Readability scores turn one judgment — 'is this easy to read?' — into a number. They work by measuring surface features of your text: sentence length, word length, and syllable counts. Longer sentences and longer words push the score down. That's crude, but it's remarkably predictive, because short sentences and common words are exactly what makes text easy for a broad audience to parse.
The two scores you'll see most
- Flesch Reading Ease — scored 0–100; higher is easier. 90–100 is children's-book simple, 60–70 is plain English for general readers, and below 30 is academic-legalese territory.
- Flesch-Kincaid Grade Level — converts the same measurements into a US school grade. A score of 8 means an 8th grader (age ~13–14) can read it comfortably.
For most online content — product pages, help articles, marketing copy — the sweet spot is a Reading Ease around 60–70 and a grade level of 8 or below. That's not dumbing down; it's matching the reading level of the general population and the reality that people skim on phones.
What moves the needle
- Sentence length is the biggest lever — split long sentences at the second independent clause.
- Replace multi-syllable words with plain alternatives where the meaning doesn't suffer ('utilize' → 'use', 'commence' → 'start').
- Cut filler — every 'in order to', 'it should be noted that', and 'as a matter of fact' lengthens text without adding meaning.
- Keep domain terms. A medical or legal article needs its technical vocabulary; readability scores punish it, but your audience needs the terms. This is the main reason to treat the score as a guide, not a target.
Scores don't measure clarity
You can write gibberish with short words and score a perfect 90. Readability measures ease, not sense. Always read the text for meaning and flow — the score is a check, not the editor.
Score your own writing
Paste your draft into Writing Studio for live word count, Flesch scores, keyword density, and a Hemingway-style readability analysis — all in your browser.
Open Writing StudioWhy two formulas disagree about the same text
Run one paragraph through Flesch-Kincaid, Gunning Fog, and Coleman-Liau and you will usually get three different grade levels. That is expected, not a malfunction. Each formula weights its inputs differently: Flesch-Kincaid counts syllables per word, Gunning Fog counts the percentage of words with three or more syllables, and Coleman-Liau ignores syllables entirely and works from characters per word — which makes it more reliable on text where syllable counting is unreliable.
The practical rule is to look at direction rather than any single number. If all three say the text is heavy, it is heavy. If they disagree by a couple of grades, that spread is normal and not worth chasing. Picking whichever formula flatters your draft is possible and pointless.
The blind spots worth knowing
Syllable counting is approximate. These formulas use heuristics based on vowel groupings rather than a pronunciation dictionary, so they miscount irregular words and proper nouns routinely. A document full of unusual names or technical terms gets a harsher score than its actual difficulty warrants.
More importantly, no readability formula reads word order. "The cat sat on the mat" and "mat the on sat cat the" score identically, because both have the same words and the same sentence length. Nothing in the calculation checks whether sentences follow logically, whether pronouns have clear referents, or whether the argument holds together — which are the things that actually make writing hard to follow.
This is why the score is best used as a flag rather than a target. A grade level far above your audience usually means sentences have grown long enough that readers lose the thread partway through — worth investigating. Chasing a lower number by mechanically chopping sentences produces choppy, disconnected prose that scores well and reads badly.
Choosing a target for your audience
For general public writing, a grade level around 8 to 10 is a reasonable aim — not because readers are incapable of more, but because lower cognitive load means more people finish reading. Government and healthcare guidance often targets lower still, because comprehension failures there carry real consequences.
Technical documentation written for specialists can sit considerably higher without being a problem. If your readers know the vocabulary, using the precise term is clearer than paraphrasing around it, even though the formula penalises the longer word. The question is never "what score should this hit" but "does this match the person reading it".