What it does
The checker reads one published page and scores it out of 100 on how well a machine can use it. That covers the things an answer engine needs before it can quote you: a clear question in the heading, an answer close to it, sentences short enough to lift, structured data that says what the page is, and a source behind any figure.
It is the same scorer we run on client content. Not a cut-down version, and not a different set of rules with the same name.
Why scoreability, not readability
A readability score tells you whether a person can follow your writing. That matters, and it is one pillar of the six here.
The other five are about whether a machine can use the page: whether the heading matches a real question, whether the answer sits directly under it, whether the page declares its own type, whether claims carry sources, and whether the structure survives being chopped into passages. An engine quoting you does not read the page top to bottom. It lifts a passage. Scoreability measures how liftable your passages are.
What the score means
| Range | What it usually indicates |
|---|---|
| 80 to 100 | Structurally sound. Remaining gains are in depth and sourcing |
| 60 to 79 | Readable but hard to quote. Usually headings that describe rather than ask |
| 40 to 59 | The content is there, the structure is not. Most published B2B blog content sits here |
| Under 40 | An engine has little to work with. Often long paragraphs with no schema |
For reference, when we scored our own client work the average across ten recent posts on one account came out at 67, losing points on the same three things every time: reading ease, sentence length, and no FAQ section.
What it does not do
It does not tell you whether you rank, or whether an engine cites you today. Those are measured separately, by tracking citations across engines over time. A page can score 90 and never be quoted because nobody is asking the question it answers.
It does not judge whether the content is any good. It measures structure and clarity. A well-structured page saying nothing scores well.
It scores one page at a time. Estate-wide scoring, change tracking and citation panels are what the paid engagement covers.
The full rubric, written up
Every check and what it is worth is published in How to Improve Readability Scores, along with the four checks that cost the most points and the fixes that recover them.
How the score is calculated
Every criterion is shown in the result rather than rolled into a number you have to trust. If a page loses 15 points, the tool names which pillar lost them and what would recover them.
The rule we hold ourselves to on this site applies to the tool as well: no figure without a method behind it. If the checker cannot explain a deduction, the deduction is a bug.