· 6 min read
Does a domain rating actually predict anything useful?
What a domain rating measures, what it correlates with, and why a site can hold a respectable score while sending nobody anywhere. Including our own number.
A domain rating is a number between 0 and 100 that describes the strength of a site's backlink profile, and it is the most-quoted figure in link building for a reason that has little to do with its accuracy: it is a single number, it is free to look up, and it lets two strangers agree on a price in one message. Whether it predicts anything you care about is a separate question, and the honest answer is "one thing, weakly, and not the thing most people use it for." (Disclosure: published by BidSurvivor, which displays its own domain rating on the site, so read the section where that gets examined with appropriate suspicion.)
What the number actually measures
Every major SEO tool has its own version — Ahrefs calls it Domain Rating, Moz calls it Domain Authority, Semrush calls it Authority Score — and none of them is a Google metric. Google does not publish one and has said repeatedly that it does not use these scores. They are third-party estimates of one thing: how many other sites link to you, weighted by how many sites link to those.
Two consequences follow immediately, and both are widely ignored.
It is logarithmic. Moving from 0 to 20 takes a handful of decent links. Moving from 60 to 80 takes thousands. A site at DR 18 and a site at DR 36 are not "half as strong" — the gap is far larger than the numbers suggest, and the gap between 70 and 80 is larger still.
It says nothing about traffic. A site can accumulate links from directories, aggregators and profile pages, register a respectable score, and receive almost no visitors. The reverse also happens: sites with strong traffic and weak link profiles are common in any niche where the audience arrives from a community rather than from search.
The correlation that exists, and its limits
There is a real relationship between link profile strength and search visibility. Links have been a ranking input since PageRank and remain one. Sites with many good links do, on average, rank for more things.
But "on average, across millions of sites" is a much weaker claim than the one people make when they buy a link. Two things break the inference at the level of an individual site:
Relevance dominates at small scale. A link from a DR 30 site in your exact niche routinely outperforms a DR 70 link from a general directory, because the first one sits in a context Google can read as topical and the second sits in a list. Ahrefs' own study of search traffic found that 96.55% of pages get no traffic from Google at all — and a large share of those pages have links pointing at them.
Search is sending fewer clicks anyway. Whatever a strong link profile bought two years ago, it buys less of now, because more queries resolve without a click. We went through the arithmetic in where small sites get traffic after AI Overviews, and the short version is that ranking and receiving visitors have come apart.
Where the number is genuinely useful
Three uses that survive scrutiny.
As a rough filter against obvious junk. A site at DR 2 with a thousand outbound links is a link farm. The number is a fast way to see that. It is a floor test, not a quality score.
As a negotiating unit. When two people who have never met need to agree that a trade is roughly fair, a shared public number does that job in one line. It does not need to be accurate to be useful for this; it needs to be public and hard to fake.
As a trend on your own site. Your own rating moving up over six months means links are accumulating. That is worth knowing. Comparing your absolute number to somebody else's is worth much less.
Our own number, examined
We display a domain rating on this site, which makes the argument above slightly awkward, so here is the position stated plainly.
The number is real, it is fetched from a third party rather than typed in, and it is small — this is a young domain. It is on the site for the second use above: people evaluating whether to trade a link with us can see it without asking, and a public number that anybody can verify is better than a claim in an email. It is not evidence that a link from here will send anybody to your site.
The thing we publish that is evidence of that is different and more specific: every ad slot's card opens and click-throughs, counted separately and shown on the front page and on each brand's own page. That is a record of humans arriving somewhere, which is the fact a rating only gestures at. The same distinction drives the backlink tracker — it joins the links that exist to the links that have ever sent a visitor, because those are different sets and almost nobody puts them in the same table.
The question worth asking instead
Before accepting or paying for a link, the useful questions are not about the rating:
- Would a reader of that page plausibly want what I sell? If not, the link is decoration.
- Does that page get read? A link on a page nobody visits is worth whatever a link on a page nobody visits is worth.
- Has that site ever sent anybody anywhere? This is answerable from your own referrer log after the fact, and it is the only one of these that produces a number you own rather than one you looked up.
- Would I want this link if scores did not exist? The most reliable filter on the list.
A domain rating answers none of those. It is a proxy for the first half of the first one, and it has become the whole conversation because it is the only part that fits in a table. If you take one thing from this: a link that never sent a visitor is not a weaker version of a good link, it is a different thing entirely, and no rating will tell you which one you have.