{
 "study": "s3",
 "version": "1.1",
 "collected": "2026-09-26",
 "statistics": {
  "sample": {
   "list": "Tranco top 1M, list ID L5PZ4 (downloaded 2026-09-26)",
   "domains_attempted": 10000,
   "domains_with_live_homepage": 5902,
   "crawl_window": [
    "2026-09-26T04:16:26.234430Z",
    "2026-09-26T04:38:00Z"
   ]
  },
  "s3_md_negotiation": {
   "n": 191,
   "pct": 3.2,
   "domains": [
    "cloudflare.com",
    "wordpress.org",
    "workers.dev",
    "sentry.io",
    "b-cdn.net",
    "hosting24.com",
    "pages.dev",
    "linktr.ee",
    "hostinger.com",
    "zendesk.com",
    "gitlab.com",
    "addtoany.com",
    "bitdefender.net",
    "vercel.app",
    "jotform.com",
    "intercom.io",
    "oxylabs.io",
    "mixpanel.com",
    "life360.com",
    "agora.io",
    "readthedocs.io",
    "netlify.app",
    "trendyol.com",
    "onesignal.com",
    "supabase.co",
    "usercentrics.eu",
    "typeform.com",
    "siemens.com",
    "noodlemagazine.com",
    "tgju.org",
    "ukdevilz.com",
    "elsevier.com",
    "classlink.com",
    "fastdl.app",
    "kir2kos.net",
    "indown.io",
    "intercom.com",
    "bitdefender.com",
    "themeisle.com",
    "metamask.io",
    "onrender.com",
    "laravel.com",
    "customer.io",
    "generatepress.com",
    "cookiebot.com",
    "asana.com",
    "sendbird.com",
    "bunny.net",
    "inmotionhosting.com",
    "samba.tv",
    "miro.com",
    "okx.com",
    "home-assistant.io",
    "polymarket.com",
    "railway.app",
    "slashdot.org",
    "aditude.io",
    "cookieyes.com",
    "padlet.com",
    "webflow.com"
   ]
  },
  "s3_md_servers": [
   [
    "cloudflare",
    136
   ],
   [
    "vercel",
    14
   ],
   [
    "nginx",
    7
   ],
   [
    "unknown",
    7
   ],
   [
    "framer/26fa766",
    6
   ],
   [
    "netlify",
    5
   ],
   [
    "bunnycdn-bd1-1043",
    2
   ],
   [
    "c2 comms cloud (82ec06f1)",
    2
   ]
  ],
  "s3_md_with_vary_accept": {
   "n": 183,
   "pct_of_md": 95.8
  },
  "s3_html_n": 5710,
  "s3_alt_markdown_link": {
   "n": 28,
   "pct_of_html": 0.5
  },
  "s3_capped_refetched": 790,
  "s3_jsonld": {
   "any_pct": 45.8,
   "organization_pct": 36.1,
   "website_pct": 31.0,
   "sameAs_pct": 28.2,
   "invalid_block_pct": 0.6,
   "top_types": [
    [
     "Organization",
     1783
    ],
    [
     "WebSite",
     1770
    ],
    [
     "SearchAction",
     1226
    ],
    [
     "ImageObject",
     1222
    ],
    [
     "WebPage",
     921
    ],
    [
     "ListItem",
     742
    ],
    [
     "PostalAddress",
     653
    ],
    [
     "BreadcrumbList",
     625
    ],
    [
     "EntryPoint",
     624
    ],
    [
     "ContactPoint",
     615
    ],
    [
     "PropertyValueSpecification",
     328
    ],
    [
     "Person",
     291
    ],
    [
     "ReadAction",
     281
    ],
    [
     "NewsMediaOrganization",
     267
    ],
    [
     "CollectionPage",
     217
    ]
   ],
   "noai_meta_pct": 0.3,
   "noai_meta_n": 18
  },
  "s3_note": "Homepages over 600 KB were re-fetched in full (up to 8 MB) for the JSON-LD count.",
  "s3_capped_pages": 807,
  "s3_jsonld_by_tier": {
   "1-1,000": {
    "n": 511,
    "jsonld_pct": 43.8
   },
   "1,001-5,000": {
    "n": 2235,
    "jsonld_pct": 45.3
   },
   "5,001-10,000": {
    "n": 2964,
    "jsonld_pct": 46.5
   }
  },
  "s3_derived": {
   "md_behind_cloudflare_pct": 71.2,
   "no_jsonld_pct": 54.2
  },
  "s3_md_vs_html": {
   "n_md_sites": 191,
   "n_compared": 185,
   "median_md_bytes": 13081,
   "median_html_bytes": 401073,
   "median_size_reduction_pct": 96.1,
   "p25_reduction_pct": 90.0,
   "md_with_frontmatter_pct": 79.6,
   "md_sites_with_valid_llms_txt_pct": 43.5,
   "md_sites_with_valid_llms_txt_n": 83,
   "md_with_frontmatter_n": 152
  },
  "s3_capped_not_refetched": 17,
  "s3_representation_check": {
   "collected": "2026-09-28",
   "sample_n": 191,
   "analyzed_n": 186,
   "excluded": {
    "webflow.com": "M1 no longer Markdown (text/html; charset=utf-8)",
    "webflow.io": "M2 no longer Markdown (text/html; charset=utf-8)",
    "zeotap.com": "H1 request failed (ConnectTimeout)",
    "datacamp.com": "H1 returned 403",
    "circle.com": "M1 no longer Markdown (text/html; charset=utf-8)"
   },
   "no_longer_markdown": {
    "n": 3,
    "of": 191,
    "domains": [
     "circle.com",
     "webflow.com",
     "webflow.io"
    ]
   },
   "equivalent": {
    "k": 2,
    "n": 186,
    "pct": 1.1,
    "wilson95": [
     0.3,
     3.8
    ]
   },
   "equivalent_without_page_chrome_exploratory": {
    "k": 7,
    "n": 186,
    "pct": 3.8,
    "wilson95": [
     1.8,
     7.6
    ]
   },
   "sentences_kept": {
    "n": 178,
    "median_pct": 83.8,
    "ci95_pct": [
     77.0,
     87.9
    ]
   },
   "sentences_kept_without_page_chrome_exploratory": {
    "n": 176,
    "median_pct": 98.3
   },
   "links_kept": {
    "n": 182,
    "median_pct": 29.9,
    "ci95_pct": [
     22.6,
     35.8
    ]
   },
   "headings_kept": {
    "n": 178,
    "median_pct": 88.9,
    "ci95_pct": [
     85.9,
     92.7
    ]
   },
   "under_20pct_of_sentences": {
    "n": 23,
    "of": 178,
    "pct": 12.9
   },
   "markdown_identical_on_repeat": {
    "n": 176,
    "of": 186
   },
   "html_identical_on_repeat": {
    "n": 97,
    "of": 186
   },
   "noise_floor_median_pct": 100.0,
   "frontmatter": {
    "n": 147,
    "of": 186
   },
   "jsonld": {
    "pages_with_jsonld": 137,
    "markdown_includes_jsonld": 76,
    "jsonld_has_facts_not_in_text": 127
   },
   "by_platform": {
    "cloudflare": {
     "n": 133,
     "equivalent_k": 2,
     "pct": 1.5,
     "wilson95": [
      0.4,
      5.3
     ],
     "median_claims_cov_html_to_md": 0.862
    },
    "framer": {
     "n": 6,
     "equivalent_k": 0,
     "pct": 0.0,
     "wilson95": [
      0.0,
      39.0
     ],
     "median_claims_cov_html_to_md": 0.821
    },
    "netlify": {
     "n": 5,
     "equivalent_k": 0,
     "pct": 0.0,
     "wilson95": [
      0.0,
      43.4
     ],
     "median_claims_cov_html_to_md": 0.87
    },
    "nginx": {
     "n": 10,
     "equivalent_k": 0,
     "pct": 0.0,
     "wilson95": [
      0.0,
      27.8
     ],
     "median_claims_cov_html_to_md": 0.632
    },
    "other/unknown": {
     "n": 18,
     "equivalent_k": 0,
     "pct": 0.0,
     "wilson95": [
      0.0,
      17.6
     ],
     "median_claims_cov_html_to_md": 0.474
    },
    "vercel": {
     "n": 14,
     "equivalent_k": 0,
     "pct": 0.0,
     "wilson95": [
      0.0,
      21.5
     ],
     "median_claims_cov_html_to_md": 0.833
    }
   }
  }
 }
}