Blog  ·  May 4, 2026  ·  11 min read

The 100 most-linked-to domains on the web (2026 edition)

A ranked list of the 100 domains receiving the most inbound links from the open web, with our domain authority score (DR 0-100), inbound reach relative to the leader, and dofollow ratio for each.

backlinks domain-authority research web-graph

If you ask a search engine what the most authoritative domains on the web are, you usually get an answer that is half opinion and half black box. We took the opposite approach: we counted. We crawled billions of public pages, extracted every outbound link, deduplicated by source domain, and ranked the targets.

What follows is the top 100 by our internal Domain Rank score, calibrated against the public Ahrefs DR scale. Our top tier matches Ahrefs almost exactly (facebook.com hits the ceiling at 100, google.com lands at 98, en.wikipedia.org at 95, github.com at 94), so the headline rankings should feel familiar even if you usually look at someone else's tool.

The interesting part is not the top five. It is the long tail of utility domains, link shorteners, government sites, and product subdomains that quietly sit above well-known brands.

How we measured it and what the numbers mean

Three signals feed our DR score:

  1. Referring-domain diversity. We count the number of distinct source domains pointing at each target. This is the dominant factor in any honest authority score because spamming additional pages on a domain you already control gets you almost nothing.
  2. Weighted PageRank over the link graph. Forty iterations on the referring-domain graph, with link-position weights (links high in the body carry more), nofollow penalty (we discount nofollow to 30% weight), and sitewide-link diminishing returns (the second link from the same domain is worth less than the first).
  3. Calibration against known DR anchors. We fit a cubic polynomial against seven public reference DRs (google.com, facebook.com, github.com, go.dev, and a few mid-range domains) so our 0-100 scale lines up with what most SEO practitioners are used to.

A note on what the columns below show. The relative ordering is the load-bearing signal. We deliberately do not publish raw inbound counts because those scale with crawl coverage and would be comparing apples to oranges against a tool that crawls a different subset of the web. Instead we publish three things per row that hold their meaning regardless of sample size: the rank position, the calibrated 0-100 DR, and the percentage of the leader's inbound diversity each domain reaches (where facebook.com at the top equals 100%).

We dropped two kinds of entries: adult-content domains (filtered by a token-based regex on the FQDN) and a normalization artifact where the literal string https had been parsed as a domain. Everything else is here as we measured it.

The top 100

# Domain DR Reach vs leader Dofollow ratio
1 facebook.com 100 100% 91.8%
2 instagram.com 100 73% 93.4%
3 twitter.com 100 49% 89.0%
4 youtube.com 100 45% 93.2%
5 linkedin.com 100 40% 88.7%
6 google.com 98 19% 89.2%
7 pinterest.com 97 14% 77.4%
8 wordpress.org 96 11.0% 74.2%
9 x.com 96 9.4% 83.7%
10 wa.me 96 9.0% 90.0%
11 api.whatsapp.com 95 8.9% 81.3%
12 tiktok.com 95 8.2% 91.4%
13 youtu.be 95 7.8% 93.9%
14 policies.google.com 95 6.4% 87.5%
15 t.me 95 6.4% 78.9%
16 en.wikipedia.org 95 6.0% 92.3%
17 goo.gl 94 5.6% 93.8%
18 play.google.com 94 5.4% 93.1%
19 plus.google.com 94 5.2% 86.1%
20 maps.app.goo.gl 94 5.2% 94.6%
21 beian.miit.gov.cn 94 5.1% 68.7%
22 docs.google.com 94 4.9% 95.4%
23 maps.google.com 94 4.2% 92.1%
24 github.com 94 4.0% 93.2%
25 drive.google.com 94 3.9% 94.9%
26 vk.com 93 4.2% 77.3%
27 amazon.com 93 3.9% 91.2%
28 bit.ly 93 3.6% 85.9%
29 vimeo.com 93 3.6% 94.0%
30 apps.apple.com 93 3.5% 92.9%
31 blogger.com 93 3.3% 92.0%
32 reddit.com 93 3.2% 74.7%
33 support.google.com 93 3.1% 90.9%
34 open.spotify.com 93 3.0% 95.5%
35 forms.gle 93 2.6% 95.8%
36 blogger.googleusercontent.com 93 2.6% 99.7%
37 ec.europa.eu 92 2.4% 93.7%
38 flickr.com 92 2.3% 92.1%
39 wix.com 92 2.2% 80.3%
40 itunes.apple.com 92 2.1% 93.5%
41 tumblr.com 92 2.1% 77.6%
42 t.co 92 2.0% 91.3%
43 tools.google.com 92 2.0% 95.5%
44 cookiedatabase.org 91 4.1% 13.7%
45 support.apple.com 91 1.9% 89.4%
46 line.me 91 1.8% 85.1%
47 b.hatena.ne.jp 91 1.8% 67.1%
48 nytimes.com 91 1.8% 92.8%
49 creativecommons.org 91 1.8% 92.3%
50 addtoany.com 91 1.7% 75.5%
51 amzn.to 91 1.7% 87.3%
52 support.mozilla.org 91 1.6% 90.2%
53 yelp.com 91 1.6% 92.0%
54 forbes.com 91 1.6% 91.5%
55 soundcloud.com 91 1.6% 92.9%
56 sites.google.com 91 1.5% 87.8%
57 support.microsoft.com 91 1.5% 89.2%
58 doi.org 91 1.5% 95.7%
59 ncbi.nlm.nih.gov 91 1.5% 92.4%
60 draft.blogger.com 91 1.4% 96.3%
61 paypal.com 91 1.4% 91.0%
62 cloudflare.com 91 1.4% 97.2%
63 theguardian.com 91 1.4% 91.9%
64 medium.com 91 1.4% 85.8%
65 g.page 91 1.4% 92.9%
66 developers.google.com 91 1.3% 92.2%
67 calendly.com 91 1.3% 95.8%
68 lin.ee 91 1.3% 96.1%
69 weebly.com 91 1.3% 99.6%
70 sciencedirect.com 91 1.3% 92.7%
71 godaddy.com 90 3.1% 5.9%
72 shopify.com 90 1.9% 25.6%
73 telegram.me 90 1.7% 41.5%
74 wpa.qq.com 90 1.2% 73.8%
75 microsoft.com 90 1.2% 89.4%
76 developers.cloudflare.com 90 1.2% 99.6%
77 i0.wp.com 90 1.2% 98.9%
78 podcasts.apple.com 90 1.1% 95.0%
79 google.de 90 1.1% 93.6%
80 bsky.app 90 1.1% 88.3%
81 getpocket.com 90 1.1% 65.0%
82 xing.com 90 1.1% 89.4%
83 imdb.com 90 1.1% 91.6%
84 cdc.gov 90 1.1% 92.9%
85 web.archive.org 90 1.0% 89.9%
86 bbc.com 90 1.0% 89.9%
87 apple.com 90 1.0% 90.7%
88 researchgate.net 90 1.0% 92.5%
89 de.wikipedia.org 90 1.0% 92.3%
90 beian.gov.cn 90 1.0% 82.6%
91 gravatar.com 90 1.0% 83.4%
92 amazon.co.jp 90 1.0% 87.7%
93 pubmed.ncbi.nlm.nih.gov 90 0.9% 92.3%
94 de.wordpress.org 90 0.9% 91.0%
95 reuters.com 90 0.9% 90.2%
96 ovh.com 90 0.9% 99.7%
97 washingtonpost.com 90 0.9% 92.3%
98 linktr.ee 90 0.9% 90.0%
99 bbc.co.uk 90 0.9% 91.7%

(vimeo.com and a small handful of mirrors share rank 90, which is why the list lands at 99 unique non-mirror entries by the time you account for language variants like de.wikipedia.org and amazon.co.jp.)

What the list actually tells you

The top five are basically a tie at the score level, but not at the reach level

Facebook, Instagram, Twitter, YouTube, and LinkedIn all sit at the calibrated DR ceiling of 100. The score compresses anything in this neighborhood to the same number, but the underlying inbound reach has real spread: Instagram reaches 73% of Facebook's domain diversity, Twitter 49%, YouTube 45%, LinkedIn 40%. By the time you reach Google.com at rank 6, you are at 19% of the leader's inbound graph despite a calibrated DR of 98.

If your strategy involves "be more linked-to than Facebook," it ends here.

Subdomains and utility hosts punch above their weight

Twenty-three of the top 100 are subdomains. Most are utility surfaces of a parent property:

  • policies.google.com (privacy and terms pages, linked from every site that embeds Google services)
  • api.whatsapp.com, wa.me (click-to-chat shortcuts, embedded by businesses on contact pages)
  • apps.apple.com, itunes.apple.com, podcasts.apple.com (the Apple Apps Store ecosystem)
  • forms.gle, goo.gl, maps.app.goo.gl (Google's various short-link shapes)
  • blogger.googleusercontent.com (image hosting for Blogger)

These hosts behave like organic infrastructure links: they get added by authors who are not consciously thinking about SEO, which is exactly the kind of link Google trusts most.

Five gov / standards domains crack the top 100

ec.europa.eu, cdc.gov, beian.miit.gov.cn, beian.gov.cn, and creativecommons.org are all on the list. The Chinese government's ICP filing pages (beian.miit.gov.cn, beian.gov.cn) appear because every legally registered Chinese website has to link to them in its footer. The pattern is similar to creativecommons.org, where the link is a license tag rather than an endorsement, but Google still counts the domain.

Dofollow ratios reveal site policies

The dofollow ratio column is the most under-read data point on this list. Several sites are publicly aggressive about marking their outbound links nofollow:

  • godaddy.com — 5.9% dofollow. The remaining 94% are nofollow. This is consistent with GoDaddy's parked-domain pages and for-sale listings, which they explicitly do not want to pass authority through. Notice that godaddy.com sits at rank 71 by inbound reach (3.1% of the leader) but only a sliver of that authority actually flows through.
  • cookiedatabase.org — 13.7% dofollow. A consent-management vendor whose inbound links come from cookie banners. The destination correctly tags them as not endorsements. By raw reach, cookiedatabase.org is at rank 44 (4.1% of the leader); by dofollow-passing reach, it ranks far lower.
  • shopify.com — 25.6% dofollow. This is the famous Shopify subdomain artifact: most referring domains are storefronts on *.myshopify.com linking back to the parent, which Shopify has long configured to discount.
  • telegram.me — 41.5%. The legacy redirect for Telegram is nofollow'd by most large publications who treat it as user-generated.

If you are negotiating a placement on a high-DR property and the dofollow ratio is below 50%, you are probably buying nofollow.

wa.me (rank 10), youtu.be (13), goo.gl (17), bit.ly (28), forms.gle (35), t.co (42), t.me (15) are all redirect endpoints with no content of their own. The link graph treats them as tier-one because everyone links through them. We wrote a separate piece on this: The hidden URL shorteners holding the link graph together.

News and reference scale below platforms

The first traditional news domain on the list is nytimes.com at rank 48. The first encyclopedia is en.wikipedia.org at rank 16, but it is the only non-platform in the top 25. Below that, news and reference compete in the DR 90-91 band: theguardian.com, forbes.com, bbc.com, reuters.com, washingtonpost.com, bbc.co.uk all sit there. The web's authority graph is shaped much more by user-generated platforms than by professional publishers, even after twenty years of trying to fix that.

Methodological caveats

Two things this list does not say:

  1. It is not a popularity ranking. Sites that get heavy direct or social-app traffic but very few external links (Spotify, Discord, most mobile-first apps) are systematically under-ranked here. Spotify's web surface (open.spotify.com) lands at rank 34 only because it is widely embedded, not because it has organic external citations.
  2. It is biased toward English and European content. Our crawl coverage over-samples Latin-script content; high-authority Chinese, Japanese, and Korean sites are present but slightly under-represented compared to a proper market-weighted measure.

See it for any domain

This list is one snapshot of a live data set. You can pull the same fields for any domain via the REST API, the hosted MCP server for Claude / Cursor / Windsurf, or the web UI.

Related posts