Skip to main content
SerpGem
Editing & quality · 6 tools

Editing & quality tools

Revision is where content quality lives. These 6 tools surface the details humans miss: side-by-side diffs, word-level inline diffs, passive-voice flags, transition-word density, outline extraction, and sentence rewriting suggestions. Pair them after your first draft and before your CMS paste.

About these tools

Editing & quality questions

Why does passive voice hurt SEO rankings?
Passive voice doesn't directly hurt rankings (Google's guidelines don't mention it), but it correlates with lower engagement. The Nielsen Norman Group's 2013 F-pattern study found passive sentences are 30% less likely to be scanned, reducing dwell time and increasing bounce rate — both indirect ranking signals. Hemingway flags passive because it buries the actor. Our Passive Voice Checker uses the same detection rules.
What's the ideal transition-word ratio?
Yoast recommends 30%+ of sentences should contain a transition word (however, therefore, first, consequently). Research from the University of Amsterdam (Kamalski 2008) found that 25-35% transition density improved comprehension scores by 12-18%. Below 20% feels choppy; above 40% feels overly formal. Our Transition Words Checker reports your exact ratio and flags low-transition paragraphs.
How does word-level inline diff differ from line-level diff?
Line-level diff (git diff) marks entire lines changed even if only one word changed — noisy for prose. Word-level diff tokenizes at word boundaries and highlights only the changed words in-place, keeping unchanged context visible. For editing content, word-level is strictly better: you see exactly what you edited, and reviewers can scan for intent without toggling between files.
Can an outline extractor find the structure in un-headered content?
Our Outline Extractor can work two ways. If the document has H1-H6 tags or Markdown #s, it extracts the explicit outline. If not, it uses sentence-vector similarity (a variant of TextTiling) to identify topic shifts and proposes a synthetic outline. Accuracy is ~70% on the synthetic mode — useful as a starting point when reverse-engineering a competitor's structure.

More in Content & Writing

Related sub-groups