ReadinessIndex.io Logo, orange and white bar charts on a purple background

The Census

Census Q4 2026

1,000 organizations · 929 scanned · Index 11.0, in effect 2026-09-24 · generated 2026-09-25

Complete, with gaps

929 of 1,000 organizations returned a score.

The run is finished. The other 71 refused the scanner, timed out, or failed outright, and no further attempt is coming — so every share below is out of what was actually scanned, never out of the cohort size. That a site cannot be read at all is itself a result.

coverage
929 scored · 71 unreachable
The cohorts

Five populations, one scan

CohortDrawn fromMedian and spread ScannedGated
state-portals The primary web portal of each US state government. Cybersecurity and Infrastructure Security Agency (CISA)
38.0 · 23.0–49.0
50/50 26 of 50
universities The 100 largest US universities by total enrollment. National Center for Education Statistics (NCES)
42.0 · 26.0–58.0
94/100 42 of 94
health-systems The 50 largest US nonprofit health systems by total revenue. Internal Revenue Service (IRS), with Wikidata
38.5 · 25.0–60.0
46/50 33 of 46
nonprofits The 100 largest US nonprofits by total revenue. Internal Revenue Service (IRS)
39.0 · 26.0–60.0
93/100 55 of 93
sp500 The 500 companies in the S&P 500. Wikipedia contributors
41.0 · 25.0–59.0
463/500 271 of 463
wordpress-100 The 100 highest-ranked WordPress sites in the Tranco list. Tranco (Le Pochat et al., NDSS 2019)
47.5 · 30.0–62.0
90/100 35 of 90
drupal-100 The 100 highest-ranked Drupal sites in the Tranco list. Tranco (Le Pochat et al., NDSS 2019)
45.0 · 31.0–68.0
93/100 45 of 93
All cohorts 929 of 1,000 sites scanned
42.0 · 23.0–68.0
929 507
purple line = median · grey band = middle of the range
What gated means

Gated means a site failed AIR-1.1 or AIR-1.2 — it blocks AI crawlers outright, so nothing else about it can be reached. The score is capped rather than averaged away, because a site nothing can read is not partly ready.

Where the failures are

The tests almost nobody passes

The share is out of the sites each test actually applied to — a test that was N/A for a site isn't counted against it, and the count is shown so a share with a small denominator can't masquerade as a sector finding.

Near-universal — 95% and above 8 checks · 3 of them worth 2 points or more
TestWhat it asksFailing OfPoints
LI-3 IndexNow fires on publish and update
99.9%
929 0
LI-4 Editorial policy page referenced via publishingPrinciples
99.8%
929 0
LI-1 HowTo on procedural content
99.7%
390 0
LI-2 speakable on summaries and ledes
99.5%
929 0
AIR-7.1 Markdown companion for every canonical page
99.3%
672 2
AIR-1.4 Content Signals declared in robots.txt
99.2%
929 3
AIR-7.3 Accept: text/markdown content negotiation
98.4%
929 2
AIR-4.4 Correct X-Robots-Tag on non-HTML resources
97.9%
622 1
Common — 90 to 95% 5 checks · 4 of them worth 2 points or more
TestWhat it asksFailing OfPoints
AIR-5.1 Author entity pages with credentials
94.9%
929 2
AIR-3.3 Vertical-specific schema types
94.4%
270 4
AIR-1.10 /llms.txt exists and is generated from the content model
92.1%
929 2
AIR-1.13 About, Contact, and organizational detail are complete
91.5%
929 2
LI-5 NLWeb endpoint exposed as an MCP server with an ask method
91.0%
929 0

Coral bars mark tests worth 2.0 points or more — the failures that actually move a score. Show all tests →

Cite as

Readiness Index Census Q4 2026, TEN7, scored against Index 11.0 (2026-09-24).

Licence CC BY-SA 4.0
Generated 2026-09-25

Is your site ready for AI?

Scan a page, get a number you can verify. Free.