What Is Domain Authority? The Honest, No-Spin Answer

Every SEO tool shows you a number next to a domain and calls it "authority." Almost nobody explains what that number actually is, who made it, or why Google has nothing to do with it. Here's the honest version.

What Is Domain Authority?

Domain authority (DA) is a third-party score, built by SEO tool vendors like Moz, that predicts how likely a website is to rank in search results on a 1-100 scale. It isn't a fixed truth handed down by Google; it's a proxy for overall site strength, calculated mostly from link data. Google does not build it, publish it, or use it to rank anything.

Think of DA the way you'd think of a credit score from a private agency: useful for comparison, occasionally wrong, and never the thing the actual decision-maker — Google, in this case — consults before acting.

Most people meet DA inside a browser toolbar or a tool like Moz, Ahrefs, or Semrush, sitting next to a URL like a badge of honor. That badge is easy to obsess over. It's also easy to misread as something it was never built to be.

DA also isn't a measure of content quality, trustworthiness, or expertise, even though people often use it as shorthand for all three. A site can have a high DA from years of link building and still publish thin, outdated, or inaccurate pages. The score describes link muscle, not what the site actually says.

Who Created Domain Authority — and Who Didn't?

Moz created domain authority. Google did not — and Google has repeatedly said it plays no role in how pages rank.

That single fact resolves most of the confusion around DA. It's a vendor metric — a modeled estimate one company built to help its customers gauge competitive strength — not a leaked or official Google signal.

Moz isn't alone in building one of these scores. Every major SEO platform ships its own version:

• Moz — Domain Authority (DA), scored 1-100
• Ahrefs — Domain Rating (DR), scored 1-100
• Semrush — Authority Score (AS), scored 1-100

They're built differently, weighted differently, and updated on different schedules, so it's common to see the same site score, say, a 40 on one tool and a 55 on another — both "correct" by their own vendor's math, and both disagreeing with each other. That's not a bug. It's proof none of them measures some single, objective truth. Each is its own approximation of the same fuzzy idea: overall site strength.

How Is Domain Authority Calculated?

DA is calculated mainly from a site's backlink profile — the number and quality of other sites linking to it — run through a machine-learning model trained to correlate with search rankings.

The logic underneath is decades old: Google's original PageRank treated links as votes, and DA borrows that same idea — a link from a well-linked site should count for more than a link from an obscure one. Moz layers a trained model on top of that logic instead of using Google's actual algorithm, which nobody outside Google has access to.

The rough inputs vendors have disclosed over the years:

• Linking root domains — how many distinct websites link to you, since raw link counts are easy to inflate
• Link quality — a link from an established, relevant site counts more than one from a spam directory
• Total link profile size — the number of links pointing to the whole domain, not just one page

None of this is disclosed in full. Moz doesn't publish its exact formula, and it updates the model periodically — which is why your DA can shift even when you haven't touched your site. A competitor's link profile moved, or the model itself got recalibrated.

That opacity is worth sitting with. You're optimizing toward a number whose exact formula you can't see, issued by a company with no visibility into Google's actual index. Useful — but not gospel.

How to Use DA Correctly

DA is valid for exactly one job: comparing your link profile's rough strength against competitors, over time. It is not valid as a target, a KPI, or a stand-in for rankings and traffic.

The mistake almost every team makes is treating DA like a scoreboard — chasing the number itself instead of watching what the number is supposed to represent.

DA use case	Valid?	Better metric
Comparing your site to 3-5 direct competitors	Yes	—
Tracking your own trend over 6-12 months	Yes, directionally	—
Setting a "hit DA 50" team goal	No	Organic traffic, ranking positions
Judging one page's ranking potential	No	Page-level backlinks, content relevance
Vetting a guest post or link opportunity	Partly — one input among several	Referral traffic, topical relevance, real audience
Predicting AI answer engine citations	No	See the next section

In practice, that means pulling up DA once a quarter, next to the 3-5 competitors you actually compete with for rankings — not the biggest name in your industry, but the sites you're really up against in the search results you care about. Watch the trend line, not the single number.

If you want a second opinion on your own number, or you're deciding which tool to standardize on, our Moz alternatives comparison walks through how the major vendors' scores diverge and when each one is actually worth trusting.

What Happens to Domain Authority in the AI Era?

AI engines don't check a DA score before citing a source. ChatGPT, Perplexity, Gemini, and Google's AI Overviews have no API call to Moz running in the background before they decide what to reference.

What these engines actually weigh looks different from classic SEO signals:

• Entity consistency — does your brand show up the same way, with the same facts, across many independent sources?
• Corroboration — do other sites, forums, and publications describe you the same way, in their own words?
• Directness — does your page actually answer the question being asked, clearly and specifically?

DA correlates loosely with these signals, because sites that earn real backlinks over years also tend to be the sites people talk about accurately elsewhere. But the correlation is loose, not causal. A high-DA site with thin, generic content can still get passed over by an AI engine in favor of a lower-DA site with a sharper, more specific answer.

This is where the DA conversation usually stalls, because there's no equivalent single number for AI visibility — no score you can screenshot. What exists instead is actual testing: running your brand's real buyer questions through ChatGPT, Perplexity, and Gemini, and recording whether you show up.

If you actually want to know whether AI engines cite you — instead of guessing from a DA badge — that means measuring AI visibility directly: checking what each engine returns when someone asks a buying question in your category.

What Actually Builds Authority?

Real authority — the kind that moves DA and gets you cited by AI engines — comes from three unglamorous things.

1. Genuine expertise, demonstrated on the page. Not claimed — demonstrated, with specifics a generalist couldn't write.
2. Earned links and mentions. Coverage, citations, and backlinks other sites give you voluntarily because your content was worth referencing — the core of off-page SEO.
3. Consistency over time. The same facts about your brand, repeated across enough independent sources, build the kind of trust neither a vendor's model nor an AI engine can shortcut.

None of that is exciting advice. It's also the only kind that survives the next time a vendor recalibrates its model, or the next time an AI engine ships a new retrieval method. Work that makes you genuinely worth citing tends to hold up. Score-chasing doesn't.

Where Does AEOeye Fit In?

DA tells you how link-strong you probably are, according to one vendor's model. It has never told you whether ChatGPT, Perplexity, or Google's AI Overview would actually mention your brand when someone asks a buying question in your category.

That's a different audit — and it's the one AEOeye runs, checking whether AI search engines actually recommend you by name, and showing you exactly where the gap is if they don't.