RankEngineRankEngine Help

Brand Voice Studio

Found in Content Studio → Settings → Brand Voice. Three paste boxes for sample copy your brand has shipped before. RankEngine analyzes them and produces a "fingerprint" the blog generator injects into every prompt.

What gets analyzed

  • Tonality — formal/conversational/clinical/playful, on a 5-point scale.
  • Sentence length — average + coefficient of variation.
  • Vocabulary hints — domain-specific terms you use repeatedly.
  • Signature phrases — 5-10 phrases the model will preserve verbatim where natural.
  • Avoid phrases — phrases never to use (cliché, off-brand, competitor jargon).

How it integrates

Once analyzed, the fingerprint becomes a BRAND_VOICE_FINGERPRINT block in the blog system prompt. It sits ADJACENT to the anti-AI-detection blocks (phases 63-67) — never replacing them. So you get on-brand voice + low AI-detection score in the same draft.

Setting it up

  1. Go to Content Studio → Settings → Brand Voice.
  2. Paste 3-5 samples (each capped at 8000 chars). Mix of: a product description, an existing blog post, a customer email if you have one.
  3. Click Analyze. Wallet drops ~$0.005-0.01.
  4. Review the fingerprint preview. Edit the bullet list of "Avoid phrases" if anything seems off.
  5. Save. Future blog generations use it automatically.

When to update

  • After a major rebrand or tone shift.
  • After you ship 5+ posts and notice generated drafts drifting from your style.
  • If you launch a new product line with distinct voice (e.g. "Kids Wellness" vs "Adult Performance" — separate fingerprints planned for a future phase).

Privacy

Samples are analyzed by RankEngine AI. RankEngine stores the fingerprint (a JSON summary) but not the raw samples after analysis completes.

Related