Domain authority does not predict whether an AI tool will cite a site. A Search Atlas study published on October 29, 2026 looked at 21,767 domains across ChatGPT, Perplexity and Gemini and reported a weak negative correlation between traditional authority scores and LLM visibility: r of minus 0.058 for Moz Domain Authority and minus 0.049 for Ahrefs Domain Rating. A correlation that close to zero means the scores carry almost no signal about which sites AI answers cite.
The study compared authority scores with how often each domain appeared in AI answers, and the result was effectively no relationship. A correlation near minus 0.05 is statistically small, and it cuts against the assumption that has guided link building for years: that a higher authority score means more visibility, in search and now in AI answers too. Search Atlas is careful about the claim, and so should anyone reporting it. A near-zero correlation is not proof that authority is irrelevant, and the study's own write-up notes that calling the metrics uncorrelated goes further than the numbers support. But it is strong evidence that authority scores are not the gate to AI citations that many sellers claim.
The finding fits a pattern across four studies gathered in a November 12, 2026 roundup. A Featured study of 22,881 AI citations across 11,499 domains found that 34.5% went to sites with Moz Domain Authority below 40, and 13.2% to sites below 20. A Wellows study of 20.5 million citations from 386,000 domains found that 47.9% of citations came from domains under DA 40, spread across 346,348 domains. A Stacker study published September 17, 2026 tracked 895 stories with about 212,000 pickups and reported that 4.7% of collected AI responses cited a story from its network, with domains under DR 40 accounting for about one-fifth of those citations.
Read together, the four studies tell two stories at once. High-authority sites earn far more citations per domain, because a single strong domain can be cited hundreds of times. But most citations in the largest datasets come from lower-authority domains, because there are so many of them. Authority is a rough stand-in for the signals AI tools actually read, not a gate: a small site with the right page can be cited, and a strong domain with the wrong page will not be.
That distinction matters because of how the scores are built. Domain Authority is a Moz score, Domain Rating is an Ahrefs score, and Semrush calls its version Authority Score. Each estimates the strength of the links pointing at a domain on a scale of 0 to 100. None of them measures editorial quality, and none is an input any AI engine is known to use. Google, OpenAI, Perplexity and Anthropic have not confirmed using any of them. Featured's study states plainly that DA was used to describe the sites that got cited, not as a factor the AI systems applied.
If authority is not the gate, the next question is what is, and the evidence points at retrieval and content structure. A 2026 analysis of 158,099 answers and 468,037 citations found that the engine's own retrieval explains 32% of citation variation, while every page property combined explains 2.8%, and only 9% to 13% of variation lies between websites in the first place. In other words, most of what decides a citation happens inside the AI system and on the page, not at domain level.
That finding lines up with what Ahrefs found in its own research. A study of 75,000 brands found branded web mentions correlated with AI Overview visibility at 0.664, while backlinks sat near zero, and YouTube mentions correlated at 0.737. The pages that get cited are the ones that match the question, carry clear structure, and are mentioned across the web, not necessarily the ones sitting on the strongest domain.
For SEOs, the Search Atlas result is permission to stop optimising for a score and start optimising for the answer. A page that directly answers a question, with clear headings, quotable facts and original data, has a better chance of being cited than a thin page on a strong domain. The work that matters is at the page level: structure, clarity, evidence, and the brand mentions that tell an AI system a source is worth trusting.
For link builders, the finding changes what a placement is worth and how it should be sold. A link from a high-authority domain still has value in classic search, but it is not a ticket to AI visibility, and selling it as one is selling something the data does not support. The placements that earn AI citations are the ones on pages that AI tools actually retrieve and quote, whatever the domain's score. Pitch the page and the content, not the number, and measure success in citations and mentions, not in the authority score of the linking domain.
Watch for more studies that separate per-domain averages from total citation share, because the two tell different stories and both are needed. Watch also for whether the tool vendors change how they market their scores, and whether AI visibility tools start reporting retrieval and content factors instead of authority. The scores will not disappear, they are too useful as a shorthand, but their role is shrinking to what they always were: a rough estimate of link strength, not a measure of whether an AI will cite you.
Sources: Spotlight on Startups, November 12, 2026, reporting the Search Atlas study published October 29, 2026.
Related reading on this site: November Core Update Collides With Black Friday, and a Clean Read Will Not Come Until December, Core Update Volatility Started Days Before Google Announced It, November Search Console Click Spikes May Be Bots, SEJ Reports, SEJ Pulse: Core Update, ChatGPT Business Hub and AI Mode's New Model, AI Overview Citations: What 15 Studies Agree On, and Don't, and Black Friday SEO Starts Eight Weeks Out, and November Is Already Late.
Linkbaits helps you create content that earns citations and backlinks. Start with the toolkit and build your first linkable asset this week.
Start free trial