[AI Search Visibility]

Domain Authority and AI Citation Rate: Is There a Statistically Significant Correlation?

Two independent studies measured the actual correlation between domain authority and AI citation rate. Neither found the relationship link builders are counting on.

The short answer

Across the two largest independent studies on this question, domain authority correlates weakly with AI citation rate at best and close to zero at worst. Surfer SEO's analysis of roughly 5 million citations put PageRank and Domain Score correlations at -0.065 and -0.051. Ahrefs' study of 75,000 brands found Domain Rating at 0.326, respectable but still less than half the strength of branded web mentions at 0.664. Domain authority is not nothing, but it is not the lever GEO strategy should be pulling first.

Does Domain Authority Correlate With AI Citation Rate?

Weakly, and less consistently than most SEO teams assume. The two largest independent studies published on this question, one from Surfer SEO in July 2026 and one from Ahrefs in 2025, both measured the actual statistical relationship between domain authority metrics and how often a domain gets cited in AI answers. Neither found a relationship strong enough to justify domain authority as a primary citation strategy.

Surfer's research, built on roughly 5 million citation URLs, found correlations close enough to zero to call noise. Ahrefs' research, built on 75,000 brands, found a real but modest positive correlation for Domain Rating, well behind the strength of branded mentions across the open web. Put the two together and the honest answer is: domain authority explains a small, inconsistent slice of why AI engines cite what they cite. It is not the dominant factor, and it is not the mechanism worth engineering content around.

That matters because link building is expensive and slow, and a lot of it is currently being sold as a GEO tactic on the assumption that what ranks Google gets cited by ChatGPT. The data below says that assumption does not hold.

What Does the Largest Study on This Question Actually Show?

Surfer SEO analyzed three months of AI Tracker data, around 5 million unique citation source URLs pulled from 20,000 prompts, across four systems: AI Mode, AI Overviews, ChatGPT, and Perplexity. They tested three separate domain authority proxies against citation likelihood using Spearman correlation, the standard statistic for measuring how consistently two variables move together on a scale from -1 to 1.

The results, published July 14, 2026:

  1. PageRank vs. citation likelihood: -0.065
  2. Harmonic Centrality vs. citation likelihood: +0.010
  3. Domain Score (Surfer, via DataForSEO) vs. citation likelihood: -0.051
Domain authority and trust signals: correlation with AI citation and visibility outcomes
SignalCorrelation (Spearman r)Outcome measuredStudy
Branded web mentions0.664AI Overview brand visibilityAhrefs, 75,000 brands
Branded anchors0.527AI Overview brand visibilityAhrefs, 75,000 brands
Branded search volume0.392AI Overview brand visibilityAhrefs, 75,000 brands
Domain Rating0.326AI Overview brand visibilityAhrefs, 75,000 brands
Backlinks0.218AI Overview brand visibilityAhrefs, 75,000 brands
Harmonic Centrality0.010AI citation likelihood (ChatGPT, Perplexity, AI Mode, AI Overviews)Surfer SEO, about 5 million citations
Domain Score-0.051AI citation likelihoodSurfer SEO, about 5 million citations
PageRank-0.065AI citation likelihoodSurfer SEO, about 5 million citations

Does Ahrefs' Research Agree, or Complicate the Picture?

It complicates it, in a useful way. Ahrefs analyzed 75,000 brands, filtered to domains with Domain Rating above 40, and measured which signals correlated with AI Overview brand visibility. Domain Rating came back at 0.326, a real, moderate positive correlation, not noise the way Surfer's PageRank figure was.

But Domain Rating was not the strongest signal in that same dataset. It was fourth. Branded web mentions correlated at 0.664, more than double Domain Rating's figure. Branded anchors came in at 0.527 and branded search volume at 0.392. Backlinks, the raw material most link building campaigns are built to produce, correlated at just 0.218, the weakest signal Ahrefs tested. The full comparison is below.

Read the two studies side by side and the pattern holds even where the numbers differ: whatever domain authority contributes to AI citation, something else contributes more. In Ahrefs' data that something is being talked about, not linked to. In Surfer's data, domain authority barely contributes at all.

Why Doesn't Google Ranking Authority Carry Over to AI Citation?

Because AI systems are not re-running your Google ranking, they are running their own retrieval. A separate Ahrefs study, published August 11, 2025, analyzed 15,000 long-tail queries, ran them through Google and Bing, then submitted the same prompts to ChatGPT, Gemini, Copilot, and Perplexity, and cross-referenced every citation against search ranking position.

Only 12% of URLs cited by ChatGPT, Gemini, and Copilot ranked in Google's top 10 for the identical query. Perplexity was the outlier, with close to a third of its citations pointing to top-10 pages. Overall, 80% of AI citations did not rank anywhere on Google for the original prompt at all. Ahrefs attributes this to query fan-out: AI assistants retrieve against dozens of reformulated variations of a question, not the single query a rank tracker measures.

That is the mechanical reason domain authority, a metric built entirely from Google's link graph and ranking behavior, is a poor predictor of a citation process that barely consults Google's rankings in the first place.

Is the Correlation Statistically Significant, or Just Noise?

Neither study published p-values, so 'statistically significant' is the wrong question to ask of this data. The right question is effect size, and on that measure the answer is clear. A Spearman r of -0.065 or -0.051, Surfer's findings for PageRank and Domain Score, explains essentially none of the variance in whether a page gets cited. Even Ahrefs' 0.326 for Domain Rating, the strongest domain-authority figure either study produced, explains roughly 10% of the variance at best (r squared of about 0.11), leaving the other 90% to factors domain authority does not capture.

For comparison, branded web mentions at 0.664 explain closer to 44% of the variance in AI Overview visibility in the same Ahrefs dataset. That is the gap that matters more than any significance test: domain authority is a weak-to-moderate signal sitting well behind a stronger one, in both studies, on both sides of the Atlantic of AI platforms tested.

Not as a primary strategy, based on what these two studies actually measured. Backlinks specifically correlated at just 0.218 with AI Overview visibility in Ahrefs' data, the weakest of five signals tested, and domain authority overall performed inconsistently across both studies. If the goal is Citation Share, the share of relevant AI answers that cite you, the data points toward brand mentions, coverage depth, and freshness as the higher-leverage spend, not link acquisition.

This is not a claim that domain authority is worthless or that AI models can be tricked into citing weak content. Omnicite does not build citation strategy around gaming a model; the honest read of this data is that AI systems weight quality, coverage, and freshness more heavily than the link graph a domain authority score is built from. That is a harder thing to buy than backlinks, and it is the actual mechanism behind why some sites get cited across ChatGPT, Perplexity, and AI Overviews and others with higher Domain Ratings do not.

The practical takeaway for a growth team deciding where to spend the next quarter's content and outreach budget: put it into being mentioned, cited, and quoted across the sources AI models already trust, not into acquiring more links to a domain score that two independent studies just found barely moves the needle.

Domain authority metrics vs. AI citation likelihood: Spearman correlation across about 5 million citations
00.00.0-0.1Spearman r-0.1Spearman r0.0Spearman rPageRankDomain ScoreHarmonic Centrality

Source: Surfer SEO, 2026-07-14

Key takeaways

  • Surfer SEO's analysis of about 5 million citations found domain authority proxies (PageRank, Domain Score) correlate close to zero, -0.065 and -0.051, with AI citation likelihood.
  • Ahrefs' study of 75,000 brands found a moderate positive correlation for Domain Rating (0.326) with AI Overview visibility, but it ranked fourth of five signals tested.
  • Branded web mentions correlated at 0.664 in the same Ahrefs study, more than double Domain Rating and about three times backlinks (0.218).
  • Only 12% of URLs cited by ChatGPT, Gemini, and Copilot rank in Google's top 10 for the same query; 80% do not rank at all, per Ahrefs' 15,000-query study.
  • Neither study reports statistical significance testing, but effect size alone rules out domain authority as a strong predictor of AI citation.
  • The data favors spending on brand mentions, coverage depth, and freshness over link acquisition as a GEO investment.

Omnicite Editorial. "Domain Authority vs AI Citation Rate: The Data" The Citation Report, Omnicite. https://omnicite.co/blog/domain-authority-ai-citation-rate-correlation/

Sources

Domain authority metrics (PageRank, Domain Score) correlate close to zero, -0.065 and -0.051 Spearman r, with AI citation likelihood across ChatGPT, Perplexity, AI Mode and AI Overviews, based on about 5 million citations. Surfer SEO, 2026-07-14

Domain Rating correlates at 0.326 with AI Overview brand visibility, versus 0.664 for branded web mentions and 0.218 for backlinks, across 75,000 brands. Ahrefs, 2025-05-26

Only 12% of URLs cited by ChatGPT, Gemini, and Copilot rank in Google's top 10 for the same query; 80% of AI citations do not rank at all, based on 15,000 long-tail queries. Ahrefs, 2025-08-11

Frequently asked questions

Does domain authority affect how often you get cited in AI search?

A little, but not consistently. Ahrefs found a moderate positive correlation (0.326) between Domain Rating and AI Overview visibility across 75,000 brands. Surfer SEO found domain authority proxies correlate close to zero with citation likelihood across ChatGPT, Perplexity, AI Mode, and AI Overviews, based on about 5 million citations. The honest summary: it is a minor factor, not a dominant one.

Is there a statistically significant correlation between domain authority and AI citation rate?

Neither major study published p-values, so significance testing is not the right lens here. Effect size is: Surfer's correlations (-0.065, -0.051, +0.010) are close enough to zero to be noise, and even Ahrefs' strongest domain-authority figure (0.326) explains only about 10% of the variance in AI Overview visibility.

Should I invest in link building to improve GEO?

Not as a first move. Backlinks were the weakest of five signals Ahrefs tested against AI Overview visibility, correlating at just 0.218. Branded web mentions, at 0.664, correlated more than three times as strongly. Budget aimed at earning mentions and coverage will likely outperform budget aimed at acquiring links.

What matters more than domain authority for AI citations?

In Ahrefs' data, branded web mentions, branded anchors, and branded search volume all correlated more strongly with AI Overview visibility than Domain Rating did. Ahrefs' separate query-overlap study also found that AI citation barely tracks Google ranking position at all, with 80% of AI citations coming from pages that do not rank for the original query.

Does a high domain rating guarantee AI citations?

No. Surfer SEO's analysis of about 5 million citations found domain authority metrics correlate close to zero with citation likelihood, and Ahrefs' study found Domain Rating explains only a modest share of AI Overview visibility, well behind brand mentions.

How is domain authority different from Citation Share?

Domain authority is an input metric built from a site's link graph. Citation Share is an outcome metric: the percentage of relevant AI answers in a category that actually cite you. The data in this study is exactly why the two diverge, a link-graph score is a weak predictor of the citation outcome that matters.