Domain Authority vs Domain Rating vs Authority Score: which number should you actually trust?

Domain Authority is Moz's, Domain Rating is Ahrefs', Authority Score is Semrush's, and none of the three is a Google ranking factor. They disagree because each vendor crawls its own index of the web. They disagree less than people assume: a peer reviewed study of 3,151 domains found every pairwise correlation between them above 0.9, meaning they rank sites in nearly the same order. Which one you check matters far less than what you do next.

That last sentence is the part worth sitting with, because most of the argument about these metrics is an argument about nothing. Below is what each score actually measures, how far apart they really get, and the specific situations where the number still earns its place in a decision.

What each authority score actually measures

These are not competing implementations of one standard. They are separate products that happen to share a 0 to 100 shape, and two of them are not even trying to answer the same question. Domain Authority is a prediction: Moz's own definition is a metric that "predicts how well a domain will rank in Google based on a machine learning algorithm of link metrics." Domain Rating is a description: Ahrefs defines it as the strength of a site's backlink profile, on a logarithmic scale, with nothing about rankings in it at all.

Semrush's Authority Score is the odd one out and the most interesting of the three. It blends link power with estimated organic traffic and a set of spam checks, which means a site with modest links but real search visibility can score higher than its backlink profile alone would justify. That is either a feature or a contamination depending on what you wanted the number for. If you are evaluating a link prospect, blending in traffic is helpful. If you are trying to isolate link strength, it is noise.

Majestic sits outside the comparison entirely by splitting the idea in two. Citation Flow counts links, Trust Flow weights them by how close the linking sites sit to a manually reviewed set of trusted seeds. Reading them as a pair tells you something a single blended score hides: a domain with Citation Flow 45 and Trust Flow 8 has a lot of links from places nobody trusts, which is a profile worth avoiding rather than buying.

How far apart do the scores actually get?

This is the part of the debate that has an actual answer, and almost nobody quotes it.

In 2023, Danilo Reyes-Lillo, Alejandro Morales-Vargas and Cristofol Rovira published "Reliability of domain authority scores calculated by Moz, Semrush, and Ahrefs" in Profesional de la informacion, volume 32, issue 4. They ran 61 neutral queries, collected 16,937 results across 3,151 unique domains, and compared the three vendors' scores for every one of them.

Every pairwise Spearman correlation came out above 0.9. In practical terms the three tools put the same sites in nearly the same order. If Moz says site A is stronger than site B, Ahrefs and Semrush usually agree, even when the raw numbers look nothing alike. The famous disagreements are mostly scale artifacts: DR is logarithmic and DA is not, so the same site can plausibly read DR 62 and DA 41 without either vendor being wrong.

The study's second finding is the one that should change behavior. All three scores correlate only weakly with actual ranking positions. They agree with each other and they do not agree with Google. That is exactly what you would expect from three estimates built on backlinks when backlinks are one input among many, and it is the reason a page with DA 20 routinely outranks a page with DA 60 on a specific query.

Worth being precise about one thing: Majestic was not among the three tested. Trust Flow and Citation Flow are outside the scope of that finding, so do not extend the 0.9 result to them.

Why the numbers disagree at all

Three reasons, in order of how much they matter.

Different indexes. Every vendor crawls the web itself and owns whatever it finds. A link that Ahrefs discovered and Moz has not simply does not exist inside Moz's calculation. Nobody has the whole web, and the missing slice is different for each of them. This accounts for most of the gap, and it is worse on smaller and newer sites where a handful of links moves the score several points.

Different formulas. Ahrefs excludes nofollow links from DR. Moz's model does not work that way. Semrush folds in traffic and spam checks that neither of the others considers. These are genuine methodological choices, not bugs.

Different update cadences. The scores refresh on their own schedules, so a site that earned twenty links last week may show the gain in one tool and not yet in another. If you are comparing a screenshot from March against a live number in August, you are measuring the calendar.

Is domain authority a Google ranking factor?

No, and this is not a close call. Domain Authority is a Moz product. Google has never used it, has no access to it, and has said repeatedly that there is no single site-wide authority score in its ranking systems. The same applies to Domain Rating and Authority Score.

The confusion is understandable, because DA does correlate with rankings. But it correlates the way umbrella sales correlate with wet pavement. Both DA and rankings are downstream of the same thing, which is links, so they move together without one causing the other. Buying a higher DA, in the literal sense of acquiring links chosen to raise the number, is the clearest way to learn this: the score goes up and the rankings do not follow.

So when is the number actually useful?

Three situations, and they are all comparative rather than absolute.

Qualifying a link prospect at speed. When you are working through a list of two hundred sites, a first-pass filter that removes the obviously worthless ones saves real time. Pair it with Trust Flow if you have Majestic access, because a low Trust Flow beside a high Citation Flow is the signal that catches link farms.

Reading the competitive gap on a keyword. Not your score in isolation, but your score against the sites currently on page one for the query you want. This is the only version of "what is a good domain authority score" that has an answer. If the page-one average is 25 and you are at 18, the content is what decides it. If the average is 70, you need a different keyword before you need a better page.

Tracking direction over time. A score that has climbed steadily for six months tells you the link acquisition is working, even though the absolute value means little. Direction is more trustworthy than level, because your own index coverage stays roughly constant.

Outside those three, the number is decoration. It is the thing people screenshot for client reports precisely because it is easy to explain, which is not the same as being informative. If the report needs a number that reflects reality, the impressions and average position sitting in Search Console are first-party, free, and about your site rather than a vendor's estimate of it. Teams that report across paid and organic together usually end up pulling those first-party numbers into a single dashboard rather than pasting three vendor scores into a slide.

What to do instead of checking your DA again

The honest summary of the research is that these scores are a shared vocabulary, not an instrument. They let two SEOs describe a site to each other in one word. That is genuinely useful and it is roughly the whole value.

Meanwhile the things that decide whether a specific page ranks are all page-level and all under your control: whether the title actually targets the query, whether the headings cover the subtopics the query implies, whether the page answers the question in the first paragraph or three screens down, whether the internal links pointing at it use words a searcher would type. None of that shows up in a domain-level score, and all of it can be fixed this afternoon.

If you want to see the browser extensions that put these numbers on your screen while you browse, and which of them gate the metrics behind a paid plan, we compared seven of them on the MozBar alternative page. For the two suites behind DA and AS, the pricing and feature comparison is on Moz Pro vs Semrush, and the wider set of Moz competitors on the Moz alternative page. Majestic's Trust Flow and Citation Flow, and why a link-only index is a different purchase, are covered in Majestic SEO vs Semrush.

Frequently asked questions

What is the difference between domain authority and domain rating?

Domain Authority is Moz's 1 to 100 machine learning prediction of how well a domain will rank, built from link metrics. Domain Rating is Ahrefs' 0 to 100 logarithmic measure of backlink profile strength. DA tries to predict ranking, DR only describes links. Different vendors, different indexes, no shared scale.

Is domain authority and authority score the same?

No. Domain Authority belongs to Moz and is calculated purely from link data. Authority Score belongs to Semrush and blends three inputs: link power, estimated organic traffic and spam signals. Because Authority Score includes traffic, a site can score well on one and poorly on the other.

Is domain authority a Google ranking factor?

No. Domain Authority is a Moz metric and Google has never used it, and the same is true of Domain Rating and Authority Score. Google representatives have said repeatedly there is no site-wide authority score in the algorithm. These are third-party estimates of something Google does not publish.

What is a good domain authority score?

There is no absolute threshold, only a relative one: the useful comparison is against the sites already ranking for your keyword. If page one averages DA 40 and you sit at 18, that gap is meaningful. If page one averages 25, DA 18 is close enough that content quality decides the outcome instead.

What is the most reliable domain authority checker?

Whichever one the people you talk to already use, because the research says the choice barely matters. A 2023 peer reviewed study of 3,151 domains found every pairwise correlation between the Moz, Semrush and Ahrefs scores above 0.9 Spearman. They rank sites in nearly the same order.

What is the difference between domain authority and page authority?

Scope. Domain Authority predicts how the whole domain will rank; Page Authority predicts how one specific URL will rank. Both are Moz metrics on a 1 to 100 scale built from the same link model. Page Authority is the more useful of the two, because Google ranks pages rather than domains.

What is the difference between domain authority and topical authority?

Domain Authority is a vendor number derived from backlinks. Topical authority is not a metric at all, it is a description of covering a subject thoroughly enough that a search engine treats you as a credible source on it. One is measured by a third party, the other is built by publishing.

Who owns domain authority?

Moz owns Domain Authority, and it is a trademark of theirs rather than an industry standard. Ahrefs owns Domain Rating, Semrush owns Authority Score, and Majestic owns Trust Flow and Citation Flow. Each is calculated from that vendor's own crawl of the web, which is why they never quite agree.

Why do Moz and Semrush give very different results?

Mostly because they crawl different subsets of the web and hold different link indexes, and partly because they measure different things. A link Semrush found and Moz did not simply does not exist in Moz's calculation. The gap widens on smaller sites, where a handful of links moves the score a lot.

Should I use domain authority to judge a backlink?

As a first-pass filter, yes, and as a decision, no. A high score tells you a domain has strong links overall, not that a link from it will help your page. Relevance to your topic, where the link sits on the page and whether the page gets any traffic all matter more than the number.

The short version

Domain Authority, Domain Rating and Authority Score are three vendors' estimates of the same rough idea, built from three different crawls of the web. They agree with each other above 0.9 correlation and agree with Google's actual rankings only weakly. Pick whichever one your team already has access to, use it comparatively rather than absolutely, and stop treating a change of three points as an event.

Then go and fix a page. That part still works.