Voice DNA

Voice DNA is the term for the small fingerprint of features that make someone's writing identifiable. It includes word choice (do they say "stuff" or "elements"?), sentence rhythm (short and choppy or flowing and qualified?), default tone (dry, enthusiastic, analytical, casual), and the patterns they reach for under pressure. Voice DNA is harder to replicate than topic knowledge because it lives at the level of phrasing, not ideas. A ghostwriter can match topic; matching voice requires either many examples of the writer's existing work to study or a recording of how they actually talk. Edgar uses voice as the source material for this reason: how someone speaks contains more voice signal than the polished text they have written.

Examples

• A founder who tends to undersell with "kind of" and "a bit" reads as understated, even on bold claims.
• A writer who uses semicolons and dependent clauses reads as careful and qualifying.

Frequently asked questions

Can voice DNA be learned by AI?

Partially. AI can capture lexical features and sentence-length patterns reasonably well from a corpus. Tonal features and patterns of thought are harder; voice samples (audio) capture more than written text alone.

How long does it take to develop a voice on LinkedIn?

Writers tend to find their voice after 30-50 published posts. Before that, the writing drifts across registers as the writer experiments.