What the checker actually scores
The tool takes whatever text you paste in and works out a reading age, using a formula called Flesch-Kincaid (a measure of reading difficulty based on average sentence length and the number of syllables per word). It also reports the number of sentences, the average words per sentence, and how many sentences are written in the passive voice, where the subject of the sentence has something done to it. Together these give you a rough map of how hard your copy is to read.
A reading age of 9 means a competent nine-year-old reader could follow the text without much strain. That is not an insult to your adult readers. Most people, whatever their education, read faster and retain more when the sentences are shorter and the words are familiar, which is why plain writing tends to perform better even for a technical audience.
Reading the number against the job the copy has to do
The right reading age depends entirely on what you are writing and who is reading it. A consumer product page selling kitchen knives can sit at a reading age of 8 or 9 without losing any authority. A page explaining pension drawdown rules to existing clients can reasonably run higher, because the subject itself requires longer sentences to stay accurate, and simplifying past a certain point starts losing meaning.
So the score is a starting point for a decision. If your homepage comes back at a reading age of 16 and you are selling to the general public, that is worth investigating. If your reading age is 16 because you are writing for actuaries, it may be exactly right.
What the score cannot tell you
Flesch-Kincaid counts syllables and sentence length. It does not know what the words mean, so it cannot tell you that a sentence is confusing because it uses a term the reader has never encountered, provided that term happens to be short. "The APR excludes any early repayment charge" scores as easy reading, because every word is brief, and yet a reader unfamiliar with APR (annual percentage rate, the yearly cost of borrowing including fees) will not follow it any faster for that.
The reverse also happens. Breaking a sentence into three short fragments lowers the score, but three short fragments in a row read like a list of instructions barked at the reader. A slightly longer sentence that uses because or so to join two related ideas is often easier to follow than the same idea chopped into pieces, even though the checker will mark it down for length.
None of this makes the score useless. It means the number is a prompt to look at the sentence. Where a term is doing real work in your copy, explain it in plain words first and put the technical term afterwards in brackets, the way this guide has done with Flesch-Kincaid and APR above. The glossary holds a working set of these terms if you want the reader-facing version of a phrase you use often.
What to fix first when the score comes back poor
Work through the flagged issues in this order, because each one has a bigger effect on how the copy reads than the next.
- Long sentences carrying more than one idea. Split them at the natural join, usually where a second thought was joined on with "and" or a comma. "We deliver next-day and offer free returns and our support team is available seven days a week" becomes two sentences: one about delivery and returns, one about support.
- Passive constructions that hide who is doing the work. "Your order will be processed within 24 hours" tells the reader something happens, but not who does it. "We process your order within 24 hours" says the same thing with an actor attached, and reads faster because the reader is not left waiting to find out who acted.
- Unexplained jargon. Any term your reader would not use themselves, defined in plain words on first use.
- Long runs of short, choppy sentences. Three fragments in a row read like a checklist. Rejoin one pair with "because" or "so" where the ideas are genuinely connected.
The judgement the tool leaves to you
A readability score measures the arithmetic of a sentence: its length, its syllables, whether its subject is doing something or having something done to it. It has no way of judging tone, whether a claim needs a source, or whether a sentence that scores well is actually true. Those calls stay with the person writing the copy, and no score, however low, substitutes for reading the paragraph aloud and asking whether it says what you meant.
Run your draft through the readability and clarity checker as a first pass, then read the result yourself before you publish it. Treat a low reading age as a target worth aiming for on most consumer copy, and treat a high one as a question worth asking. For copy where the words carry legal weight, such as advertising claims, a passing readability score is no guide to whether the claim itself will stand up. That is a separate check, covered in the UK advertising rules guide.
Using it alongside the rest of your toolkit
Readability is one dimension of copy quality. A sentence can score well and still open weakly, undersell the benefit, or ask the reader for the wrong action. The headline scorer and the call-to-action tester look at those questions specifically, and using them together with the readability checker gives you a fuller picture than any one tool on its own. None of them replaces reading the copy back and asking whether it sounds like something a person would actually say.