The short version
- Domain Authority is a Moz metric, not a Google one. Google does not use it and has said so repeatedly.
- It is a prediction of ranking ability built from link data — useful as a rough comparative signal, misleading as a target.
- Every vendor has its own version: Moz DA, Ahrefs Domain Rating, Majestic Trust Flow. They disagree, and all are estimates.
- For a real decision, look at the specific pages ranking for your specific query rather than at any domain-level score.
Domain Authority has achieved something unusual: it is the metric everybody in and around SEO understands, including people who know nothing else about it. Clients ask about it. Link sellers price against it. Agencies report it monthly. And it is a third-party prediction that the search engine it predicts does not use.
That combination — universally understood, not actually a ranking factor — makes it genuinely useful and genuinely dangerous, depending on what you do with it.
What it actually is
Moz built a model that predicts how likely a domain is to rank, trained largely on link data. The output is a 1–100 logarithmic score. Ahrefs has Domain Rating; Majestic has Trust Flow and Citation Flow. Each is computed from a different index with a different model, which is why they disagree about the same site.
None of these is used by any search engine. They are commercial estimates of a property search engines assess with data nobody outside them has.
Majestic’s split is the most informative design of the four, because separating trust from volume distinguishes a site with a thousand poor links from one with fifty excellent ones — a distinction a single composite score cannot express.
Where it is genuinely useful
- Rough comparison at scale. Triaging a list of 200 link prospects, a score is a reasonable first filter.
- Trend on your own domain. Rising over quarters suggests your link profile is strengthening.
- Stakeholder communication. A number non-specialists already understand has real practical value.
- Spotting obvious problems. A very low score on an established domain is worth investigating.
Where it misleads
- As a ranking target. Raising DA is not a goal; it is a side effect of things that are.
- Comparing across niches. A DA 40 site in a small technical niche may outrank a DA 70 generalist for a specific query.
- Buying links by score. Pricing link placements on DA created an entire industry of manufactured scores. Google’s spam policies address the underlying practice.
- Page-level decisions. Domain-level scores say nothing about whether a specific page can rank for a specific query.
The metric is gameable and gets gamed
Because link vendors price on DA, there is a direct financial incentive to inflate it, and techniques for doing so are well established. A high score on an unfamiliar domain is a reason to look more carefully, not less. Check whether the site has real traffic and real editorial output before valuing its links at all.
What to look at instead
- The actual pages ranking for your actual query. Not domain scores — the specific competing pages.
- Whether those pages are better than yours, honestly assessed. Usually more informative than any metric.
- Your own ((google-search-console|Search Console)) data — impressions and positions are real, not modelled.
- Referring domains over time, which is a fact rather than a prediction.
- Whether links come from places your customers actually read.
None of those requires a subscription to an authority metric, and all of them survive the next change to how a vendor computes its score. That durability is the point: a metric that can be recalculated by its owner overnight is a poor foundation for a strategy, whereas "are the pages outranking us actually better" is a question whose answer does not depend on anyone’s model.