Readability checker for help articles and emails

How to use the readability checker

1. Paste the text. Use plain text, with a blank line between paragraphs. Remove menus, footers and code blocks first, because they change the score.
2. Read the two scores. Flesch Reading Ease (higher is easier) and the Flesch-Kincaid grade (lower is easier).
3. Fix the listed sentences. The checker lists sentences over 25 words and sentences that look passive. Rewrite those first: they move the score most.

The formulas

Both scores use the same two numbers: average words per sentence and average syllables per word.

• Flesch Reading Ease = 206.835 − 1.015 × (words ÷ sentences) − 84.6 × (syllables ÷ words)
• Flesch-Kincaid Grade = 0.39 × (words ÷ sentences) + 11.8 × (syllables ÷ words) − 15.59

Syllables are counted with a rule based on vowel groups, as most checkers do. English spelling defeats every rule sometimes, so treat a difference of a few points as noise.

Targets for support content

• Help articles: Reading Ease 60 or more, grade 8 or lower.
• Customer emails and chat replies: 70 or more. People read them on a phone, often while annoyed.
• Developer docs: scores run lower because of technical terms. Keep sentences short to compensate.

Our own documentation follows ASD-STE100 (Simplified Technical English), a standard written for maintenance manuals. One of its rules is a limit of 25 words per sentence, and our docs build fails when a sentence is longer. This checker uses the same limit.

How to raise the score

1. Split long sentences. One idea per sentence. A sentence with "and", "which" or a semicolon is usually two.
2. Use shorter words. "Use" for "utilize", "help" for "facilitate", "about" for "approximately".
3. Prefer active voice. "We send the invoice" is clearer than "The invoice is sent", because it says who acts.
4. Put the answer first. It does not change the score, but it changes whether the reader stops reading too early.
5. Use lists for steps. Numbered steps are shorter sentences by design.

A before and after example

Before (one sentence, 35 words, passive, grade 16):

Refund requests that are submitted after the 30-day period has elapsed will be reviewed on a case-by-case basis by our billing team, which may take up to ten business days, and approval is not guaranteed.

After (four sentences of 5 to 9 words, active, grade 2):

You can ask for a refund after 30 days. Our billing team checks each request. This can take up to ten business days. We cannot promise a refund.

Same facts, fewer words. Paste both into the checker to see the difference: the second version scores in the "very easy" band, and each fact can be quoted on its own.

What a score cannot tell you

A formula sees only lengths. It cannot tell whether a sentence is correct, whether a technical term is necessary, or whether the text answers the question. Short sentences full of jargon still score well. Use the score to find hard passages, then read them aloud.

Why clear text matters for AI answers

AI assistants and support chatbots quote passages from your pages. A long sentence that mixes a rule, an exception and a link is easy to quote out of context. Short sentences with one fact each are easy to quote correctly. Check a rewritten article with the AI answer generator: paste it as the source and ask a real customer question.