Every number on this page is an aggregate of automated scans Antileak ran against real, publicly accessible business homepages — restaurants in Chelsea, Manhattan, NY. Each scan fetched the homepage the way AI crawlers fetch it — the raw HTML, with no JavaScript executed — and read the site's robots.txt, schema markup and public DNS records. Nothing here is an estimate, a survey answer, or an industry benchmark borrowed from somewhere else.
| Finding | Share | Measured on |
|---|---|---|
| had structured (schema.org) business data on the homepage | 60% | 40 sites |
| had a DMARC record on their domain | 18% | 40 sites |
| had an SPF record | 40% | 40 sites |
| were missing a page title | 0% | 40 sites |
| had no meta description for answers to quote | 28% | 40 sites |
| declared a mobile viewport | 78% | 40 sites |
AI assistants build answers from what a site's raw HTML gives them: structured business data they can quote, pages they are allowed to crawl, and titles and descriptions that say what the business is. A site missing those isn't penalised by AI assistants — it simply isn't there. The table above is how often those building blocks were present in this group.
Where does your site stand? The same scan behind these numbers is free, takes about 60 seconds, and needs no account. It fetches your homepage exactly the way GPTBot, ClaudeBot and PerplexityBot do and tells you what they receive.