RankScore: 96 metrics in six pillars

Six pillars, 96 metrics

A metric that cannot be measured for an article says so and is left out of the score, never filled in with a made-up number. The page shows how much of each pillar was measured.

Checks that block publishing

A high score is not enough. An article is only marked ready when these pass:

• Every fact has a source that supports it
• First-hand claims say whose experience they are
• Nothing copied from a page that ranks
• No hidden or look-alike characters
• Health, money, legal and safety topics have a qualified reviewer
• Australian Consumer Law and TGA wording checks

The writer never grades its own work

Most metrics are computed from your text, your sources and the live search results. The judgement calls and the fact-check are made by a separate review that never sees the writer's instructions. Where the site runs a single AI model, the score says so.

Human, honestly

The Human pillar measures writing readers notice: varied rhythm, plain words, no stock phrases, your own voice. It is earned by better writing, never by tricks to fool an AI detector. On Pro, Agency and Enterprise plans you can run an outside detector as a second opinion.

Beside the score

Four panels explain the text from other angles without changing the score: all six readability formulas with what each one means, predicted engagement (reading time, words before the answer, title appeal), accessibility problems such as skipped headings or images without descriptions, and the technical checks on schema and meta tags. The free content scorer shows the readability and human-writing panels too.

"Beats 8 of the top 10"

We run the same checks, with the same code, on your article and on the full text of each top-10 page, and show how many score lower on those comparable metrics, before you publish. Checks that need your sources, your brief or a reviewer are not compared, so the number is never stretched.