"What's the readability of this?" sounds like one question, but the answer depends on which of half a dozen formulas did the asking. Flesch, Flesch-Kincaid, Gunning Fog, and SMOG all claim to measure how hard a text is to read, they all return different numbers on different scales, and they're all built from the same two ingredients. Once you see the pattern, every score becomes readable itself.
What readability formulas actually measure
Every formula in this guide is a proxy: instead of testing comprehension with real readers, it counts two surface features that correlate with it — average sentence length (longer sentences hold more clauses in the air) and word complexity, usually measured in syllables per word (longer words are rarer and slower to decode). Rudolf Flesch built the founding formula in 1948, Robert Gunning adapted it for business in 1952, and the rest are variations on his insight: sentence length plus syllable count predicts reading difficulty surprisingly well.
The key word is predicts. These formulas never look at meaning, logic, organization, or whether your argument is any good — they measure the mechanical load a text places on a reader, which is necessary for clarity but nowhere near sufficient. Keep that caveat in your pocket; we return to it at the end.
Flesch Reading Ease: the 0–100 classic
The original and still the most quoted. Flesch Reading Ease combines average sentence length and average syllables per word into a score from 0 to 100, where higher is easier. The rule of thumb most editors use: 60–70 is plain English, roughly what a 8th–9th grader handles comfortably — and the right default target for most business and web writing. The full ladder:
- 90–100: very easy (5th grade) — children's books, simple instructions
- 70–89: easy (6th–7th grade) — popular fiction, consumer content
- 60–70: standard / plain English (8th–9th grade) — the sweet spot for most writing
- 50–59: fairly difficult (10th–12th grade) — needs motivated readers
- 30–49: difficult (college level) — academic and legal prose often lives here
- 0–29: very difficult (graduate level) — the register of journals and tax codes
Because the scale is continuous and familiar, Flesch Reading Ease works well as an everyday dashboard number: paste a draft, glance at the score, and you immediately know whether you're writing for everybody or only for specialists.
Grade-level formulas: Flesch-Kincaid and ARI
In 1975 the US Navy wanted Flesch's math expressed in a unit commanders could act on: school years. The result, Flesch-Kincaid Grade Level, converts the same two inputs into a US grade number — a score of 8.4 means roughly an 8th-grade reader can follow along. That framing made it the most widely deployed readability number in the world: Microsoft Word has shipped it since the 1990s, and it's the default score in most word processors and learning platforms. We break down its formula, its blind spots, and how to move the number in our dedicated Flesch-Kincaid guide.
Its close cousin, the Automated Readability Index (ARI), grew out of 1960s US Air Force research and does the same job with one swap: it counts characters per word instead of syllables, because characters are trivial for software to measure. The two scores usually land within a grade or so of each other — if one drifts far from the other, suspect an unusual vocabulary rather than a real difference in difficulty.
Gunning Fog and SMOG
Gunning Fog (1952) was built for journalism and business: Robert Gunning made his name rewriting bureaucratic prose into English. The Fog Index reports the years of education a reader needs — 8 means an 8th-grade education suffices — and it counts as "complex" every word of three or more syllables. Its bias is toward hard words: a text with short sentences but heavy jargon still fogs up, which is exactly the failure mode Gunning cared about. Under 12 is the accepted ideal for business writing; anything above 16–18 reads as institutional.
SMOG — formally Simple Measure of Gobbledygook, from G. Harry McLaughlin in 1969 — counts polysyllables (words of three or more syllables) in a sample of ten sentences each from a text's start, middle, and end, then converts the square root of that count into a grade level. Its sampling method and its validation against clinical comprehension studies made it the standard in healthcare plain-language work, where the standing target is roughly grade 8 for patient materials — consent forms, discharge instructions, medication guides. If you write for the public in a regulated industry, SMOG is likely the score your compliance team will ask for.
The big comparison
Same inputs, different dialects. Here is the whole family on one card:
| Formula | Scale | Sweet spot | Best used for |
|---|---|---|---|
| Flesch Reading Ease | 0–100, higher = easier | 60–70 | General web and business writing |
| Flesch-Kincaid Grade Level | US grade years | 7–9 | Education and technical documentation |
| Automated Readability Index | US grade years | 7–9 | Quick automated checks |
| Gunning Fog Index | Years of education | Under 12 | Business and government communication |
| SMOG Index | US grade years | About 8 | Healthcare and plain-language work |
Caveats — and which score to use when
The honest limits first. Readability formulas cannot judge the clarity of your ideas: a text can score a Flesch 75 and still be nonsense, because none of these measures knows what your words mean. Short sentences aren't automatically clear either — chop everything to five words and you get a staccato fog, plus a score that flatters you. And syllable counts misjudge genuinely common long words ("everything," "September") while excusing short, rare ones ("roan," "gyre"). Treat every score as a smoke test: a great place to find friction, a terrible place to find truth.
Practical defaults, then: writing for the general public — aim Flesch 60–70. Writing curricula, documentation, or anything with a stated audience grade — watch Flesch-Kincaid. Writing internal business comms — keep Gunning Fog under 12. Writing patient or public-health materials — target SMOG near grade 8. And don't chase one score alone: run the family together, because when Flesch praises you but Fog doesn't, the diagnosis is specific — your sentences are short but your vocabulary is heavy. WordlyCount computes all of these live in the analyzer as you type, so the comparison costs nothing. One last distinction worth making: if your worry is how long a piece takes to read rather than how hard it is, that's reading speed — a different measurement with its own average reading speed guide.