domain authority, domain rating, trust flow, citation flow, and our own cg authority are five 0-to-100 numbers sold as the same thing. they are not the same thing, they are not convertible, and none of them is google's. each is one company's prediction, computed with its own formula over its own crawl. this page is what each one actually measures and how to read them without fooling yourself.
google does not publish or use a site-wide authority score of this kind. every number on this page is a third-party estimate of an outcome, not an input to it. that is why raising one is not a strategy: you would be optimising somebody else's guess.
the five, side by side
| scale | vendor | what it is computed from | |
|---|---|---|---|
| Domain Authority (DA) | 0-100, logarithmic | Moz | a model trained to predict ranking strength, over Moz's Link Explorer index |
| Domain Rating (DR) | 0-100, logarithmic | Ahrefs | the strength of a domain's inbound link profile, over Ahrefs' own crawl |
| Trust Flow (TF) | 0-100 | Majestic | link volume weighted by proximity to a curated set of trusted seed sites |
| Citation Flow (CF) | 0-100 | Majestic | link volume alone, with no trust weighting |
| CG Authority | 0-100, log-rank | CrawlGraph | a percentile of Common Crawl's published harmonic centrality, on open data |
why they never agree
two reasons, and they compound. the first is the input: each score is computed over a different crawl, and no crawler sees the whole web, so the underlying link sets differ before any math runs. the second is the formula, and these are genuinely different kinds of thing. a model trained to predict rankings and a centrality measure over a graph are not two estimates of one quantity; they are answers to two questions.
so a site scoring 45 on one and 62 on another is not evidence that one vendor is wrong. it is what you should expect. the graph math behind centrality-style scores, including the exact formula behind ours, is in what is a backlink graph, which is the right page if you want to see how the numbers are actually produced rather than what they are for.
the logarithmic scale is the most misread part
these scales are logarithmic, which has a consequence people consistently get wrong when setting goals. the distance from 20 to 30 is not the distance from 60 to 70. the upper end compresses hard, so each additional point costs dramatically more than the last.
this is why "get our DA to 50" is a bad objective. it is a target on someone else's scale whose difficulty is not linear and whose definition can be revised by its owner without warning, and it has been revised before.
what counts as a good score
there is no absolute threshold, and a page that gives you one is guessing. the number only means something in comparison, so make the comparison the useful one: look at the scores of the sites currently ranking for the queries you want. that tells you what the bar is in your subject, which is the only version of the question with an answer.
the other comparison worth making is between Trust Flow and Citation Flow on the same domain. a large gap, plenty of citation and little trust, describes a link profile with volume and no credibility behind it. that gap is more informative than either number alone, and it is the one genuinely diagnostic reading in this whole family.
how to actually use them
- sort with them, do not decide with them. all of these are good enough to rank a list of candidate domains from strong to weak. none is precise enough to justify a decision resting on a few points.
- never compare across vendors. a DR from one report and a DA from another are different units. putting them in one column produces a spreadsheet that looks rigorous and means nothing.
- quote whichever one your counterparty uses. if a client asks about DA, answer about DA. arguing about which vendor metric is truer is a conversation with no destination.
- treat a sudden jump as a recalculation. vendors update their models and everyone's number moves. that is a change in the estimate, not in your site.
what to measure instead
the metric that survives all of this is not a score, it is a set: which credible domains link to your competitors and not to you. it needs no model, it cannot be revised by a vendor, and every entry is an action rather than a number. that is a link gap, and it is the thing an authority score is a lossy summary of.
faq
What is the difference between Domain Authority and Domain Rating?
Domain Authority is Moz's 0-100 score and Domain Rating is Ahrefs' 0-100 score. They are different formulas computed over different crawls by competing companies, so they are not convertible and a site routinely scores differently on each. Both are predictions of ranking strength on a logarithmic scale, and neither is used by Google.
Does Google use Domain Authority?
No. Domain Authority is a Moz metric, Domain Rating is an Ahrefs metric, and Trust Flow and Citation Flow are Majestic metrics. Google has repeatedly said it does not have a website-wide authority score of this kind that it uses in ranking. These scores are third-party estimates of the outcome, not inputs to it.
What is a good Domain Authority score?
The question has no absolute answer because the scale is logarithmic and relative. Moving from 20 to 30 is a far smaller achievement than moving from 60 to 70, and a score that is strong in a niche subject would be weak in a competitive commercial one. The useful comparison is against the sites currently ranking for the queries you want, not against a number.
What is the difference between Trust Flow and Citation Flow?
Citation Flow is Majestic's measure of how much link volume points at a site, and Trust Flow weights that by proximity to a curated set of trusted seed sites. The gap between the two is the point of having both: high Citation Flow with low Trust Flow describes a profile with plenty of links and little credibility behind them.
Which authority metric should I use?
Whichever one your counterparty quotes, and only for sorting rather than for deciding. All of these scores are reasonable at ranking a list of candidate domains from strong to weak, which is what you actually need when choosing outreach targets. None of them is precise enough to justify a decision that hinges on a few points of difference.
writes the queries we run internally. ships one tactical post a week.
plus one when a new common crawl release lands. that is all.