Readability scores explained: every formula in plain English

Most readability scores were invented before computers existed. They have survived because the answer they give is genuinely useful and because nobody has replaced them with anything better. Honestly, if you write for the web or the public sector or any audience that did not study your topic at university, knowing how these scores work is one of the cheapest investments you can make in your writing.

This guide walks through every formula that still matters. For each one you get the maths in plain English, what the number actually tells you and where it falls down.

What is a readability score?

A readability score is a number that estimates how hard a piece of writing is to read. It does not measure whether the writing is correct or interesting or true. It measures structure: how long the sentences are, how long the words are and how many of those words come from a list of familiar everyday English.

Different formulas weigh those signals differently. None of them read the text the way a person does. They count syllables and tokens, multiply by a constant and return a single number. That is the whole trick. It is also why a fluent paragraph about quantum field theory will still score badly.

The six readability formulas you actually need to know

Web tools quote dozens of variants. In practice almost every score you will see is one of these six or a tweak of one.

Flesch reading ease

The one you have probably already met. Rudolf Flesch published it in 1948 to help US government writers produce clearer documents. Microsoft Word built it in, Yoast SEO surfaces it inside WordPress and almost every readability tool starts with it.

The formula is 206.835 minus (1.015 times average words per sentence) minus (84.6 times average syllables per word). The output sits on a 0 to 100 scale where higher means easier. A score of 60 is roughly a UK reading age of 13 to 15, which matches a national tabloid news article. A score of 30 reads like a peer-reviewed paper.

For a public website aim for 60 or above. For a health or government or legal-services site that has to reach the whole population, push for 70.

Flesch-Kincaid grade level

The US Navy commissioned Kincaid in 1975 to rework Flesch into a number that maps to school grade level. Same inputs as Flesch reading ease, different output. A score of 8 means “the average eighth-grader can read this”, which roughly equals UK Year 9.

Use Flesch-Kincaid when you need to defend a target to stakeholders who think in school years. “Year 9 reading level” lands better in a committee meeting than “ease score of 65”.

Gunning fog

Robert Gunning was a publishing consultant who thought business writing was failing its readers. He built the fog index in 1952. The formula multiplies the sum of (average sentence length) and (percentage of words with three or more syllables) by 0.4.

Fog is more sensitive to long sentences than Flesch is. If your draft is full of clause stacks, fog will catch it where Flesch is more forgiving. Aim for fog 12 or below for public writing. Push for 8 if you need to be properly accessible.

Dale-Chall

Dale and Chall took a different route. Instead of counting syllables they built a list of 3,000 words that fourth-graders in 1940s America reliably understood. The formula compares your text against that list. If your writing uses words outside the list, the score climbs.

The good thing about Dale-Chall is that it reflects real vocabulary load rather than syllable counts. The bad thing is that the familiar-word list is dated and US-centric. “Lorry” and “petrol” register as unfamiliar to Dale-Chall even though every UK seven-year-old uses them.

SMOG

The Simple Measure of Gobbledygook, published by Harry McLaughlin in 1969, was built for healthcare writing where the cost of being misunderstood is high. SMOG takes three samples of 10 sentences each, counts the polysyllabic words across all 30, takes the square root and adds 3.

SMOG is the formula the NHS plain English guide quietly favours. It is harsher than Flesch-Kincaid by a year or two of schooling, so if you are writing health information and want to be safe, hit SMOG 8 or below.

Spache

George Spache built his formula in 1953 for content aimed at children below US fourth grade. It uses a different familiar-word list (smaller and more concrete) than Dale-Chall. If you write for early-reader audiences, use Spache. For anything older than around age 10, ignore it.

Which score should you actually use?

Match the formula to the audience. There is no universally best score.

• General public, blog, marketing site: Flesch reading ease. Aim for 60 or higher.
• Government, charity, plain English: SMOG or Flesch-Kincaid. Aim for grade 8 or lower.
• Legal, financial, technical documentation: Gunning fog. Aim for 12 or lower if the audience is informed adults.
• Children’s content under age 10: Spache.
• Children’s content above age 10 or general adult content: Dale-Chall.

What readability scores will not tell you

The scores are blunt instruments. They miss the things that actually decide whether a reader finishes the page.

They cannot judge meaning. A sentence can be short and clear and completely wrong. A formula will give that sentence a high score and walk away whistling.

They cannot tell you whether a word is jargon. “Bovine” and “horse” both score the same number of syllables but only one of them needs explaining. This is the gap where a tool like WP AI Explainer earns its keep. It lets readers tap the words they do not know without you having to rewrite the article for the lowest common denominator.

They cannot tell you whether you have lost your audience emotionally. Boring is not the same as hard. A score cannot see boring.

And they cannot tell you whether the structure of the page helps a reader skim. Subheads and bullets and bold key terms change how a real reader experiences difficulty. None of them touch the score.

How to actually improve a low readability score

You do not improve readability by running the score and panicking. You improve it by editing in this order.

Cut long sentences first. Find every sentence longer than 20 words. Look for “and” or “but” or “, which” that joins two complete thoughts and break them apart. Most sentences over 25 words are two sentences pretending to be one.

Then swap long words for short ones. Strike “utilise” for “use”, “commence” for “start”, “facilitate” for “help” and “approximately” for “about”. Every one of those swaps drops syllables without losing meaning.

Then leave the jargon that has to stay and give the reader an escape hatch. If you are writing about a regulated industry and “fiduciary duty” cannot be replaced, leave it in. WP AI Explainer lets any reader highlight that word and see a plain-English tooltip without leaving the page. Your score does not change. Your reader’s comprehension does.

Putting it into practice on a WordPress site

Three steps move a typical WordPress blog from “we have no idea” to “we know our readability”.

Switch on Yoast’s readability tab if you have not already. It gives you a per-post Flesch score in the editor. The score is rough but the per-paragraph and per-sentence flags are useful.

Run the post you are about to publish through readable.com for a second opinion. If Flesch is above 60 and Gunning fog is below 12, you are in good shape.

Install WP AI Explainer. It will not change your scores. It will close the gap between your scores and what your readers actually understand, which is the thing readability was always trying to be a proxy for.

Frequently asked questions

What is a good readability score?

For a public website, aim for a Flesch reading ease of 60 or higher, which roughly matches a US reading grade of 8 or lower. Government and healthcare sites should target 70.

Which readability formula is best?

There is no single best formula. Flesch reading ease is the most widely cited. Gunning fog is more sensitive to long sentences. Dale-Chall uses a list of familiar words and is better for early-reader content. Pick the one that matches your audience.

Do readability scores measure comprehension?

No. Readability scores only measure structural difficulty (syllable counts, sentence length, common word lists). They cannot tell you whether the meaning of the writing is clear or correct.

How do I check the readability of my WordPress posts?

Yoast SEO shows a Flesch reading ease score in the editor. For a deeper check, paste the post into readable.com or hemingwayapp.com. To help readers who hit hard words, install WP AI Explainer so they can highlight any word for a plain-English tooltip.