Domain Authority History: Reading a Score Over Time

A single authority score is a snapshot, and on a dropped domain it is the least useful reading of one, because the links behind it were earned years before the drop and some of them have since gone. A history view answers two questions a snapshot cannot: was the authority earned steadily or in one burst, and is it still there. Both are checkable free, from archived snapshots by date and from referring-domain counts across successive crawl releases. What is not checkable is a vendor's own score history, because no vendor publishes one, and no third party can recompute it. In the Revised directory, as of 14 September 2026, the 409 listings on the aged hub had a median of 100 to 250 referring domains, against 10 to 50 across all 11,255 listings.

Why one number is the wrong reading

Two domains can read the same today for opposite reasons.

One earned forty references over twelve years from sites that still exist, was cited by a university in 2014 and still is, and lost a handful of links to redesigns. The other got two hundred links in a single month in 2019 from one network, most of which are now dead pages, and the vendor has not recrawled them.

Same score. Completely different asset. The snapshot cannot tell them apart, and the pattern of how a profile was built is exactly what a buyer is trying to read.

Google's own framing is the reason this is not pedantry. The ranking systems guide says ranking works mostly at page level:

Our ranking systems are designed to work on the page level, using a variety of signals and systems to understand how to rank individual pages. Site-wide signals and classifiers are also used and contribute to our understanding of pages. Having some good site-wide signals does not mean that all content from a site will always rank highly.

And on the scores themselves, from the March 2024 core update post:

Please note: some third-party services provide "reputation" or "authority" scores for sites. These don't correspond to any of Google's own signals nor come from Google.

So the number is a third party's estimate of one input, at one moment. Its trend is more informative than its value, and its trend is the thing tools mostly do not show you.

What can you check for free?

Three things, and each answers a different question.

Archived snapshots, by date. What the site was, when it changed, and when it stopped. This is the strongest free signal about a dropped domain and the only one that shows intent. A site that was a working business for a decade and then a parked page for two years reads completely differently from one that flipped subjects four times. Our expired domain checker shows the snapshot timeline alongside registry state.

Referring-domain counts across crawl releases. Successive public crawl releases are dated, so the same domain measured in two releases gives you a direction. Rising means references were still being added. Flat over years means the profile was finished long ago. Falling means pages holding the links are disappearing, which on a dead domain is the normal case.

Whether the links are still live now. Fetch the linking pages and look for the link. This is the check that separates a real profile from a historical one, and it is the one no score includes. Checking backlinks for free covers the method.

What cannot be reconstructed?

Be clear about the gaps, because plenty of pages imply otherwise.

Anyone showing you a domain's authority score over ten years is showing you their own tool's reading of their own crawl, which began when they started crawling, and which cannot be audited against anything.

How to read a history in practice

Six questions, in order, on a domain you are considering.

1. How long was it a site at all? First snapshot to last active snapshot. Under two years and the references are probably thin whatever the count says.
2. Did the subject stay the same? Sample snapshots across the span. Subject flips are the shape evaluating link history treats as a disqualifier.
3. Were references added steadily or in a burst? A burst in one month, from many domains, with similar anchors, is a purchased campaign. Google names "using automated programs or services to create links to your site" in its link spam policy.
4. Are the strongest references still live? Check the top five by hand. If the university page that carried the link was redesigned in 2022, the link may be gone and the score may not know.
5. How long has it been dead? Longer means more decay. This is the one variable that only moves against you.
6. Does the subject match yours? If not, stop. The rest of the reading does not matter.

That sequence takes twenty minutes per domain and it is the difference between buying a profile and buying a number.

Using the Revised directory for this

Every listing carries the history signals alongside the score, before the name is revealed. Practically:

1. Open the aged domains hub. It holds every listing at least five years old, which was 409 listings as of 14 September 2026, median 100 to 250 referring domains, 79% passing a low spam screen.
2. In History, set a minimum Age. Age is the proxy for how long a profile had to be built, and it is the only part of a history view you can filter on directly.
3. Still in History, set Checked to a recent window, so you are not reading the history of a name somebody already registered.
4. In Quality, set Referring domains rather than the score, then set Spam signal to the low band. A burst-built profile shows up here more often than on the score.
5. Compare Referring domains against the Revised Score on each row. A high count with a low score is the profile to look at twice: it is consistent with volume from places that carry little weight.
6. Read the archived snapshot history and the Linked by column on the shortlist, then run the finalists through the expired domain checker and the domain authority checker.

Filtering and sorting are free signed in or not. A plan buys reveals, so do the history reading first.

For the vendor differences behind any score you are comparing, how the authority metrics differ sets out what each one measures and why two tools disagree on the same domain.

FAQ

Can I see a domain's authority score history? No vendor publishes one, and no third party can recompute a historical score. What you can reconstruct is the direction, from archived snapshots and from referring-domain counts across dated crawl releases.

How do I check a domain's history before buying it? Read the archived snapshots by date for what the site was and when it changed, check whether the strongest referring pages still carry the link, and compare referring-domain counts across two crawl releases for direction.

Why does a dropped domain's authority score mean so little? Because the links behind it were earned before the drop and have been decaying since. The score reflects the last crawl, and on a dead domain the trend is downward.

Is a high score with few referring domains good or bad? It usually means a handful of strong references rather than volume, which count-based metrics undersell. It is worth a closer look rather than an automatic yes.

What does a high referring-domain count with a low score tell me? That the volume is coming from places that carry little weight, or that the pages holding the links have gone quiet. Open the linking domains rather than trusting the gap.

Do authority scores ever go down? Yes. They are recomputed on a crawl cycle, and a domain loses ground when linking pages are deleted, redesigned or re-rated. Our own score ships as versioned releases and can fall between them.

If reading a history rather than a snapshot is the point, the Revised directory of expired domains with backlinks shows archived snapshot history, referring-domain band, age and the named referring sources on every listing, all before the name is revealed.