Domain Authority

Domain Authority is fundamentally a link graph metric. It aggregates the total number of unique linking root domains and total backlinks pointing to a site. The algorithm does not merely count links; it models the quality and authority of the linking domains. A single link from a high-authority, relevant domain (like a .edu or .gov site) is exponentially more valuable than hundreds of links from low-quality, unrelated sites. The model is trained against actual Google search results to determine which link profiles correlate most strongly with high rankings.

The Domain Authority score operates on a 100-point logarithmic scale. This means it is significantly easier to grow a score from 20 to 30 than it is to grow from 70 to 80. The difference between higher scores is exponentially more difficult to bridge. For example:

• DA 20 to 30: Achievable with a solid foundational link profile.
• DA 70 to 80: Requires acquiring links from a critical mass of highly authoritative, established domains, often taking years of sustained effort. This scaling reflects the real-world difficulty of competing in the most competitive search verticals.

Domain Authority is not an absolute measure of quality and is not used by Google in its ranking algorithm. It is a comparative metric designed to predict relative ranking potential. A DA of 60 does not mean a site is 'good'; it means it is more likely to outrank a site with a DA of 40 in a head-to-head comparison for a given keyword. Its primary utility is in competitive analysis, allowing SEO strategists to benchmark their site's link equity against competitors in the same search engine results page (SERP) landscape.

The underlying algorithm is a supervised machine learning model trained on a vast dataset of actual search engine result pages. Moz uses a proprietary neural network that identifies the link-based features that best predict a domain's likelihood of ranking. The model is periodically retrained to adapt to changes in Google's algorithm. This predictive nature means DA can fluctuate not only because a site's link profile changed, but also because the model's understanding of which link signals are most important has been recalibrated.

Domain Authority measures the predictive strength of an entire root domain (e.g., example.com) or subdomain, not individual pages. This is distinct from Page Authority, which predicts the ranking potential of a single URL. A new blog post on a high-DA domain inherits a baseline level of trust and ranking potential from the domain's overall link equity, giving it an immediate advantage over an identical post on a low-DA site. This domain-level signal is a key factor in strategic site architecture.

The accuracy of a Domain Authority score is entirely dependent on the size, freshness, and quality of the Mozscape web index that powers it. This index crawls and stores a fraction of the total web. Consequently, DA is an estimation based on discovered links. If the Mozscape crawler has not discovered a significant portion of a site's backlinks, its DA score will be artificially low. This limitation is why DA should be used as a directional indicator, not a precise, infallible metric.

The foundational link analysis algorithm developed by Larry Page and Sergey Brin at Stanford. PageRank measures the importance of a web page by treating links as votes, where links from high-importance pages carry more weight than those from obscure sources. The algorithm models a random surfer who clicks links and occasionally jumps to a random page, computing a probability distribution that represents the likelihood of landing on any given page. Unlike Domain Authority—which is a proprietary Moz metric—PageRank is a query-independent, global importance score that Google still uses as one of hundreds of ranking signals.

A semi-automatic link analysis technique designed specifically to combat web spam by separating reputable pages from low-quality ones. The algorithm begins with a manually curated set of highly trusted seed pages, then propagates trust outward through the link graph. Trust diminishes as distance from the seed set increases, meaning pages linked from trusted sources inherit trust, while pages in spam neighborhoods receive low scores. This concept directly influences modern authority metrics by establishing that not all links are equal—a link's value depends on the trustworthiness of its source.

A network theory measure that forms the mathematical backbone of both PageRank and Domain Authority. Eigenvector centrality scores nodes based on the principle that connections to high-scoring nodes contribute more than connections to low-scoring nodes. In the context of the web graph:

• A backlink from a high-authority domain (like nature.com) passes more equity than one from an unknown blog
• The score is computed recursively until convergence
• The dominant eigenvector of the adjacency matrix represents the centrality values This recursive weighting is why Domain Authority emphasizes linking root domains over raw link count.

A complementary metric to Domain Authority that measures the percentage of sites with similar features to a target site that have been penalized or banned by search engines. Developed by Moz, Spam Score analyzes 27 common flags found in penalized domains, including:

• Low ratio of content to external links
• Presence of a high number of external links in navigation
• Thin or scraped content patterns
• Domain name characteristics associated with spam A high Spam Score combined with a high Domain Authority suggests manipulated link building rather than genuine authority.

Hyperlink-Induced Topic Search (HITS), developed by Jon Kleinberg at Cornell, introduced the dual-concept of hubs and authorities. Hubs are pages that link to many high-quality resources on a topic, while authorities are pages that receive links from many good hubs. This mutual reinforcement creates a bipartite graph where:

• A good hub points to many good authorities
• A good authority is pointed to by many good hubs This hub-authority distinction predates and conceptually parallels how Domain Authority evaluates both the quantity and quality of linking root domains.

An emerging algorithmic framework that evaluates the quality, relevance, and trustworthiness of sources cited by AI systems and search engines. Unlike Domain Authority—which focuses on link equity—citation integrity scoring assesses:

• The factual accuracy of the cited source's historical claims
• Whether the source is primary or merely aggregating
• The author's demonstrated expertise on the specific topic
• Recency and staleness of the cited information This concept extends traditional authority metrics into the generative AI era, where models must distinguish between high-confidence and low-confidence reference material.