Ahrefs DR4 explained: the new Domain Rating and what to do next
What DR4 changes, why your score may look different, and how founders and SEO providers can prepare without confusing a new metric with lost traffic.
DR4 is Ahrefs' new model for Domain Rating. Announced on November 5, 2026, it is available as a preview, with the standard Ahrefs interface rollout scheduled for December 1. This guide reflects the information available on November 9, 2026. Source: Ahrefs' announcement.
Start by comparing DR3 and DR4 for your exact website host. Keep both readings and their dates. A different score under a different model is not a measurement of how much your business grew or declined.
What changes with DR4?
The model remains link-based. It gives more weight to links from hosts estimated to receive organic search traffic, and scores subdomains separately instead of broadly inheriting the root domain's rating. Fewer domains can influence other sites' scores. Ahrefs describes the changes here.
For a founder evaluating a possible link, those changes suggest two useful questions: does the actual linking host have an audience, and is the specific page worth appearing on? A well-known parent domain alone is a weak reason to buy a placement on an unrelated hosted page.
| Decision | What to check now |
|---|---|
| Comparing your score with a competitor | Use the same model, similar dates and comparable hosts |
| Assessing a blog or directory | Review its relevant pages, audience and editorial standards |
| Assessing a hosted subdomain | Check that host rather than borrowing the parent site's reputation |
| Buying a service | Specify the metric version and inspect the proposed methods |
| Reporting progress | Separate model changes from changes measured over time |
Is DR4 better because it considers traffic?
We think the direction is useful for evaluating link opportunities: it encourages a closer look at whether a linking site participates in real search discovery. That is an interpretation, not proof that every high-scoring link will send customers.
It also does not mean that buying visits, running ads or increasing your own page views directly increases DR4. The announced traffic consideration concerns the sites linking to you. Traffic estimates are not your private analytics, and a niche site can be useful to your audience even when an SEO tool estimates little traffic.
Ahrefs reports a stronger correlation with Google ranking positions across all hosts. Its published root-domain-only comparison favors DR3, however. So “better in every situation” overstates the evidence. See the original comparison.
Our practical approach is to use DR as one screening signal, then inspect relevance, the actual linking page, search visibility and referral results. None of those checks should disappear because a new score looks more sophisticated.
Why did my Domain Rating drop?
First establish whether you are comparing different models. Two readings taken together under DR3 and DR4 are not a before-and-after experiment.
On November 9, 2026, we compared 100 selected project domains listed on VCL: 93 scored lower under DR4, five were unchanged and two scored higher. Read the full sample, methods and limitations before applying those numbers to your own site.
Do not delete links, change domains or replace a working acquisition channel just because two versions disagree. Start with evidence about the problem you actually have:
- A score difference: compare the same host under both models.
- Fewer Google clicks: inspect Search Console by page and query, including dates and seasonality.
- Fewer signups: inspect the relevant acquisition and conversion funnel.
- Lost links: verify the referring pages and when those links changed.
Ahrefs' DR is not a Google ranking factor. A change in this third-party metric does not itself establish a Google penalty. Ahrefs' explanation of authority metrics and ranking factors.
Do directory submissions still make sense?
Sometimes, as distribution. A maintained directory with a relevant category and genuine visitors can help the right person find your product. The submission itself is not a promise of a particular DR4 increase.
Evaluate each opportunity rather than buying a count of listings. Look for useful descriptions, appropriate categories, maintained pages and a credible reason your audience would visit. Record referral visits and qualified signups when you can. Do not assume that an impressive root-domain score transfers to every page or hosted subdomain.
Google identifies low-quality directory links, excessive link exchanges and paid links that pass ranking credit as link-spam risks. Paid placements should be appropriately qualified. A provider's DR promise does not override those rules. Google's link-spam policy.
How should I try to increase DR4?
Start with a versioned baseline, publish something useful to a defined audience, and earn relevant editorial mentions. Review the linking pages and measure business outcomes alongside the score. Our practical Domain Rating guide turns that into a working plan.
For a new app, a useful integration example or a reproducible product comparison can give relevant publishers a reason to mention it. A generic article produced only to hold a link does not offer the same reason. There is no reliable conversion table from “number of submissions” to “DR4 points.”
What this means for DR Growth on VCL
We are keeping new paid DR Growth campaigns paused while we assess targets and provider methods against the changed metric. The free comparison remains available. Historical DR3 results are not evidence of DR4 delivery, and joining either waitlist creates no paid campaign.
Building a product? Compare your scores, then join the buyer waitlist if you want updates about future DR Growth availability. No ranking, traffic or revenue outcome is promised.
Providing SEO services? Join the provider waitlist. Be ready to explain your methods, show dated results using the same metric version, and distinguish referral or search gains from a score change. Joining is an expression of interest, not provider approval.