Original, anonymized audit data
2026 AI Citation Readiness Benchmark
A view of page signals we can measure, such as access, content, and schema. These gaps are worth checking, but a score cannot predict inclusion in an AI answer.
Scoring view version 2026.07. Scores are aggregated; URLs, brands and users are not published.
By Evan Weber · Updated October 20, 2026
What an average can tell you
Imagine three fictional audit scores of 20, 40 and 60. Their arithmetic mean is 40. That does not mean the pages have a 40% chance of being cited, and it does not tell you which individual page has an access problem.
Use an aggregate to notice a dimension worth investigating. Then inspect the actual finding on your page. A low structured-data average does not make schema the first task if your important page is inaccessible or leaves a customer’s question unanswered.
Read the sample before comparing yourself
The sample counts completed URL audits. It is not a count of unique companies and is not presented as a representative survey of the web. Different dimensions may have different usable records, including older audits without the current crawler-access dimension.
A readiness score is a product diagnostic. It is not observed citation share, search traffic or conversion performance. Use the methodology to interpret the weights and the measurement guide for those separate outcomes.
Average readiness scores
Based on 68 completed audits. Legacy audits without the current crawler dimension are excluded from that dimension only.
Overall score distribution
Methodology
The benchmark uses completed URL audits and reports arithmetic means for the current scoring dimensions. It contains no inferred traffic, ranking, or conversion data. Scores indicate citation readiness, not guaranteed inclusion in any AI answer.
The public endpoint is cached and supplies aggregate statistics. Version 2026.07 identifies this scoring view, not a claim that every audit was collected in July.
View the public aggregate JSON. Compare the score with the underlying page findings before choosing your next edit.