Domain Rating: What DR Predicts, and What It Does Not

Two sites in our sample both carry a Domain Rating of 91. One draws an estimated 2.37 million US organic visits a month. The other draws 393,893. That is the problem with reading DR as a performance number: it ranked the sample about right and missed the size of every gap.

We pulled DR, referring domains, US organic traffic and ranking keyword counts for ten SEO and AI visibility sites through the Ahrefs API on 1 November 2026, in a single batch so every figure shares one date. The full table is below, including our own site, which comes last on every column.

What DR actually measures

Domain Rating scores one thing: the strength of a backlink profile, as Ahrefs has crawled it. In Ahrefs' own words, DR "represents the strength of the website's backlink profile on a logarithmic scale from 0 to 100, with the latter being the strongest."

Three consequences follow from that definition, and they explain most of what the data shows.

• It is a link metric, not a traffic metric. Content quality, search demand and commercial intent are not inputs.
• The scale is logarithmic. Ahrefs states that "the higher the DR, the more difficult it is to increase it", so the distance from 75 to 85 is nothing like the distance from 25 to 35.
• The formula is private. Ahrefs says "the exact formula of Domain Rating is not disclosed", which means no one outside Ahrefs can reconcile a DR number with a link count.

Ahrefs itself claims only a correlation: it "found a correlation between the DR of the website and the number of keywords it ranks for". That is a careful claim, and the measurement below is consistent with it.

What we measured, and how

Ten domains, all in SEO software or AI visibility measurement, queried through the Ahrefs batch analysis endpoint in subdomains mode with the country set to the United States, on 1 November 2026. Every number comes from the same call, so none of the drift between tools or dates applies here. Traffic and keyword figures are Ahrefs estimates, not Search Console data.

Domain	DR	Referring domains	US organic traffic	US keywords
semrush.com	92	138,711	2,996,939	78,863
ahrefs.com	91	118,895	2,369,209	19,000
moz.com	91	98,220	393,893	12,812
screamingfrog.co.uk	87	26,313	17,795	884
surferseo.com	85	24,883	17,323	991
tryprofound.com	81	11,781	22,522	276
otterly.ai	78	7,459	679	59
sitebulb.com	75	6,367	6,901	583
athenahq.ai	72	4,212	566	30
growthhasten.com	31	514	0	0

Two notes on reading that table honestly. The 19,000 US keywords for ahrefs.com is the figure the API returned, not a number we rounded. And the referring domain count for ahrefs.com was 118,682 on an identical call a few hours earlier the same day, against 118,895 here: a 213-domain move inside one day, on the metric that feeds DR.

Finding 1: the same DR hides a six-fold traffic gap

Ahrefs and Moz both sit at DR 91 with referring domain counts of the same order, 118,895 and 98,220. Their estimated US organic traffic differs by about six times: 2,369,209 against 393,893. Nothing in the DR figure hints at that, because nothing in the DR formula is about which queries each site covers or how well.

Semrush, one point higher at DR 92, ranks for 78,863 US keywords against Ahrefs' 19,000. A single DR point sits on top of a four-fold difference in ranking footprint.

Finding 2: DR 81 drew more US traffic than DR 87 and DR 85

tryprofound.com, at DR 81, drew more estimated US traffic than screamingfrog.co.uk at DR 87 and surferseo.com at DR 85: 22,522 against 17,795 and 17,323. It did it with 276 ranking keywords, under a third of either of the other two, which rank for 884 and 991.

More traffic from fewer keywords implies a position on queries with more demand behind them, though the four columns here cannot confirm that on their own; a top-pages pull would. What the table does establish is that if you were picking which competitor to study, DR would have sent you to the wrong one.

Finding 3: DR 78 can mean 679 visits

otterly.ai carries DR 78, three points above sitebulb.com at 75. Sitebulb draws an estimated 6,901 US visits a month; Otterly draws 679, which is a tenth of it. The same pattern repeats at athenahq.ai: DR 72, and 566 estimated visits. Strong link profiles, with a ranking footprint far behind them.

DR measures the links pointing at a domain. Traffic measures whether anyone searching has a reason to arrive.

What the ranking order does get right

Being fair to the metric: across the ten sites, DR ordered the sample close to correctly. Sorting by DR and sorting by traffic produces the same sequence except for three pairs, all of them in the middle of the table. As a coarse sort of "which of these sites has the stronger link profile", DR did its job.

That count depends on how you break the DR 91 tie between Ahrefs and Moz, which is the piece's own opening example: put Moz first and it is four discordant pairs rather than three. The failure is one of resolution, not direction. DR tells you the order. It does not tell you whether the gap between two adjacent rows is forty per cent or four thousand per cent, and in this sample it was both.

What to use DR for, and what to use instead

Keep DR for the question it was built to answer, and use a different number for everything else.

1. Link prospecting: DR is a reasonable first filter for comparing the link profiles of sites you might earn a link from. Pair it with traffic so you do not chase a strong profile on a dead site. Our guide to earning backlinks covers the qualification step.
2. Competitive benchmarking: use ranking keyword counts and estimated traffic, not DR, to decide who is actually winning the queries you care about. Start with competitor analysis.
3. Your own site: use Search Console. Impressions, positions and clicks for your property are measurements rather than estimates, which is the difference our piece on what an impression is worth gets into.
4. Reporting to anyone who pays you: if a DR number is in the report, pair it with the traffic line in the same table. A DR that rose while traffic sat still is a fact the report should surface rather than lead with.

What this does not prove

Ten domains in one niche on one day is a small sample, and every traffic and keyword figure is an Ahrefs estimate rather than a measurement from the sites themselves. The pattern here is about the resolution of a link metric, not about any of these companies: nothing in these four columns says where a link profile came from or when it was built. A larger sample across several niches might tighten or loosen the ordering, and we have not run one.

Our own row is the honest footnote: DR 31, 514 referring domains, zero estimated US traffic. The score is not the outcome, which is the whole point of the exercise.