We measured 107 small and mid-size websites under Ahrefs’ current Domain Rating and under DR4, the model Ahrefs adopts on December 1, 2026. Their combined rating fell by 35 percent. Seventy-seven sites lost rating, 59 lost at least half of it, and the sites with the highest ratings were among the hardest hit.
Ahrefs’ own analysis of DR4 looked at the largest sites on the web and found most of them gaining. Our sample looks at the other end: the small content sites, newsletter sites, tools and expired domains that make up most of the web. The picture there is very different.
What changes on December 1
Domain Rating is Ahrefs’ 0 to 100 score for the strength of a site’s backlink profile compared with every other site in its index. DR4 is the fourth version of that model. Ahrefs previewed it on November 5, 2026, and it replaces the current model, DR3, in the Ahrefs interface on December 1.
Two changes explain most of the movement. First, DR4 is still calculated from links, but it puts more weight on links from sites that Ahrefs estimates receive organic search traffic. A link from a site that real people find through Google now counts for more than a link from a site nobody visits. Second, subdomains get their own score instead of inheriting the rating of their root domain.
Ahrefs reports that DR4 correlates better with Google ranking positions than DR3, and that among its top 1,000 domains by organic traffic, 64.3 percent gained rating, 15.8 percent stayed the same and 19.9 percent fell, by a median of 4 points. Those are the biggest sites on the web. We wanted to know what happens to everyone else.
The sample and the method
The sample is 107 websites that we track or have evaluated at expired domain auctions. They are small and mid-size sites: content sites, newsletter sites, small software tools and directories, many of them built on expired domains with years of backlink history. None of them is a household name.
For each site we took both scores from Ahrefs’ free Domain Rating endpoint, which returns DR3 and DR4 side by side, on November 14, 2026. Sites that already had a DR3 of 0 were left out, because there was nothing for the new model to change. We do not name the sites: the point is the pattern, not the individual domains.
The results at a glance
Across the sample, DR4 removes about a third of the rating that DR3 showed. Most sites fall, many fall a long way, and the median site ends up with a rating close to the bottom of the scale.
- Combined Domain Rating falls from 2,339 to 1,528 points, a drop of 35 percent.
- 77 sites lose rating, 30 gain, and none stays exactly the same.
- 59 sites, more than half of the sample, lose at least half of their rating.
- 41 sites drop below DR 1.
- The median site falls from DR 22 to DR 6.
Higher ratings were not safer
The most useful finding is how the change splits by starting rating. If a high DR3 meant a strong, trustworthy profile, high-rated sites should hold up best. They did not.
| DR3 range | Sites | Fall below DR 1 | Lose half or more | Gain rating |
|---|---|---|---|---|
| Under 10 | 28 | 22 | 22 | 4 |
| 10 to 19 | 22 | 8 | 12 | 7 |
| 20 to 29 | 30 | 2 | 11 | 12 |
| 30 to 39 | 16 | 4 | 8 | 4 |
| 40 or higher | 11 | 5 | 6 | 3 |
Of the 11 sites rated DR 40 or higher, five fall below DR 1 under DR4 and only three gain. Half of the 16 sites in the 30 to 39 band lose at least half of their rating. The largest single fall in the sample is a site that goes from DR 92 to DR 26.
The sites at the bottom fare worst of all in relative terms: 22 of the 28 sites that started below DR 10 end up below 1. For them, DR4 does not lower the score so much as erase it.
The middle held up best
Sites that started between DR 20 and 29 are the clear exception. Only two of the 30 fall below 1, and twelve gain rating under DR4, the highest share of any band.
That fits what DR4 is designed to reward. A rating built from a moderate number of links from living sites survives the change. A rating built from a large number of links from dead or traffic-free sites does not, however high it was.
Why high ratings collapsed
Under DR3, a link counted according to the rating of the site it came from, whether or not anyone visited that site. That made it possible to build a high rating from links on sites with high ratings and no audience: link-selling sites, abandoned blogs, networks of directories that link to each other. Many expired domains carry exactly this kind of history.
DR4 discounts those links because the sites they come from have no estimated organic traffic. When most of a profile’s strength sat in links like that, the rating falls almost to nothing. Others measuring DR4 have seen the same effect in directories: an analysis of 2,405 directories found that DR4 cut the pool of rated sites from about 287 million to about 132 million, and a separate check of 359 SaaS launch directories found 93 percent of them scored lower than before.
None of this means the sites in our sample lost rankings on December 1. Domain Rating is a third-party metric, not a Google ranking factor, as we explain in our guide to what backlinks actually do. What changes is the number that agencies, link buyers and domain investors use to price links and judge sites. For how links fit with everything else that moves rankings, see our guide to how SEO works.
What this means in practice
For anyone buying expired domains, DR3 is now a poor screen on its own. Check both scores before you bid. A domain whose DR4 is far below its DR3 is selling you links that the new model has already stopped counting. Our guide to rebuilding or redirecting expired domains covers what to look for beyond the number.
For link building, the lesson is to judge prospects by their traffic rather than their rating. Sort your referring domains and your prospect lists by estimated organic traffic. The links that still count under DR4 are the ones from sites with readers, which are also the links most likely to send visitors and to help rankings.
For reporting, expect a visible drop in December even if nothing about your site has changed. Explaining it in advance, with both numbers side by side, is easier than explaining it afterwards.
Limits of this data
This is a sample of 107 sites, not a census, and it leans towards small sites and expired domains. It is not representative of the web as a whole, and it is not meant to be: it describes the kind of sites that Ahrefs’ top 1,000 analysis does not. Scores are a snapshot from November 14, 2026, and Ahrefs notes that live scores can move by a few points as data updates.
Figures in this article reflect the data available when it was last updated.
Frequently asked questions
When does DR4 replace the current Domain Rating? On December 1, 2026, in the standard Ahrefs interface. Ahrefs previewed the model on November 5, and its free Domain Rating endpoint already returns DR4 when you ask for it.
Why did my Domain Rating drop under DR4? Most often because many of your links come from sites with little or no organic traffic, which DR4 counts for much less. Links from subdomains of large platforms can also be worth less, because those subdomains now have their own scores.
Does a lower DR4 mean I will lose rankings? No. Domain Rating is an Ahrefs metric and Google does not use it. A lower score changes how others value your site, not how Google ranks it.
How can I check my DR4 before December 1? Ahrefs’ free Domain Rating endpoint accepts a version parameter and returns DR4 alongside the current score, so you can compare both for any domain.
