Domain Authority: The Moz link-graph metric explained, calculation methodology, and how aged-domain buyers read it

· Last reviewed · 13 min read

Domain Authority is the Moz link-graph metric that predicts the relative ranking strength of any website’s domain on a logarithmic scale from 1 to 100.

Moz introduced Domain Authority in 2012 as the first widely-adopted third-party predictive ranking score after Google deprecated the public PageRank toolbar.

The metric is calculated by a machine-learning model trained on dozens of link-graph signals plus the Moz Spam Score adjustment.

The SEO Domains analytical position: Domain Authority is one of four cross-validated link-graph proxies that aged-domain buyers read alongside Domain Rating, Trust Flow, and Citation Flow before any acquisition decision.

SEO Domains operates the curated marketplace with a 220,000+ pre-screened catalogue from $100 entry-level domains through $1.5 million premium acquisitions, ICANN-accredited transfer, and Managed Account expert support; every listing surfaces Domain Authority alongside the multi-metric screening framework that filters link-spam patterns at inventory ingestion.

What does Domain Authority measure?

Domain Authority measures the relative ranking strength of a website’s domain on a logarithmic scale from 1 to 100. Moz designed Domain Authority as a comparative predictor: a domain with DA 60 is, on average across the test query population, more likely to outrank a domain with DA 30 than the reverse.

The metric does not predict a specific SERP position; it estimates the probability of stronger versus weaker ranking outcomes when other factors are held equal.

Moz Free Domain Authority Checker showing seo.domains with Domain Authority 23, Linking Root Domains 143.2k, Ranking Keywords 292, Spam Score 1%
Figure 1. Moz Free Domain Authority Checker surfacing Domain Authority alongside Linking Root Domains, Ranking Keywords, and Spam Score. The example query against seo.domains returns DA 23, 143.2k linking root domains, 292 ranking keywords, and 1 percent Spam Score. The checker provides 3 free reports per day to unauthenticated users.

Domain Authority is a comparative predictor, not a ranking-position estimator.

Moz designed Domain Authority as a comparative tool. The score answers “which of these two domains is more likely to rank stronger” but does not predict the exact SERP position either domain will occupy on a specific query.

Practitioners reading Domain Authority as a ranking-position predictor misuse the metric; practitioners reading it as a comparative link-graph proxy use the metric correctly.

Domain Authority aggregates link-graph signals across the entire domain.

Moz publishes Domain Authority for an entire domain and Page Authority for individual URLs. Domain Authority aggregates link-graph signals across the whole domain. Page Authority isolates the signals for a single page.

Aged-domain buyers acquiring a domain check both: a domain with high Domain Authority and weak Page Authority on key URLs signals that the link equity concentrates on a small subset of pages.

Domain Authority is a third-party predictive proxy calculated by Moz.

Domain Authority is calculated by Moz on the Mozscape link graph. Moz operates an independent crawl of the public web and applies a machine-learning model to the link-graph signals it observes.

The score reflects Moz’s calculation methodology and Mozscape index state; it is independent of any search engine algorithm. The SEO industry standard treats DA as a cross-validated proxy alongside Domain Rating, Trust Flow, and Citation Flow.

Moz launched Domain Authority in 2012 as the first widely-adopted predictive ranking proxy

Moz introduced Domain Authority in 2012 as part of the Mozscape API release. The 2012 launch positioned Moz as the first major SEO toolset to publish a third-party predictive ranking score based on independent link-graph analysis.

Before 2012, SEO practitioners relied on Google’s own PageRank toolbar value, which Google deprecated and eventually removed from public visibility. Domain Authority filled the gap by offering a public-facing predictive score based on Moz’s proprietary crawl.

The 2012 launch filled the gap left by Google PageRank toolbar deprecation.

Google began deprecating the public PageRank toolbar value through 2011-2012, removing the only public-facing predictive ranking metric the SEO industry had.

Moz launched Domain Authority in early 2012 as the replacement: a Mozscape-calculated, publicly visible, predictive ranking proxy.

The market timing of the 2012 launch is one of the reasons Domain Authority became the industry-standard reference metric for the next decade.

Moz surfaces Domain Authority across Link Explorer, Pro tools, and the MozBar.

Domain Authority appears in Moz Link Explorer for single-domain audits, the Moz Pro tools suite for ongoing tracking, the MozBar browser extension for in-browser quick reads, and the Mozscape API for programmatic access.

The single metric is consumed across the Moz product line, contributing to its industry adoption among SEO teams using Moz as their primary toolset.

How Moz calculates Domain Authority: 6 algorithm inputs

Moz calculates Domain Authority through a machine-learning model trained on 6 factor categories. The training target is correlation with Google search rankings across a large sample of query-result pairs.

The model outputs a 1-to-100 score normalized to the logarithmic scale. The 6 factor categories below describe what the machine-learning model evaluates; the precise weighting is closed-source and updated through periodic model retrainings.

01Linking root domains
Number of unique domains with at least one referring link. Direct positive contribution weighted by referring-domain quality. The primary volume signal in the model.
02Total links
Total inbound link count across all referring URLs. Positive contribution with diminishing returns: 1,000 links from one referring domain count less than 100 links from 50 referring domains.
03MozRank (link equity flow)
Page-level link-equity propagation through the link graph, conceptually similar to the original Google PageRank algorithm. Captures recursive link-graph relationships beyond direct first-hop links.
04MozTrust (trust signal)
Distance from seed trusted domains in the link graph. Closer to trust seeds equals higher contribution. The trust-weighted complement to raw MozRank link-equity flow.
05Spam Score adjustment
Moz Spam Score detection across 27 documented spam-pattern flags. High Spam Score reduces Domain Authority below the raw link-count signal. The adjustment is what separates DA from pure-volume metrics on manipulative profiles.
06Anchor text distribution
Diversity and naturalness of incoming anchor text. Healthy distribution contributes positively; commercial-anchor over-optimisation patterns trigger algorithmic flags that lower the score.
Figure 2. The 6 documented factor categories in the Moz Domain Authority machine-learning model. The precise weighting is closed-source; Moz publishes the factor categories in its Learn Center documentation.
1
Collect the link-graph signals from the Mozscape index
Moz reads linking root domains, total links, MozRank link-equity flow, MozTrust trust distance, and anchor text distribution from its independent crawl of the public web.
2
Feed the 6 factor categories into the machine-learning model
The model was trained on a large sample of query-result pairs with documented ranking outcomes. It adjusts the factor weights to maximise correlation between the score and the observed ranking position. The precise weighting is closed-source.
3
Apply the Spam Score adjustment
The Moz Spam Score evaluates 27 documented spam-pattern flags. A high cumulative Spam Score pulls the score below what the raw link-count signal alone would suggest. This adjustment is what separates Domain Authority from pure-volume link metrics.
4
Normalise to the 1-to-100 logarithmic score
The model output is mapped to the logarithmic scale. Movement from DA 70 to DA 80 requires exponentially more high-quality referring domains than movement from DA 20 to DA 30.
Because the score is relative to every domain in the Mozscape index, a Moz model retraining can shift a domain’s score even when its own link profile has not changed.
Figure 3. How the Domain Authority score is built. The 6 documented factor categories feed the machine-learning model, the Spam Score adjustment is applied, and the result is normalised to the 1-to-100 logarithmic scale. Factor categories and the Spam Score 27-flag count are from the Moz Learn Center documentation.

The machine-learning model is trained on documented ranking outcomes.

Moz trains the Domain Authority machine-learning model on a sample of query-result pairs with documented ranking outcomes. The training target is correlation between the score and the observed ranking position.

The model adjusts factor weights to maximise correlation across the training set. The trained model then predicts new domains’ scores based on the same factor categories.

This is the methodology Cyrus Shepard has documented through his time at Moz and subsequent independent writing on link-metric correlation.

The Moz Spam Score adjustment penalizes link-manipulation patterns

The Moz Spam Score adjustment is the structural feature that separates Domain Authority from raw link-count metrics. Moz Spam Score evaluates 27 documented spam-pattern flags across the link profile and the on-page content.

High Spam Score reduces Domain Authority below the raw link-count signal would suggest. The adjustment limits the gap between honest authority and link-spam manipulation.

Moz Free SEO Tools dropdown menu showing Domain Analysis, Keyword Explorer, Link Explorer, Competitive Research, MozBar, and More Free SEO Tools options
Figure 4. The Moz Free SEO Tools menu surfaces Domain Analysis (where the Free Domain Authority Checker reads DA and Spam Score together), Link Explorer (full backlink data including Spam Score breakdown), Keyword Explorer, Competitive Research, and the MozBar browser extension. Spam Score evaluation typically starts from the Link Explorer entry in this menu.

The Spam Score evaluates 27 documented spam-pattern flags.

The 27 Spam Score flags cover patterns like high link count from low-quality referring domains, anchor text over-optimisation, thin content on the target domain, mismatched language signals, and unusual outbound link profiles.

Moz publishes the flag list in its Learn Center materials. Each flag adds to the cumulative Spam Score; high cumulative scores trigger the Domain Authority reduction.

Charles Floate documented Spam Score manipulation detection through PBN A/B testing.

Charles Floate documented the Spam Score detection mechanics through PBN A/B testing across 2018-2024.

The testing demonstrated that link-farm referrers and pattern-injected referring domains trigger the Spam Score flags and produce the Domain Authority reduction in observable timeframes.

The Spam Score adjustment is one of the strongest single signals that a domain’s link profile has been manipulated, which makes the Domain Authority-versus-raw-link-count gap a high-leverage pre-acquisition audit signal.

Domain Authority 2.0 retraining and the model update cadence

Moz launched Domain Authority 2.0 in March 2019 as a major model retraining that improved correlation with Google rankings. The 2.0 retraining incorporated improvements to the Spam Score adjustment, the anchor text distribution analysis, and the machine-learning architecture.

Subsequent model revisions are deployed without version-number announcements at irregular intervals. SEO practitioners observe Domain Authority shifts of 5-15 points when Moz refreshes the model, even when the underlying link profile has not changed.

2012
Domain Authority launches with the Mozscape API
Moz introduces Domain Authority as the first widely-adopted third-party predictive ranking proxy, filling the gap left by Google deprecating the public PageRank toolbar value.
5 Mar 2019
Domain Authority 2.0 retraining
The largest single update since the 2012 launch. The retraining improved the Spam Score adjustment, the anchor text distribution analysis, and the machine-learning architecture, with documented correlation improvements over the original DA 1.0 model.
2019 onward
Unversioned model refreshes 2-4 times per year
Moz refreshes the model at irregular intervals without version-number announcements. A refresh can shift a domain’s score 5-15 points with no change to its link profile. If Domain Rating from Ahrefs is stable while Domain Authority moved, the cause is a Moz model refresh.
Figure 6. The Domain Authority model timeline from the 2012 launch through the 5 March 2019 DA 2.0 retraining to the ongoing unversioned refresh cadence. All dates are from the Moz announcements record cited in this article.

The 2019 DA 2.0 retraining improved correlation with documented ranking outcomes.

Moz announced the Domain Authority 2.0 retraining on 5 March 2019 with documented correlation improvements over the original DA 1.0 model.

The 2.0 retraining was the largest single update to the Domain Authority methodology since the 2012 launch.

Practitioners running historical comparisons across the 2019 transition account for the model retraining as a separate variable from any underlying link-profile change.

Subsequent model revisions ship without version announcements.

Moz refreshes the Domain Authority model at irregular intervals without version-number announcements. The refreshes typically occur 2-4 times per year.

Practitioners observing unexpected Domain Authority shifts cross-check against the Moz announcements page and the Domain Rating value from Ahrefs: if DR is stable while DA shifted, the cause is a Moz model refresh, not a link-profile change.

Domain Authority compared to Domain Rating: scope and methodology differences

Domain Authority and Domain Rating both score domains on a 0-to-100 logarithmic scale but the methodologies, index sources, and refresh cadences differ. Domain Authority (Moz, 2012) incorporates the Spam Score adjustment that Domain Rating does not apply.

Domain Rating (Ahrefs, 2014) uses recursive referring-domain weighting that Domain Authority handles differently through its machine-learning model. The two metrics agree on link-profile-strong domains and diverge on the link-manipulation edge cases.

DimensionDomain Authority (Moz, 2012)Domain Rating (Ahrefs, 2014)
Calculation methodologyMachine-learning model on 6 factor categoriesRecursive referring-domain weighting
Spam-pattern adjustmentMoz Spam Score reduces DA on manipulative profilesNone applied at metric level
Index sourceMozscape link indexAhrefs proprietary link graph (2011-launched)
Refresh cadenceRoughly monthly with model recalibration eventsContinuous near-real-time
URL-level companion metricPage Authority (PA)URL Rating (UR)
Strongest single use caseSpam-pattern detection on candidate acquisitionsNear-real-time link-profile change tracking
Figure 5. Side-by-side comparison of Domain Authority and Domain Rating methodology dimensions. The full comparison guide covers when each metric surfaces signals the other misses.

Wide DA-DR gaps signal link-profile manipulation.

A domain with significantly higher DR than DA shows one of the strongest single signals of link-profile manipulation that aged-domain buyers can audit pre-acquisition.

The Domain Rating calculation does not apply a spam-pattern penalty equivalent to the Moz Spam Score. The DA-DR gap measurement is the operational audit shortcut that practitioners apply on every candidate domain.

For the full comparison framework see Domain Authority vs Domain Rating.

Domain Authority for aged-domain evaluation: pre-acquisition link-profile screening

Aged-domain buyers read Domain Authority as one input in a multi-metric screening framework alongside Domain Rating, Trust Flow, and Citation Flow. Single-metric reliance on Domain Authority produces blind spots that the multi-metric cross-check surfaces.

The Spam Score adjustment built into Domain Authority makes DA the strongest single-metric signal for link-manipulation patterns; the multi-metric cross-check completes the audit.

Domain Authority is the strongest single-metric signal for link-manipulation patterns.

The Spam Score adjustment built into Domain Authority makes DA more sensitive to link-spam manipulation than Domain Rating, Trust Flow, or Citation Flow taken alone.

Practitioners running pre-acquisition audits start with Domain Authority and the DA-DR gap, then cross-check against Trust Flow and Citation Flow for the multi-angle confirmation.

The screening framework is documented across the Backlinko link-correlation studies and the Authority Hacker affiliate-site case publications.

Aged-domain Domain Authority targets vary by niche competition level.

A good aged-domain Domain Authority depends on the niche competition.

Finance and insurance verticals require DA 70+ for competitive entry; mid-competition niches like SaaS or technology accept DA 50+; long-tail niche-content sites compete at DA 30-50.

For the niche-relative benchmarks see Domain Rating scoring benchmarks which applies the same SERP-relative read across the metric family.

Cyrus Shepard has documented Domain Authority correlation analyses across enterprise SEO programmes.

Cyrus Shepard, formerly at Moz and now at Zyppy, has documented Domain Authority correlation analyses across enterprise SEO programmes through 2015-2024.

The analyses demonstrate that Domain Authority predicts ranking probability at scale across query populations but does not predict specific SERP positions on individual queries.

Practitioners applying the comparative-probability interpretation make correct acquisition decisions; practitioners applying a ranking-position interpretation misuse the metric.

5 frequently asked questions about Domain Authority

The 5 top Domain Authority questions practitioners and aged-domain buyers ask. Answers reflect the SEO Domains analytical position alongside documented industry-leader interpretation.

Q1What is a good Domain Authority score?

A good Domain Authority depends on the niche. As a general guide: DA 30-50 indicates established link-building presence, DA 50-70 indicates well-established authority sites, DA 70-100 indicates top-tier media and enterprise domains.

Niche-relative comparison against top-3 SERP competitors is more useful than the absolute number for acquisition decisions. Finance and insurance verticals require DA 70+ for competitive entry; niche content sites compete at DA 30-50.

Q2How is Domain Authority calculated?

Moz calculates Domain Authority through a machine-learning model trained on 6 factor categories: linking root domains, total links, MozRank, MozTrust, Spam Score adjustment, and anchor text distribution.

The model is trained on documented ranking outcomes across a large query-result sample. The 0-to-100 logarithmic scale normalises the output. Moz refreshes the model 2-4 times per year at irregular intervals.

Q3How can I check Domain Authority for free?

Moz Link Explorer provides 10 free Domain Authority checks per month for unauthenticated users. The Moz Pro tools suite includes unlimited DA checks for subscribers. The MozBar browser extension surfaces DA on every page the user visits.

Multiple third-party tools also expose DA via the Mozscape API. For aged-domain acquisition decisions, SEO Domains catalogue listings show DA alongside DR, TF, and CF at the listing level.

Q4Why does my Domain Authority go up and down?

Three reasons typically cause Domain Authority shifts. First: link-profile change (gained or lost referring domains). Second: Moz model retraining (4 times per year typical, can shift scores 5-15 points without any link-profile change).

Third: relative-position shift (other domains gained authority, pushing the relative position downward). Cross-check against Domain Rating from Ahrefs: if DR is stable while DA moved, the cause is a Moz model refresh.

Q5Is Domain Authority worth the screening cost for aged-domain acquisition?

Domain Authority is the strongest single-metric signal for link-manipulation patterns because of the Spam Score adjustment.

For aged-domain acquisition, DA is one of four cross-validated metrics (alongside DR, TF, CF) the SEO Domains catalogue applies at inventory ingestion. The multi-metric screen filters spam-pattern domains before listings reach buyers.

The cost of running the manual multi-metric audit per candidate exceeds the catalogue’s pre-screening cost at typical acquisition budgets.

How aged-domain marketplaces use Domain Authority in inventory screening

Curated aged-domain marketplaces use Domain Authority, Domain Rating, Trust Flow, and Citation Flow together in inventory screening. Multi-metric screening filters out the link-spam patterns that single-metric reliance would miss.

The SEO Domains catalogue applies the multi-metric screen at inventory ingestion, before listings reach buyers, with the DA-DR gap as the primary spam-pattern detection signal.

Screening dimensionMetrics referencedFilter logic
Link-spam pattern detectionDA Spam Score + DA-DR gap + TF:CF ratioWide DA-DR gap with high Spam Score or low TF:CF ratio excludes the listing
Niche-relative authority baselineDA + niche-median top-10 SERP DAListings below niche-median DA excluded from competitive-entry inventory
Refresh-history stabilityDA historical trajectory across 12-24 monthsVolatile DA history (sudden spikes or drops) flags the listing for manual review
Trust signal strengthDA MozTrust component + Trust Flow valueLow trust signals exclude listings even with strong raw link counts
Figure 7. The 4 screening dimensions where Domain Authority is read alongside Domain Rating, Spam Score, Trust Flow, and Citation Flow in catalogue construction. Multi-metric screening produces stronger pre-acquisition signal than any single-metric filter.
Moz
Domain Authority
Carries the Spam Score adjustment, so it is the strongest single signal for link-manipulation patterns. It never decides a listing alone.
Ahrefs
Domain Rating
Recursive referring-domain strength with no spam penalty at the metric level. A wide DA-DR gap flags the manipulation Domain Rating misses.
Majestic
Trust Flow
Weights link quality by distance from a curated seed set of trusted domains. A low Trust Flow excludes a listing even on strong raw link counts.
Majestic
Citation Flow
Counts raw link strength without trust weighting. The Trust Flow to Citation Flow ratio adds a fourth independent angle on the link profile.
Curated catalogue multi-metric screen
The four metric angles cross-validate at inventory ingestion. A listing reaches buyers only on the consensus signal across Domain Authority, Domain Rating, Trust Flow, and Citation Flow, never on a single score. The SEO Domains catalogue spans $100 entry-level domains through $1.5 million premium acquisitions.
Figure 8. Domain Authority is one of four cross-validated angles feeding the curated catalogue’s multi-metric inheritance screen. A single score is never the arbiter; the screen resolves on consensus across the four metrics at inventory ingestion.

The catalogue construction process applies the multi-metric DA screen at inventory ingestion.

SEO Domains catalogue listings surface Domain Authority, Domain Rating, Trust Flow, Citation Flow, and Majestic Topical Trust Flow at the listing level. The catalogue construction process applies the multi-metric screen during inventory ingestion.

The buyer reviews pre-screened listings instead of running the multi-metric audit manually across raw inventory. Each listing shows the consensus quality signal across the four metrics.

ICANN-accredited transfer preserves Domain Authority attribution through ownership change.

SEO Domains operates with ICANN-accredited registrar transfer protocols. The Mozscape crawler records link profiles by URL and propagates the inherited Domain Authority value to the new owner without re-validation.

Buyers acquire the established Domain Authority along with the link profile that produced it, validated by the catalogue’s 7-vector inheritance screen at inventory ingestion.

Hristo Bogdanov, Head of SEO at SEO Domains

Hristo Bogdanov

Head of SEO @ SEO Domains · CEO & Co-founder of SEO.bo

Hristo has spent 15+ years building aged-domain acquisition workflows for SEO professionals, brand owners, and domain investors.

He leads SEO at the SEO Domains marketplace, which operates a 220,000+ curated catalogue from $100 entry-level domains through $1.5 million premium acquisitions, penalty-screened across the catalogue, with Managed Account expert support for premium-tier clients.

· Last reviewed