Check this page with an assistantOpens a chat asking it to summarise this article and name the evidence behind each claim.
Claude opens with the prompt on your clipboard: Anthropic does not support prefilled prompts on the web, and we would rather copy it than ship a button that drops it.
Whose Score Is It?
Domain Authority belongs to Moz. Domain Rating belongs to Ahrefs. Authority Score belongs to Semrush. Each is a model built by a tool company to help its customers compare domains. None of them is a Google metric, and Google has stated repeatedly that they play no part in crawling, indexing, or ranking.
That should end the argument and does not, because the scores are presented on a nought-to-hundred scale that looks like a grade. A number out of a hundred implies an authority conferring it. People reasonably assume that authority is the search engine, and every conversation downstream inherits the mistake.
The vendors are not being dishonest about this; the documentation says what the scores are. The confusion is downstream, where a score built for comparison gets quoted as though it were a measurement of how Google sees you, most often by someone selling links priced against it.
Why the Correlation Is Real and Misleading
Sites with higher scores do tend to rank better, which is exactly what makes the misunderstanding durable. The explanation is straightforward: the scores are largely models of backlink strength, and links are something Google genuinely uses. Both track the same underlying thing.
The distinction matters because it determines whether improving the score does anything. If the score is a thermometer, raising the reading by holding a match to it does not warm the room. Buying links to lift a score produces a higher number and no ranking change, because Google was never reading the number and largely discounts the links that moved it.
It also explains why the correlation is loose at page level. These are domain-level models, and Google assesses pages. A strong domain with a weak page loses to a weak domain with a strong page more often than the scores suggest, which is the effect described in page-level authority.
Legitimate Uses, and the Misuses
These scores are genuinely useful within their design. The failures happen when a comparison tool is used as a prediction tool, which is most of the time.
| Use | Sound? | Why |
|---|---|---|
| Comparing two similar domains | Yes | What the model was built for |
| Evaluating a domain before acquiring it | Yes, alongside real checks | One input among several, per SEO due diligence |
| Tracking your own profile over time | Yes, as a trend | Consistent methodology makes the direction meaningful |
| Predicting whether a page will rank | No | Domain-level model, page-level question, wrong input |
| Setting a minimum threshold for links | No | Cannot see relevance, which matters more |
| Reporting it to a client as performance | No | A vendor score moving is not a business outcome |
| Explaining a traffic decline | No | Google was not reading it before or after |
The threshold row is where real money gets wasted. Filtering outreach by a minimum score excludes relevant sites in your field with modest profiles while including irrelevant general sites that score well, which is close to the opposite of what you want. It is also how link sellers price inventory, which is worth noticing.
They Are Cheap to Inflate
Because these scores model backlinks, they respond to link volume regardless of whether that volume represents anything genuine. That makes them manipulable in a way the thing they proxy for is not, and there is an entire market built on the gap.
Expired domains are sold on exactly this basis: a domain with an inflated score, presented as an asset, where the score reflects links accumulated in a previous life that Google may well discount entirely. A buyer using the score as the valuation is paying for a number nobody at Google reads. The checks that actually matter are in SEO due diligence.
The same gap explains why link marketplaces quote these scores. It is the only comparable number available across wildly different sites, it is manipulable by the seller, and the buyer treats it as a quality guarantee. Every part of that arrangement works better for the seller, and the underlying trade is examined in how to get backlinks.
What to Use Instead
For the question these scores get used to answer, is this site worth a link from, the better inputs are relevance, whether it has a real audience, and whether a person would find your mention useful there. None is a number, which is why the number keeps winning.
For measuring your own progress, use outcomes rather than proxies: referring domains earned, impressions and clicks from Search Console, and revenue attributable to organic. Those are harder to move and they are what you actually care about, which is the whole argument in the SEO reporting framework.
And if you want to know whether you are gaining standing in a category rather than in a tool, the assistant-era version of that question is measurable by sampling rather than by score, as set out in share of model. It has the same property as everything else worth measuring here: it is slower, less satisfying, and about the real world.
Questions About Domain Authority
- Is domain authority a Google ranking factor?
No. Domain Authority is a score built by Moz, Domain Rating is built by Ahrefs, and equivalents exist at other vendors. Google representatives have said repeatedly that these third-party scores play no part in crawling, indexing, or ranking. Google does not read them and has never used them.
- Then why does a higher score correlate with better rankings?
Because the scores are largely models of backlink strength, and links are something Google genuinely uses. The score and the rankings share a common cause rather than one producing the other. Improving the score by any route that does not involve genuinely earning references produces no ranking benefit at all.
- What are these scores actually good for?
Comparison and tracking. Judging two similar domains against each other, evaluating a domain you might acquire, and watching your own move over time are all legitimate uses. What they cannot do is predict whether a specific page will rank for a specific query, which is what they get used for.
- Should I filter link prospects by DA or DR?
Only loosely, and never as the primary criterion. A relevant link from a modest site in your field is usually worth more than an irrelevant one from a high-scoring general site, and the score cannot see relevance at all. Filtering purely on score is also exactly how link sellers price inventory, which should tell you something.
- Can these scores be manipulated?
Yes, and cheaply, which is the strongest argument against relying on them. Because they are largely backlink models, they respond to link volume regardless of whether that volume represents anything real. Expired domains with inflated scores are sold on exactly this basis, and the score says nothing about whether Google agrees.
