AI Web Checkby noviKEY
Menu

Research · aggregate data

What AI Web Check observations show

Public statistics from aggregated AI Web Check results: AI Readiness, AI-commerce and optional machine-readable signals. These observations describe the AI Web Check sample, not the entire web.

Important: counts represent checks and reports, not unique websites. One site may be checked more than once. Site addresses and individual report data are not published on this page.

Period
2026-07-19 — 2026-08-17
Latest data
2026-08-15
Methodology version
4.2

Methodology 4.2

AI Readiness

AI Readiness is still gathering enough data for public statistics. Percentages will appear when the sample is large enough.

Days with aggregate data
13
Methodology version
Methodology 4.2
Minimum for publication
20 observations

AI Readiness uses results from the current methodology version, so this sample can differ from the overall report count.

Once enough data is available, this section will show: the share with AI Readiness ≥ 75, structured data, search-crawler access, canonical, content without mandatory JavaScript, and separately llms.txt and WebMCP.

Optional signals

Additional machine-readable signals

These metrics show how often the signals appear in the sample. They are not part of AI Readiness.

These optional-signal shares will appear when there are enough comparable checks using the current methodology.

AI-commerce 1.0

Commerce and AI-agent readiness

90 successful AI Web Check reports in the observation window.

Commerce context confirmed

6%

6%

Based on 90 evaluations.

Share of successful AI Web Check reports where commerce context was confirmed.

AI-commerce score applicable

Not enough data yet to publish this share.

Share of confirmed commerce contexts where AI-commerce 1.0 can be scored.

Product / Offer ready

Not enough data yet to publish this share.

Share of commerce checks without a significant Product / Offer issue.

UCP detected

Not enough data yet to publish this share.

Share of applicable commerce checks where a valid UCP profile was detected.

A2A Agent Card detected

Not enough data yet to publish this share.

Share of applicable commerce checks where an A2A Agent Card was detected.

OpenAPI detected

Not enough data yet to publish this share.

Share of applicable commerce checks where an OpenAPI description was detected.

Commerce context was confirmed in 6% of 90 successful reports in the AI Web Check sample. This describes service usage, not the share of commerce websites on the web.

How to read these data

  • Not unique websites. One site may be checked more than once, so the number of checks is not the number of websites.
  • Sample scope. The statistics include only sites that users chose to check with AI Web Check, so the sample does not represent the entire web.
  • Methodology version. AI Readiness statistics use results calculated with the current methodology version.
  • Small samples. Percentages are not published until there is enough data for a reliable public result.
  • Privacy. Public statistics do not include site addresses, raw HTML or individual report contents.

Criterion definitions: public methodology. Practical remediation: technical guides.