A ranking chart showing a gibberish domain placed above major websites, illustrating how misleading backlink data can be

A domain made of gibberish — w2rt6byl5nf1opmc-a.com — ranked number 8 in the Majestic Million, above microsoft.com at number 10 and github.com at number 11. It was the fastest-growing domain in the top 100. And it does not even resolve: type it into a browser and you get a DNS error. The explanation, when it came, was simpler than any conspiracy: Majestic's own unfiltered test data.

This story is funny, but the lesson is serious: backlink counts alone prove nothing. A single metric, from a single tool, put a non-existent domain above two of the biggest sites on the internet. This article retells the mystery, then teaches you the five-check method to investigate any strange domain you find in your own logs or backlink reports — so you can tell a real threat from bad data.

The Domain That Ranked Above Microsoft

The mystery was documented on jack.cab, whose author noticed the domain while browsing the Majestic Million — a ranking of the top million domains, ordered by the number of different referring subnets pointing at each one. At number 8 sat w2rt6byl5nf1opmc-a.com, a string that looks like a password, not a website.

Majestic's own numbers for the domain were staggering. Site Explorer reported 241,085,378 nofollow backlinks and just 6 follow backlinks, spread across 5,163,708 referring domains. All of them were "fresh", meaning recorded within the last 120 days. The domain also held rank 38 on Tranco, a ranking list that uses Majestic data — the same bad data, echoed downstream.

And it was not alone. Six hash-like subdomains of the same gibberish also ranked: at positions 157, 424, 554, 822, 1,111, and 36,933. Whatever was generating these numbers, it was generating them at scale.

The Clues That Didn't Add Up

Every additional check made the domain stranger, not clearer. Here is what the investigation turned up, clue by clue:

Check What it showed Why it was suspicious
DNS resolution Firefox returned dnsNotFound; the domain does not resolve A top-10 domain that no browser can reach
Ahrefs backlinks Just 53 backlinks Majestic reported over 241 million — the tools disagreed by orders of magnitude
Registration date First registered November 11, per Certificate Transparency After most of the backlinks; Tranco had data from October 5
Anchor text table Added up to nine links Nowhere near the 241 million links claimed
Indexed pages Majestic indexed 1,267 pages, crawled 27 Pages indexed on a domain that never resolved
Wayback Machine Not archived No trace it ever served a page

One more detail deepened the puzzle: an AI-generated pull request on GitHub from about two weeks earlier referenced the domain. At first glance it looked manufactured — a fake domain, fake links, fake activity. The investigation drew wide attention, and the Hacker News thread reached 328 points and 146 comments before the answer arrived.

The domain's own numbers contradicted themselves at every turn. Links reported as 100% TextLinks, an anchor table summing to nine, millions of referring domains for a site that never answered a single request. Something was wrong with the data, not with the internet.

The Answer Arrived on November 16

On November 16, Steve Pitchford, who works for Majestic, replied to the investigation. His explanation was short and honest: "Some internal experimentation. Choose a domain, make it esoteric, do some experiments. However, looks like we forgot to add a filter to avoid reporting it. We will sort that…"

The moral, as the post put it: don't test in production. Majestic had run internal experiments against a deliberately obscure domain, and the experiment data flowed into the public ranking because nobody added a filter to keep it out. The gibberish domain was never a real website, never a hack, and never a threat. It was a test that escaped.

The full investigation, with screenshots of every contradictory number, is in the original investigation. It is worth reading in full if you want to see how methodically a strange ranking can be taken apart.

How to Investigate Any Strange Domain: 5 Checks

You will meet strange domains in your own work: in server logs, in backlink reports, in referral traffic, in spam comments. Most are harmless. A few are real threats — phishing lookalikes, spam networks, or hacked sites linking to you. The five checks below, in order, separate the two. They are the same checks that cracked this mystery.

Check 1: Does it resolve?

Start with the simplest question: does the domain exist on the DNS at all? A DNS lookup shows whether the domain has records — A, AAAA, MX, TXT — and what they point to. A domain with no records at all cannot serve a website, receive mail, or do much of anything. That alone would have ended this mystery: w2rt6byl5nf1opmc-a.com returned dnsNotFound.

If a domain resolves sometimes and not others, or resolves differently from different places, the problem may be propagation rather than absence. Our guide to why a DNS change is not working walks through the usual causes, from TTL caching to nameserver mismatches.

Check 2: When was it registered?

A domain first registered last week cannot have earned backlinks last month. Registration dates come from WHOIS and RDAP records, and Certificate Transparency logs show when certificates were first issued for a name. In this case, the domain was first registered on November 11 — after most of its supposed backlinks already existed, and after Tranco had data from October 5.

That timeline contradiction is one of the strongest signals in domain research. Real authority accumulates over time. When the dates run backwards — links before registration, rankings before existence — the data is wrong, or the history is manufactured.

Check 3: What does the server say?

If the domain resolves, ask the server itself what it is. A check of the HTTP headers reveals the status code, the server software, redirect chains, and security headers — the fingerprints of whatever is actually answering. A domain that returns 200 with real content is a very different animal from one that redirects through three hops to a parked page.

For the mystery domain, this check never got to run: there was no server to answer. That absence was itself the answer. No headers, no status code, no content — nothing to investigate beyond the DNS failure.

Check 4: Do independent tools agree?

Never trust one tool's numbers. Majestic reported 241 million backlinks for this domain; Ahrefs reported 53. A disagreement of that size is not a rounding difference — it means one dataset is polluted, stale, or measuring something else entirely. When two reputable tools disagree wildly, believe neither until a third source or a manual check breaks the tie.

This applies to every SEO metric, not just backlinks. Rankings, traffic estimates, and keyword volumes all come from sampled, modelled data. Cross-checking is not paranoia; it is the basic hygiene this story exists to teach.

Check 5: Does the methodology explain it?

Finally, ask how the number is built. The Majestic Million ranks domains by the number of different referring subnets — Class C subnets — not by raw link counts. That design choice is meant to resist manipulation, but it also means a flood of links from many subnets moves a domain fast, and test data with broad subnet coverage moves it fastest of all.

Tranco, which ranked the domain at 38, uses Majestic data among its inputs, so the same pollution appeared in two lists that looked independent. Always trace a metric back to its source. Two lists that share a data feed are one source wearing two hats, and the Tranco list should be read with that in mind.

Backlink Audit Lessons for SEOs

The mystery is solved, but the habits it teaches are permanent. Four rules to take into your own work:

  • Never trust one metric. A backlink checker shows you one vendor's view of the web. Pair it with a second tool before you act on anything surprising, and treat any number that contradicts itself — nine anchors, 241 million links — as broken until proven otherwise.
  • Audit before you react. A backlink audit that starts with "disavow everything unfamiliar" wastes effort and can remove good links. Run the five checks first. Most strange domains turn out to be noise, test data, or dead sites — not attacks.
  • Question "millions of backlinks" claims. If someone sells you a package promising millions of links, remember that 241 million links pointed at a domain that never existed. Volume without verification is the oldest trick in low-quality SEO.
  • Authority scores are models, not measurements. A domain authority checker estimates strength from link data. When the underlying data is polluted, the estimate follows it. Use scores to prioritise, never to decide.

Majestic said it would add the missing filter, so the gibberish domain should disappear from future rankings. What remains is the reminder: the web's biggest datasets are built by people, and people forget filters. Check the data yourself, and you will rarely be fooled.

Frequently Asked Questions

What is the Majestic Million?

The Majestic Million is a ranking of the top million domains on the web, ordered by the number of different referring subnets — Class C subnets — linking to each domain. It is published by Majestic and is widely used as a proxy for a domain's link popularity.

How did a domain that doesn't resolve rank #8?

Majestic ran internal experiments against the esoteric domain and forgot to filter the test data out of its public reporting. The experiment generated backlink data across hundreds of thousands of subnets, and since the Million ranks by referring subnets, the domain shot to number 8. Majestic confirmed the cause and said it would fix the filter.

What is a backlink audit, and when should I run one?

A backlink audit is a review of the domains linking to your site: which are real, which are spam, and which deserve action. Run one when you see a sudden spike in referring domains, when a strange domain appears in your reports, or before disavowing anything. The five checks in this article are a good starting routine.

Why do Majestic and Ahrefs report different backlink counts?

Each tool crawls a different portion of the web, on its own schedule, with its own rules for what counts as a link. Small differences are normal. Differences of orders of magnitude — 241 million versus 53 — signal polluted, stale, or incomparable data, and mean you should verify manually before believing either number.

Should I worry if a strange domain links to my site?

Usually not. Most strange links are spam, scrapers, or test data, and search engines generally ignore low-quality links. Investigate with the five checks if the domain looks like a phishing copy of your brand or if you see a sudden flood of links. Otherwise, note it and move on — reacting to every odd link is a waste of a working day.