What the Plain Language Analyzer does
This analyser checks a draft against published plain-language guidance and ties every finding to the guideline behind it: complex words with simpler alternatives from the US federal plainlanguage.gov list, the words GOV.UK tells its writers to avoid, hidden verbs such as "make a decision", "shall" used for requirements, double negatives, passive voice, long sentences and long words. It also estimates the reading grade with the Flesch-Kincaid formula and compares it with a target for your audience.
It is built for letters, forms, policies and web pages that ordinary people have to act on. Each highlight links to the rule it comes from, so you can see why it was flagged and decide whether it applies. It runs live in your browser; the text is not sent anywhere.
How to use it
- Paste your draft, or load the benefits letter or policy update example.
- Choose the audience. General public flags sentences over 25 words and targets reading grade 8; service instructions are stricter at 20 words and grade 7; professional readers allow 30 words and grade 12.
- Read the summary by guideline. Each line gives the number of findings and a link to the guidance it comes from.
- Work through the highlighted text or the table. Each finding shows a plainer alternative where one exists and any exception - GOV.UK allows "deploy" for software, for example.
- Add your organisation's jargon under Your own jargon list, with a plainer word after an equals sign, then download the CSV for your editor.
Reading the results
The reading grade is the Flesch-Kincaid grade level, an estimate of the US school grade needed to follow the text. It depends only on sentence length and syllables per word, and syllables are counted by a rule of thumb, so treat it as a guide, not a measurement of how well real readers understand the text.
Flesch reading ease runs from about 0 (very hard) to 100 (very easy). Plain English for the general public usually sits around 60-70.
Findings are prompts. Some suggested words will not fit your sentence, and a correct technical term your readers know is fine. The goal is text your readers can understand and act on the first time they read it.
Worked example: a benefits letter with every classic problem
Load the benefits letter. It has 67 words in 5 sentences, counting the greeting, and an estimated reading grade of 11.8 against a target of 8. Its Flesch reading ease is 34 - "difficult".
The check finds 7 words with simpler alternatives: pursuant to (under), in accordance with (under), prior to (before), additional (more), comply with (follow), notify (tell) and immediately (now). "Undertaken a thorough assessment" hides the verb "assess", "shall provide" should be "must provide", "not uncommon" is a double negative, and "to be requested" is a passive that does not say who will ask.
Seven long words - such as eligibility, documentation and circumstances - are also listed. None of the sentences is over 25 words, which shows why the grade is high: the problem in this letter is word choice, not sentence length.
How this differs from the Filler Word and Cliché Detector
The Plain Language Analyzer is about comprehension: difficult or official-sounding words, hidden verbs, requirements, sentence length and reading grade. The Filler Word and Cliché Detector is about padding and tone: filler, clichés, hedges, weak qualifiers and intensifiers. A letter can pass one and fail the other, so they work well together.
Which rules come from where
Simpler-word suggestions are adapted from the plainlanguage.gov list of simple words and phrases. The GOV.UK list is the "Words to avoid" section of the GOV.UK A to Z style guide, including its exceptions. Hidden verbs, "must" for requirements, double negatives, active voice and short sentences are plainlanguage.gov guidelines. The long-word rule (four or more estimated syllables) and the audience thresholds are A2Z defaults, stated so you can judge them.
Limitations: what the result does not prove
- It checks English only, and the word lists mix US and UK guidance - both spellings of words like "utilize/utilise" are covered.
- Readability formulas measure sentence and word length, not meaning. A text can score well and still be unclear, and a short sentence full of jargon can score better than it deserves.
- It cannot tell whether a flagged word is used in a sense the guideline allows, such as "leverage" in finance. Exceptions are shown so you can decide.
- The best test of plain language is a real reader. Use this to prepare a draft for that test, not to replace it.
Privacy: where your data goes
Everything you paste, type or drop is processed in this browser tab. It is not uploaded, logged, stored or sent to analytics. Session recording and tag-manager scripts are switched off on this page.
Standards and sources
- US federal plain language guidelines (plainlanguage.gov) - checked 19 Sep 2026
- GOV.UK style guide - checked 19 Sep 2026
- plainlanguage.gov - Use simple words and phrases
- plainlanguage.gov - Use "must" to indicate requirements
Frequently asked questions
What reading grade should public information aim for?
There is no single official figure. Some public-sector guides, GOV.UK among them, aim for a reading age of around 9 - roughly US grade 4 - for information everyone needs. This page's general-public target of grade 8 is a pragmatic default that you can compare against your own standard.
Why should I write must instead of shall?
"Shall" is ambiguous: lawyers and readers disagree on whether it means a duty, a future event or a permission. US federal plain-language guidance recommends "must" for requirements, "must not" for prohibitions and "may" for discretion.
What is a hidden verb?
A verb turned into a noun and propped up by a weak verb: "make a decision" instead of "decide", "conduct an investigation" instead of "investigate". Hidden verbs make sentences longer and more abstract; the table suggests the verb to use instead.
Does GOV.UK really ban words like deliver and robust?
Its style guide lists them under "Words to avoid" because they are vague when used for abstract ideas, and gives exceptions - pizzas and post can be delivered, and a sturdy object can be robust. This page quotes those exceptions next to each finding.
Is the Flesch-Kincaid grade accurate?
It is a formula based on average sentence length and syllables per word, published in 1975 for US Navy training material. It is consistent and quick, but it cannot see meaning, layout or how familiar the words are. Syllables here are counted by a rule of thumb, so the grade is an estimate.
How is this different from a filler word checker?
This page looks at comprehension - hard words, hidden verbs, requirements, sentence length and reading grade, each tied to a published guideline. The Filler Word and Cliché Detector looks for padding such as "basically", "at the end of the day" and "very", sorted into categories.
Last reviewed by the A2Z.Tools team against the sources listed above.
