DR vs Real Authority: Why DR Lies
Domain Rating, Domain Authority, Trust Flow — every major SEO tool ships a single number summarising how strong a domain is. They are useful for rough comparison and actively misleading when buying expired domains, because the number can be bought and the thing it approximates cannot.
What these scores measure
DR and DA are computed from the link graph. More referring domains, and links from domains that themselves have many links, produce a higher score. The formula differs between vendors but the input is the same: links pointing at the domain.
Nothing in that calculation asks whether a link was editorially given or paid for, or whether the linking site has any readers.
Why that is exploitable
Because the score is a function of the link graph, anyone who can add links to the graph can raise it. This is a service you can buy. A few hundred dollars gets a domain from DR 5 to DR 40 through a private blog network — a set of sites built for exactly this, with enough links between them to pass the threshold.
The result scores well and is worth nothing. Google's systems are considerably better at spotting these patterns than a link-counting formula is, which is why a "DR 50" expired domain can rank for nothing at all.
When you sort a drop list by DR descending, you are sorting inflated domains to the top. The manipulation is the reason they are there.
What cannot be bought
Some links are not purchasable at any realistic price:
- Wikipedia citations — reverted quickly when added promotionally, and editors watch the pages that matter
- Major news organisations — a link in a national newspaper comes from a journalist deciding to cite you
- Government and academic domains — a
.govor university link means an institution referenced your work - Established code repositories — a link from a widely used project's documentation reflects genuine usage
None are perfectly clean. Wikipedia spam exists, and paid placements happen in real publications. But the cost and difficulty are orders of magnitude higher than PBN links, which changes the economics enough that these links are reasonable evidence.
If a domain has a live Wikipedia citation, something happened that was hard to fake.
Practical difference
Two domains from a real drop list:
Domain A — DR 52, 890 referring domains. Sorted to the top. The referring list is largely expired blog comments, directory listings, and thirty sites sharing an IP block. No source you recognise.
Domain B — DR 18, 34 referring domains. Buried on page four. Among them: a citation from a university department page, two regional newspaper articles, and a link from a moderately popular open-source project's docs.
Domain A costs more at auction, because auctions price on the visible number. Domain B is the one with links that were genuinely earned.
How to check
Ignore the composite score. Open the referring domains list and read it.
Count the sources you recognise without looking them up. That number — not DR — is the useful signal. Then check the links are still live, because historical totals include links that no longer exist.
It takes longer than sorting a column. It is also the only version of the check that survives contact with a manipulated domain.
Related: what to look for in a domain tool and finding domains with real traffic.