Noticeai.readability

The page is harder to read than it needs to be

Why it matters

content.thin counts words. It has nothing to say about a page with plenty of words that nobody can get through: sentences of fifty words, three paragraphs where one would do, and the particular register that creeps into a website when it is written by the people who built the thing rather than the people who buy it.

Readability formulas (Flesch, Gunning fog) measure sentence and word length and call it reading age, which punishes plain technical words and rewards short bad writing. A model reading the actual page does better, and can say which part is the problem.

An opinion, labelled as one, and only ever a notice — it costs nothing against your score. Plenty of pages are meant to be dense, and a model has no idea who your readers are.

How to fix it

Read the sentence the finding quotes out loud. If you run out of breath, split it.

Beyond that, the usual things: put the answer first and the background after it, cut the sentence that introduces the sentence, use the words your customers use rather than the ones your industry does, and let a subheading do the work of a paragraph.

Examples

These are the rule's own test pages, and what it says about each. They are run on every change to the rule book.

They differ only in what the server answered:

{"model":"Claude","readability":{"ok":false,"comment":"The second paragraph is one sentence of sixty-two words and should be three"}}

NoticeThe second paragraph is one sentence of sixty-two words and should be three (Claude's opinion, not a rule.)

fail-dense
{"model":"Claude","readability":{"ok":true,"comment":"Short sentences, and it answers the question first"}}

PassedThe page reads clearly: Claude agrees.

pass-clear