AI Readiness Across 194 Websites: The Median Site Scores 60/100

The readiness gap is 30 points

The median AI Readiness Score is 60/100. Median SEO on the same 194 sites is 90/100. A site can therefore clear the familiar SEO checks while leaving explicit entity facts, extractable answers, and structured page signals incomplete.

Foglift's Technical Audit measures both surfaces in one run. The result is a five-dimension report covering SEO, AI Readiness, performance, security, and accessibility. This study isolates the AI Readiness and SEO scores to quantify the extraction gap across the current scanned population.

One quarter of sites score 70 or higher

Forty-nine sites score at least 70, which is 25.3% of the sample. Five sites reach the 80s, and the highest observed score is 82. The largest single band is 60 to 69 with 54 sites. Another 38 sit in the 50s.

The 25th percentile is 48.3 and the 75th percentile is 69.8. That interquartile range gives operators a more useful comparison than the maximum because it shows where the middle half of current audited sites sits.

22.4% of SEO-strong sites remain below 50

The sample contains 165 sites with an SEO score of at least 70. Among them, 37 have an AI Readiness Score below 50. That is 22.4% of the SEO-strong cohort.

SEO and AI Readiness have a Pearson correlation of 0.29 in this sample. The relationship is positive and modest. This supports a practical operating model: keep SEO fundamentals healthy, then measure AI extraction signals directly instead of inferring them from SEO alone.

FAQ and entity gaps lead the issue list

A missing FAQ section appears on 141 sites (72.7%). Few internal links and missing entity identity markup each appear on 72 sites (37.1%). No structured data appears on 45 sites (23.2%). These are issue frequencies from the latest eligible audit for each domain.

The finding does not justify adding generic FAQ copy to every page. A useful FAQ answers real customer questions with self-contained facts. Entity markup and JSON-LD must also agree with visible page content.

Bottom-quartile sites carry 2.3 times as many readiness audit issues

Twenty-five sites score at or above the 90th-percentile threshold of 74. They average 2.7 SEO and AI Readiness audit issues per latest audit. The 49 sites scoring 48 or below average 6.2. The issue load is 2.3 times higher in the bottom quartile.

What the benchmark implies

Measure the extraction layer directly. A 90 median SEO score coexists with a 60 median AI Readiness Score. Foglift's free Technical Audit reports both in one five-dimension scan and identifies the exact issues behind the gap.

Answer the questions buyers actually ask. The largest issue frequency is missing FAQ content. Add concise answers only where customer language and page intent support them. Clear headings, entity facts, and matching schema make those answers easier to extract.

Connect readiness to outcomes. A Technical Audit measures whether the page is ready to be parsed. Repeated AI Visibility checks measure mentions, citations, sentiment, competitors, and answer position across ChatGPT, Claude, Perplexity, Gemini, and Google AI Overview. Teams need both measurements to know whether a shipped fix reached the answer layer.

Two score contracts, kept separate

The first edition of this report used the legacy AEO compatibility score. The August refresh uses the current AI Readiness field from Foglift's five-dimension Technical Audit. Treat the editions as separate benchmarks because their score contracts and data windows differ.

The May edition covered 1,386 scans across 344 domains, with 311 domains carrying the legacy score. In that sample, 44.5% of sites with SEO at 70 or higher had a legacy score below 50, and 29.6% had no JSON-LD. Those figures remain available as historical observations and are excluded from the August calculations above.

Methodology boundaries

• The current score field starts on July 17, so the 194-site sample is a short-window snapshot. It should be refreshed quarterly as the population grows.
• This is a self-selected audit population. It does not estimate the average score of every website on the public web.
• Hostname normalization removes a leading www. Other subdomains remain separate observations when operators scanned them independently.
• Issue frequencies are diagnostic associations. This report does not claim that fixing one issue causes a specific citation-rate increase.

Related Foglift research

• AI engines agree on brands more than sources. A cross-engine view of brand agreement and citation-domain overlap.
• Which AI search tools do five engines recommend?. A source-layer benchmark for high-intent category questions.
• AI Search Citation Benchmark, Q2 2026. The reference dataset covering 375 buyer-intent answers.