Domain Authority is a score from 1 to 100 that Moz uses to estimate how likely a site is to rank on Google compared with other sites. It is not something Google uses to order its results, and Moz says so itself on the page that documents the metric.
On its own, the number tells you nothing. It only makes sense as a comparison: your site's DA against the sites already occupying the SERP you want to break into, which is the same logic you should apply to keyword difficulty. A 28 can be enough in an industrial niche and nowhere near enough in a sector where competitors sit at 60.
Who calculates it and with what data
Domain Authority was created by Moz and is calculated with a machine learning model trained to predict which domains appear in Google's results. The data comes from Moz's link index, Link Explorer, and the heaviest weight goes to the number and quality of domains linking to the site. The current version of the model, Domain Authority 2.0, dates from 2019.
Two details in Moz's documentation change how the score should be read. First, every new site starts at 1: not zero, and not a sector average. Second, the scale is relative and compressed at the top. Moz gives its own example. If Facebook gained a billion new links, everyone else's DA would drop, because places at the top end are contested. That is why moving from 20 to 30 is far easier than moving from 70 to 80, and it has nothing to do with how hard you have worked.
Why the score drops when you haven't touched anything
Moz lists four reasons DA changes, and only one depends on your site:
- the Moz index is updated and contains more or fewer linking domains than the previous crawl;
- the site's backlink profile grows or loses links;
- other sites' scores change, especially the very high ones, and the scale is recalibrated;
- the calculation model changes.
Three of the four are outside your control. A two-point drop from one month to the next, with no lost links, is usually a recalibration of the scale. Treating it as a problem to fix means chasing the thermometer instead of the fever.
DA, DR, Authority Score and Trust Flow don't measure the same thing
| Metric | Who produces it | What it is based on | Scale | Updates |
|---|---|---|---|---|
| Domain Authority (DA) | Moz | Machine learning model on the Link Explorer index, dominated by inbound links | 1-100, logarithmic | With each Moz index update |
| Domain Rating (DR) | Ahrefs | Backlink profile only; nofollow links pass no DR | 0-100, logarithmic | Continuous |
| Authority Score (AS) | Semrush | Eight weighted factors: link power, organic traffic and six manipulation signals | 1-100 | Every two weeks |
| Trust Flow (TF) | Majestic | Proximity in the link graph to a set of manually reviewed trusted seed sites; paired with Citation Flow, which measures link volume and power | 0-100 | With each Majestic index update |
The differences are not nuances. Semrush calculates Authority Score from eight weighted factors: two are link power and organic traffic, and the other six are there to catch manipulation. Among them is a stated threshold. If more than 90% of the domains linking to a site do so without the nofollow attribute, Semrush treats the ratio as almost always artificial. Another check looks at how many linking domains share the same IP address or network, which is the classic way to spot a blog network built to sell links.
Ahrefs does not disclose the Domain Rating formula, but it states three rules: nofollow links pass no DR, a domain linking to a thousand sites passes on less than one linking to ten, and the scale is logarithmic here too.
Majestic splits the question in two. Citation Flow measures how much link power a site carries; Trust Flow measures how close its links sit to a list of trusted seed sites that Majestic reviews by hand. Neither is an authority score in the Moz sense, but the gap between them is informative: a site with a high Citation Flow and a low Trust Flow has plenty of links and few from anywhere trustworthy. If you are assessing a site, reading the two numbers together tells you more than any single score.
What Google says it does have
Google does not publish any domain authority score, and none of the four metrics feeds into its systems. Its spam policies do contain one sentence worth reading in full: "Google generally applies a presumption that individual pages (including new pages) match the overall quality of other pages on the domain".
It is not a score, it is a presumption. A new page inherits the trust earned by the rest of the site until Google has data of its own about that page. That is the practical reason publishing on an established domain pays off sooner, and it is also the reason that mechanism gets sold.
The same spam policies say what happens to those who try to exploit it. There is a site reputation abuse policy, and in August 2026 Google changed how it applies in the European Economic Area: inside the EEA, non-compliant sections no longer receive a manual action; they are treated as separate from the main domain and ranked on their own. The UK is not in the EEA, so for a UK site the policy text still describes manual actions for non-compliant pages. Inside the EEA, a subfolder rented out to a third party no longer drags the rest of the site down, but it also stops inheriting anything. The domain presumption, which is what everyone loosely calls authority, switches off in that case.
It is worth remembering, again from Google's documentation, that E-E-A-T is not a ranking factor either: the SEO Starter Guide answers "no" to the direct question. A concept can guide the work without being a number inside the algorithm, and that applies to DA as much as to E-E-A-T.
Where Domain Authority becomes a problem
The guest post market has organised itself around these metrics. Price lists read "DA 40+", "DR 50+", "TF 30", and the price rises with the score. Yet the transaction described in those lists is named in full in Google's spam policies, under link spam: "Exchanging money for links, or posts that contain links". The text leaves no room for doubt, and there is no DA threshold that makes it acceptable.
There is a legitimate way, and it is on the same page: a paid link must be marked with rel="sponsored" or rel="nofollow". At that point it passes no ranking value, which is exactly why hardly anyone does it. Anyone selling visibility off a price list is selling the part Google asks you to cancel out. That is where the line between link building and black hat SEO runs.
A second effect shows up with expired domains. Buying a domain with a link history and reusing it for unrelated content is, again under the spam policies, expired domain abuse. Third-party metrics stay high for months after the purchase, because they look at the old links and do not know what has happened to the site. This is the case where a high score is the symptom of the problem, not a guarantee.
What to measure instead of the score
On an industrial project managed by Visilay, a product query moved from position 88 to 2 and the client's share of voice reached 25.55%, ahead of Amazon: the figures and the method are in the case study on SEO for manufacturing. Over those months Domain Authority was not the indicator that showed it was working, and it caught up later.
Three numbers tell you more, and you already have them:
- rankings for the queries that bring enquiries, not the ones that bring visits;
- share of voice across your market's keyword basket, which shows how much space you hold against your real competitors;
- contact requests attributed to organic search.
The first two come from Search Console and a rank tracker, the third from your CRM. These are the SEO KPIs that hold up in front of a sales director, and none of them recalibrates itself when a tool updates its index. If you are starting now, an SEO audit tells you which of these numbers you already have and which you don't, while the ranking factors Google documents remain the reference for what the algorithm actually looks at. If someone needs to carry the work forward, our SEO services start from exactly here.
There is one use of Domain Authority almost nobody makes, and that is reading it backwards. Semrush has a label called "lacks organic traffic": it flags domains with a strong link profile and low traffic compared with sites with a similar profile. That combination describes a site that has piled up links without Google sending it visitors, and in a guest post price list it is the most useful information you can get. The high score attracts buyers; the gap between score and traffic tells you how it was built.
Domain Authority FAQs
There is no threshold that applies to everyone: Moz states that DA is a comparative metric. The benchmark is the set of sites occupying the SERP you care about. If competitors sit between 35 and 45, a 40 puts you in the game; the same 40 in a sector where everyone is at 70 is not enough.
No. In its documentation Moz writes that Domain Authority is not a Google ranking factor and has no effect on the SERPs. It is a prediction Moz builds from its own link data, not a value Google reads.
Moz gives four causes: updates to its own link index, changes in the site's backlink profile, changes in other domains' scores that recalibrate the scale, and changes to the calculation model. Only the second depends on your site, so a drop of a few points often signals no real problem.
With Moz's Link Explorer, which requires a free account, or with MozBar, the browser extension that shows DA and Page Authority directly in the search results. To compare many domains in bulk you need the Moz API or a tool that uses it.
It depends on where you start, because the scale is logarithmic: going from 20 to 30 takes far less work than going from 70 to 80. The score also only moves when Moz updates its index, so a link earned today does not show up straight away.
