FeaturedLane
Blog

Ahrefs DR4: what changes on December 1, and what it means for your Domain Rating

On December 1, 2026, the most-quoted number in SEO changes under everyone's feet. Ahrefs is switching the Domain Rating shown in its standard interface from the current model, DR3, to a new one called DR4. Glen Allsopp announced it on Ahrefs' blog on November 5, and a preview has been available through Ahrefs' free DR API since then.

If you sell links, buy links, report DR to clients, or just check your own number every Monday morning, some of what you know is about to stop being true. Most sites will barely move. Some will drop hard, and a few will drop to almost nothing without a single backlink changing.

We run a directory where DR comes up in nearly every conversation with founders, so we read the announcement line by line and compared it with early reports from people testing the preview. This is the plain version: what DR is, what DR4 changes, what the data really shows (including the part most write-ups skipped), and what to do before the switch.

First, what DR actually measures

Domain Rating is Ahrefs' own score for the strength of a site's backlink profile, on a logarithmic scale from 0 to 100. It is not a Google metric, and Google never sees it. Ahrefs is blunt about this in the DR4 post: the score is "not a reflection of the quality of your content, your impact, or anything other than your backlink profile."

A few mechanics carry over from the way Ahrefs has long explained DR, and they matter for everything below:

  • Only followed links count. A site that links to you only with nofollow links does not raise your DR.
  • Only the first link from each domain counts. Ten links from one site are worth the same as one.
  • Links from stronger sites pass more, and a site that links out to thousands of domains spreads what it passes very thin.
  • It is logarithmic. Going from DR 20 to 30 is far easier than going from 70 to 80.

None of that changes with DR4. What changes is which links Ahrefs trusts, and how it treats subdomains.

What DR4 changes

Ahrefs lists four changes. Here they are without the marketing gloss.

1. Subdomains get their own score. Under DR3, a subdomain inherited the DR of its root domain. Under DR4 it is scored on its own links. Ahrefs' own help site is the cleanest example: help.ahrefs.com goes from DR 91 to DR 75, while ahrefs.com itself sits at 92.

2. Links from sites with search traffic count for more. Ahrefs now puts "more weight" on links from sites (and individual subdomains) that it estimates get organic search traffic. A link from a blog nobody visits is worth less than it was. A link from a site people actually find through Google is worth more.

3. A smaller pool of sites can move DR at all. Ahrefs says DR4 uses "a smaller, higher-quality pool of domains that can impact DR." In practice, the number of sites that even have a DR4 score is about half the number that had a DR3 score.

4. Spam gets hit hard. DR4 "significantly reduces the scores of domains labeled as spam." Ahrefs moved its spam labelling to machine-learning detection in January 2026, and spam sites with DR3 scores between 30 and 70 saw the largest drops, a median of around 15 to 20 points.

The numbers, including the bit nobody is talking about

Ahrefs' headline claim is that DR4 tracks Google rankings much better than DR3, with an overall correlation "more than 2.5x stronger." The table behind that claim, built on 10 million hosts, is worth reading line by line, because it does not say what most summaries say it says.

Correlation with ranking in Google's top 3 (DR3, then DR4):

  • All hosts: 0.128 to 0.332
  • Root domains only: 0.382 to 0.295
  • Subdomains only: 0.126 to 0.225

And for the top 10:

  • All hosts: 0.140 to 0.359
  • Root domains only: 0.407 to 0.304
  • Subdomains only: 0.172 to 0.259

Look at the root domain lines. For ordinary root domains, the kind most of us compare when we look at a competitor or a link prospect, DR4's correlation with rankings in this dataset is lower than DR3's, not higher. The big overall jump comes from subdomains, which DR3 handled badly because it simply copied the root domain's score onto them.

That is not a criticism of the update. Scoring subdomains properly was overdue. But "DR4 predicts rankings 2.5x better" is not something you should repeat to a client about their own root domain. Ahrefs has not published a methodology beyond the dataset sizes, so treat all of these numbers as directional.

On score movement, Ahrefs tested the 1,000 domains with the most organic traffic:

  • 64.3% saw their DR go up
  • 15.8% stayed the same
  • 19.9% went down, by an average of 7.3 points and a median of 4

For sites that are not labelled spam, Ahrefs says the median change is close to zero. That is the reassuring line for most legitimate businesses.

Who is going to lose points

From the examples in the announcement and what the preview shows, three groups should expect to drop.

Sites that live on someone else's domain. Ahrefs' examples are brutal. Example sites on hostingersite.com and notion.site subdomains went from DR 92 to 1.9 and 0, because they no longer borrow the platform's score. github.io itself dropped 32 points, while a subdomain with its own strong link profile, facebook.github.io, held at DR 86. If your product, blog or docs live on a platform subdomain, your DR was never really yours. Now it shows.

Sites whose links come from places with no traffic. Private blog networks, expired-domain blogs, thin guest-post farms and low-effort directories that rank for nothing. Under DR3 those links pushed DR up. Under DR4 they carry much less weight. If you bought a "DR 50+ link package" last year, check where those links actually sit.

Sites Ahrefs labels as spam. If your own site has been flagged, expect a big drop, and expect your outgoing links to count for very little.

Who is going to gain

Sites with fewer but better links. If most of your referring domains get real search traffic, DR4 is likely to treat you better than DR3 did, because your links are now weighed by something closer to their real value. Ahrefs' examples of large SaaS companies are mostly flat or slightly up: Airtable went up to 92, Monday.com moved from 91 to 90, and Teamwork.com stayed at 82.

What to do before December 1

1. Check your DR4 now. Ahrefs' free DR API returns DR3 by default, and adding version=dr4 to the request returns the new score. Do it for your own site and your main competitors so the change does not ambush you.

2. Save your DR3 history. After the switch, Ahrefs says you will be able to compare DR3 scores through the API "for a short period of time." It does not say how long. If you report DR to clients or investors, export the numbers this week and add a note to your next report explaining the model change. A sudden drop without context reads like a disaster. With context, it reads like a recalibration.

3. Sort your referring domains by traffic. In your backlink tool, look at the domains linking to you and sort them by their own organic traffic. The ones with zero traffic are the links DR4 discounts. You cannot do much about them, but you will know why your number moved.

4. Stop buying links on DR alone. This was always weak practice, and DR4 makes it actively expensive. Before paying for or chasing a link, check that the site has organic traffic, that its traffic is not collapsing, and that it does not link out to thousands of unrelated sites.

5. If your site lives on a subdomain, plan a move. A custom domain you own is the only way your links will ever build a score that belongs to you.

How to grow DR under the new model

DR4 rewards the same thing good SEO always rewarded: links from sites that real people visit. A few approaches that hold up:

  • Make things worth linking to. Free calculators, small tools, original data and honest comparisons earn links on their own. Some of the most-linked pages on the web are a single useful calculator.
  • Fix the links you already have. Old links pointing at broken pages leak value. A free checker like Broken Link Detector finds them quickly, and a full site audit from FreeScan.app catches the technical issues that stop pages being found at all. CachePlex also has a set of free SEO tools that need no sign-up.
  • Get listed where people actually browse. Launch platforms and directories still help, but under DR4 only the ones with real traffic move the needle. Before submitting anywhere, check whether the site ranks for anything and whether its listings get visitors. Weekly launch boards like Launchpadly, IndieStore and Launchdog, and curated directories such as StackLedge and ToolVerified, are worth a look. If you are launching to a lot of them, LaunchBert saves time with a roadmap and a form auto-filler.
  • Avoid duplicate copy everywhere. Pasting the same description into fifty directories hurts more than it helps. We wrote about why duplicate descriptions make directories disappear, and the same logic applies to your own listings.
  • Earn mentions, not just links. AI assistants now drive a share of discovery, and they rely on the same kind of trusted, visited sources. Tools like ShareOfAsk track how AI assistants mention your brand. Our guide on getting your SaaS recommended by ChatGPT covers the rest.

A note on our own model, since we are a directory: FeaturedLane gives a dofollow link only to listings that verify our badge on their site and reach an upvote threshold, or that choose the paid plan. Under DR4, what matters for any directory link is whether that directory has real search traffic. Check that for us, and for anyone else, before you rely on it.

If you buy or sell links

DR4 is going to cause a noisy few weeks in link marketplaces, and it is worth thinking ahead.

If you sell links or guest posts priced by DR, expect buyers to ask for DR4 numbers from December, and expect some inventory to lose value overnight. Sites whose DR came mostly from other sites with no traffic will look very different. The sellers who come out fine are the ones with real readers, because DR4 now rewards exactly that.

If you buy links, this is a good moment to stop treating DR as a price tag. A site with DR 35 and steady organic traffic is now a better bet than a DR 60 site whose traffic chart is flat at zero. Check traffic, check what else the site links to, and check whether it would still be worth a link if DR did not exist. If the answer is no, it was never a good link.

If you are an agency, update your proposals and reporting templates before December 1. Clients who were promised "DR 50 by Q2" deserve a quick note explaining that the ruler itself changed. It is a much easier conversation before the drop than after it.

Questions people are asking

Does Google use Domain Rating?

No. DR is Ahrefs' own metric, built from Ahrefs' own crawl. Google has its own signals and does not see your DR. Ahrefs' own advice for years has been that "Google ranks pages, not websites," which is why a low-DR site can still rank a great page.

My DR dropped. Did I do something wrong?

Probably not. If your links did not change, the drop is the model, not you. Check whether your site sits on a subdomain and whether your referring domains have organic traffic. Those two things explain most of the large drops.

Should I still care about DR?

Yes, as a quick comparison number, as long as you remember its limits. Ahrefs itself warns that DR and search traffic "won't always move in the same direction." Use DR to size up link prospects and competitors. Use traffic, rankings and conversions to judge whether your SEO is working.

Will DR3 still be visible after December 1?

Ahrefs has only said that DR4 becomes the default in the standard interface on December 1, and that DR3 scores stay available through its API "for a short period of time" afterwards, without giving a date. Export what you need now.

The short version

DR4 makes DR more honest about where links come from. Subdomains stop borrowing scores, links from sites nobody visits lose weight, and spam gets cut down. Most legitimate sites will see small changes. Sites on platform subdomains, and sites built on cheap links, will see big ones.

The headline correlation gain is real, but it comes mostly from subdomains, so do not oversell it. Check your DR4 now, save your DR3 history, and judge links by the traffic behind them rather than the number next to them.

If you have built something worth linking to, submit it to FeaturedLane. A person reads every listing before it goes up.