We ran 501 expired domains through Ahrefs’ current Domain Rating and through DR4, the model Ahrefs switches to on December 1, 2026. Their combined rating fell by 49 percent. The median domain dropped from DR 26 to DR 5, and 132 domains went to exactly zero.
This is a follow-up to our DR4 study of 107 working websites, where the combined rating fell 35 percent. Expired domains fared worse, and the way they fared tells us a fair amount about what DR4 is built to do. Below we set out the data, our hypotheses about what Ahrefs is after, what DR4 actually achieves on this sample, and what we still cannot tell.
Key findings
These are the numbers that matter most from the study. All of them come from Ahrefs’ free Domain Rating endpoint, measured on November 14, 2026, for the same 501 expired domains, most of them listed in Namecheap’s expired-domain auctions that week.
- Ahrefs DR4 cuts the combined Domain Rating of 501 expired domains by 49 percent compared with the current model, DR3.
- The median expired domain in the sample falls from DR 26 under DR3 to DR 5 under DR4.
- One in four expired domains, 132 of 501, has a DR4 of exactly zero.
- Of 60 expired domains rated DR 40 or higher, only 18 keep that level under DR4, and ten drop from DR 50 to zero.
- For 68 domains, the DR shown in the Namecheap auction listing was more than five points above Ahrefs’ own current figure.
- In the six domains we examined closely, those that gained under DR4 had links from sites with real search traffic, and those that fell to zero had links from sites with none.
The sample and the method
The sample is 501 expired domains, most of them listed in Namecheap’s expired-domain auctions closing between November 16 and 17, 2026. From the auctions we kept only names Namecheap listed at DR 20 or higher, without digits or hyphens, and dropped obvious junk. We pulled DR3 and DR4 for each domain from Ahrefs on November 14.
Both scores come from Ahrefs’ free Domain Rating endpoint, which returns the current model and DR4 side by side. For six domains, three of the biggest gainers and three of the biggest losers, we went further and looked at their referring domains, the traffic of the sites linking to them, and when those links first appeared. Nearly every domain in the sample was at the $15 opening bid, so price tells us nothing here and we left it out.
One detail before the results. The DR shown in the auction listing is already out of date. For 68 domains, Namecheap’s figure was more than five points above Ahrefs’ own current DR3, and 63 domains listed at DR 20 or more were already below 20 in Ahrefs on the day we checked.
The results at a glance
Three out of four domains lost rating under DR4, and more than half lost at least half of it. Only 152 of the 501 still sit at DR 20 or above, the threshold almost the whole sample was selected on. A minority moved the other way, and some of those gains were large, which matters for what comes later.
- Combined DR3: 13,751. Combined DR4: 6,977. A fall of 49 percent.
- Median rating: DR 26 under DR3, DR 5 under DR4.
- 376 domains fell, 17 stayed the same and 108 rose.
- 291 domains lost half their rating or more.
- 178 domains ended below DR 1, and 132 of those at exactly zero.
- 54 domains gained ten points or more. The biggest single gain was 39 points.
Breaking the sample down by current rating shows that no band was safe. Every band from DR 20 upward lost between 42 and 56 percent of its combined rating.
| DR3 band | Domains | Change in combined rating | Median DR4 | DR4 below 1 |
|---|---|---|---|---|
| Under 20 | 65 | -16% | 0.4 | 34 |
| 20 to 29 | 263 | -52% | 4.8 | 88 |
| 30 to 39 | 113 | -49% | 8 | 35 |
| 40 to 49 | 35 | -56% | 13 | 11 |
| 50 and above | 25 | -42% | 35 | 10 |
The top of the table is where the losses are most striking. Of the 60 domains rated DR 40 or higher, only 18 are still at DR 40 or higher under DR4, and 21 fell below DR 1. Ten domains went from DR 50 to exactly zero.
Six domains up close
Aggregates hide the mechanism, so we opened up six domains: three that went from DR 50 to zero and three that gained between 34 and 39 points. The two groups have similar numbers of referring domains. What separates them is who is linking and whether those linking sites get any visitors.
The three that fell to zero
The three losers look almost identical. Each has around 930 referring domains, roughly one backlink per domain, about 310 of them dofollow and links from about 210 IP subnets. Profiles that close to each other suggest the same source built all three.
Sorted by Domain Rating, their strongest linking sites are link shops and “rank checker” pages on cheap domains, rated DR 50 to 76 and with no estimated organic traffic at all. Most of those links were first seen between April and September 2026. In other words, the DR 50 these domains showed at auction appears to have been built this year, from sites that seem to exist to pass rating rather than to be read. Two of the three also had a handful of links from very large sites, but not enough to hold any rating once the rest stopped counting.
The three that rose
The gainers have fewer links and, in two cases, very few dofollow ones. What they have is links from sites people actually visit. Each got there a different way, which is the useful part.
- A graphic design domain, DR 34 to 72. Its links come from design marketplaces, tutorial sites and design blogs, several with over 100,000 monthly visits. This is the textbook case: links from real sites in its own field.
- A coding-related domain, DR 12 to 51. Only 88 of its 829 referring domains are dofollow. Its gain comes with links from a large website builder, Google’s blog domain and two website-information sites with tens of thousands of visits each.
- A company domain, DR 24 to 58. Only 78 of 462 referring domains are dofollow. Its strongest links are business listings on a national yellow-pages site and on two large job portals, each with between 250,000 and 575,000 monthly visits.
None of the six domains, winners or losers, has any organic traffic of its own. The rating moved entirely on the quality of what links to them.
What Ahrefs seems to be after
Our hypothesis is that DR4 is meant to stop Domain Rating from being something you can manufacture. Under DR3, a network of rated but unvisited sites could lift any domain to DR 50. Under DR4, a link mostly counts if the site behind it gets real search traffic, which is much harder to fake at scale.
Ahrefs says DR4 puts more weight on links from sites with organic search traffic, and that it scores subdomains separately, so a free subdomain on a big platform no longer borrows the platform’s rating. Both changes point the same way. They make DR a closer proxy for links that real visitors might follow, and a worse target for anyone selling inflated metrics. The expired-domain market, where DR is often the main number in a listing, is exactly where that inflation has been easiest to sell.
What DR4 achieves on this sample
On the evidence here, DR4 mostly does what Ahrefs appears to intend. The domains with manufactured profiles collapsed, and the domains with old links from busy sites gained. A buyer filtering this auction by DR3 would have been steered toward the wrong names. A buyer using DR4 would have been steered toward better ones.
It is not airtight. One of the domains we first picked as a top gainer, and later replaced, had climbed by 37 points almost entirely on links planted in hacked pages of busy sites: a government portal, a political party branch, a regional newspaper, a hotel. The links had nothing to do with the pages they sat on. DR4 counted them, because the sites hosting them do have traffic. So DR4 raises the cost of faking a rating, but it does not appear to tell an editorial link from an injected one.
What we infer for expired-domain buyers
The practical inference is that the DR in an auction listing should now be treated as a claim to check, not a fact. A high DR3 with a low DR4 usually means the rating was built rather than earned, and that is precisely the profile that loses value fastest once everyone switches models.
- Check DR4 next to DR3 before bidding. Ahrefs offers both for free, and a large gap between them is the single quickest signal in this data.
- Look at who links, not how many. The winners here had fewer links than the losers. What they had was links from sites with visitors.
- Do not dismiss nofollow profiles automatically. Two of our three gainers were mostly nofollow and still rose sharply. We read that as DR4 giving weight to links from high-traffic sites even without the dofollow attribute, but that is our inference, not something Ahrefs has stated.
- Be wary of sudden link growth in the past year. The three losers had almost all of their strong links first seen in 2026.
- Check a few of the top linking pages by hand. That is the only way we found the injected links behind the domain we dropped.
For what to do with a domain once you own it, see our comparison of rebuilding an expired domain versus redirecting it.
What we still do not know
This is one auction window, one registrar and one snapshot in time. DR4 was still a preview when we pulled the data, so the figures may shift before or after December 1. We also only looked closely at six domains, which is enough to suggest a mechanism but not to prove one.
- Whether nofollow links count. Our two nofollow-heavy gainers suggest they do when the linking site has traffic, but we cannot isolate that effect from the few dofollow links they also have.
- Whether Ahrefs will catch injected links. The hacked-page case may be an edge case, or a gap that link sellers will now target. We have no way to tell yet.
- Whether DR4 tracks rankings any better. DR is Ahrefs’ metric, not Google’s. A domain gaining DR4 is not evidence that Google values its links more, only that Ahrefs’ model now resembles one plausible way of doing so.
- How representative Namecheap’s listings are. Other marketplaces filter and price differently, and their inventory may have more or less built rating in it.
Figures in this article reflect the data available when it was last updated.
Frequently asked questions
Is an expired domain with DR4 zero worthless?
Not necessarily. A zero says the domain’s links come from sites without search traffic. If the name itself is valuable to you, the rating matters little. If you were buying it for its links, a DR4 of zero is a strong sign those links were never worth much.
Why do auction listings show a different DR from Ahrefs?
Listings take their figures from a data feed that lags behind Ahrefs. In our sample, 68 domains were listed more than five points above their current Ahrefs DR3. Always check the live figure before bidding.
Can DR4 still be gamed?
It is harder, not impossible. Networks of unvisited sites stop working, but one domain in our sample gained heavily from links injected into hacked pages on busy sites. Manual checks of the strongest linking pages remain necessary.
Did higher-rated expired domains hold up better?
Only partly. Domains rated DR 50 and above lost 42 percent of their combined rating, slightly less than average, yet ten of them fell from DR 50 to zero. A high DR3 was no protection when it had been built quickly.
