Uniform sentence length: a scaled-content hygiene signal

What it means

The body text on this page shows very low sentence-length variance — sentences run at a consistently similar word count throughout (standard deviation below 4 words). Uniform sentence rhythm is a stylistic characteristic of templated or unedited writing. It does not indicate how the content was produced; it indicates the content has not been through the kind of editorial review that naturally introduces structural variety. This check identifies a content quality pattern. It does not determine AI authorship.

Why it matters

Sentence-length variance is not a ranking signal on its own; it is a proxy for editorial effort, which is central to Google's E-E-A-T "How" question — how was this produced, and does it show care? Human writing varies sentence length as a byproduct of working through ideas: short sentences for emphasis, longer ones for qualification and nuance. Stylistic flatness, where every sentence lands at the same beat, is a consistent feature of unedited output from templated workflows, and it tends to co-occur with other low-effort signals (missing author, missing date, generic structure) that compound into a site-level quality pattern. This is distinct from the low-variety check, which measures vocabulary diversity rather than sentence-length variance.

How to fix it

1. Read the page aloud; rewrite the sections where the rhythm feels monotonous.
2. Break some compound sentences into short ones for emphasis, and merge short parallel sentences into one where the ideas belong together.
3. Add specifics — concrete examples, qualifications, counterpoints — which naturally introduce longer, more complex sentences.
4. Cut filler; tighter writing is almost always more varied writing. Re-run the check after revising.

When it's not a problem

The check excludes pages under 300 words and pages where more than 40% of the text sits inside <ol>, <ul>, or <pre> elements — instructional guides, API references and reference tables legitimately use short, parallel sentences because the format demands it. The threshold is deliberately conservative; instructional prose commonly scores in the 4–6 range and is fine. This fires only at the extreme low end, where flatness is hard to attribute to a deliberate stylistic choice, and it is never a verdict of AI authorship.

How common is it?

119 audited sites in our corpus currently show this issue. The breakdowns below show which platforms, gatekeepers, verticals and countries are most exposed.