Definition
Domain Authority (DA) is a score from 1 to 100, created by the SEO software company Moz, that predicts how likely a website is to rank in search results compared with other sites. It is calculated by a machine learning model that mainly uses backlink data from Moz's own link index, chiefly the number and quality of unique domains linking to the site. Higher scores indicate a stronger link profile.
DA is a third-party estimate, not a search engine metric. Moz's own documentation states that Domain Authority is not a Google ranking factor and has no effect on the SERPs. It is a comparative tool: Moz recommends judging a site's DA against the sites it competes with, because there is no universal good score. The scale behaves roughly logarithmically, so moving from 20 to 30 is much easier than moving from 70 to 80. Scores can also change without anything happening on your site, when Moz's index is refreshed, when very large sites gain links and shift the scale, or when Moz updates the model, as it did with Domain Authority 2.0 in 2019.
Other tools sell similar scores built on their own link indexes, such as Ahrefs' Domain Rating and Semrush's Authority Score. Numbers from different tools are not interchangeable.
DA stays popular because the thing it approximates still matters to Google. Links remain part of how Google judges pages: in the US antitrust case against Google, a Department of Justice trial exhibit described PageRank as one of Google's original signals that 'continues to be one of the factors' in its page quality signals. A site with many strong, relevant links tends to rank more easily, get crawled more often and appear more often in the results that AI search systems retrieve from. DA is a reasonable shorthand for that link strength, and for vetting link and PR opportunities.
Its limits are sharper in AI search. Generative engines do not look up a domain score; they retrieve passages that answer a specific question, then choose sources partly on relevance, specificity and consistency with other sources. A high-DA general publisher with a thin page on a topic can lose the citation to a specialist site with a lower score and a detailed, current answer. Brand mentions without links, which DA ignores, still shape how models describe a brand. Entity SEO, topical authority and third-party coverage in the sources each platform favors often explain AI citation patterns better than any domain-level score. Our analysis of ChatGPT citation share by domain rank looks at how citations are spread across sites of different authority.
Use DA for what it is good at: benchmarking link strength against direct competitors, sizing the gap in backlinks, and screening outreach targets. Do not use it as a goal in itself, or as a proxy for AI visibility. Measure that directly by tracking whether answers mention your brand and cite your pages, using GEO performance metrics.
Examples of Domain Authority
- A specialist payroll-compliance blog with a DA of 38 is cited by Google AI Mode for state-specific payroll questions ahead of a DA 90 business magazine whose article covers the topic in two paragraphs.
- An agency uses DA to shortlist 60 industry publications for a data-led PR campaign, then checks which of them already appear as cited sources in AI answers in its client's category.
- A company's DA drops from 52 to 47 with no change in its links or rankings; Moz's index refresh had shifted scores across its competitors as well, so the relative position was unchanged.
- A startup grows from DA 12 to DA 30 in 18 months through original research that earns links, while separately tracking that its AI mention rate climbs as the same research gets quoted in reviews.
