosses 80.
Sorted from weakest to strongest. “Zero” means the dimension scored nothing at all.
| Dimension | Median | Scored zero | Below half | What it measures |
|---|---|---|---|---|
| Structured entity | 0/10 | 76.8% | 76.8% | Schema.org entity graph, sameAs profiles |
| Freshness | 0/10 | 58.9% | 94.6% | Sitemap lastmod, feed, dateModified |
| Authority anchors | 0/10 | 71.4% | 98.2% | Third-party citations of .gov/.edu/Wikipedia grade |
| Content depth | 4/10 | 35.7% | 62.5% | Extractable words, heading structure |
| Answer readiness | 4/10 | 19.6% | 91.1% | Question-form headings, tables, FAQ schema |
| Retrievability | 4/10 | 1.8% | 64.3% | Sitemap coverage, server-rendered text share |
| /llms.txt | 5/10 | 46.4% | 48.2% | Machine-readable index for AI crawlers |
| Crawler access | 10/10 | 5.4% | 5.4% | GPTBot / ClaudeBot / PerplexityBot permissions |
Crawler access scores a median of 10/10 and only 5.4% of sites block the frontier bots. AI engines are allowed in. They simply find little worth quoting.
76.8% of domains expose no structured entity graph; 71.4% have zero authority anchors. Without them an engine has no reason to prefer one brand over another.
58.9% show no freshness signal at all and 94.6% score below half. A site that never states what changed reads as dormant, whatever its traffic.
POST /api/diagnose/geo.Run the same eight-dimension audit on your own site. No signup, no email, results in seconds.