Measured client writing guided a rejected campaign's rebuild
I compared 241 reference pieces and rebuilt the copy against the studio's own voice profile.
My first round of AI-assisted copy sounded like AI, so I measured the client's own copy and wrote a rebuild that landed inside its range.
Context
The client couldn't say what was wrong with it.
Problem
The copy was statistically unlike the client's real voice, and there was no way to argue about it productively. Voice disputes die as taste debates: one person feels the slop, another defends the draft, and nobody can point at anything.
Approach
The measurements resolve into a numeric profile of how this one business sounds, and the copy got rebuilt against it.
Stack
- Python metric engine, nine metric families spanning lexical, rhythm, punctuation, and sentiment axes
- Corpus extraction across hundreds of client-authored assets
- A rebuild written to the measured numbers
Result
On a first measurement, the rebuild landed inside the client's range. Nobody had to trust a writer's ear, because the distance between the first round and the client's voice was a number anyone could check.
Impact
Writing quality at this studio became an inspectable number rather than a taste argument. When a future draft drifts from the voice, the drift shows up as a measurable distance before a customer ever sees it.
Lessons
AI-written content is converging on one voice, a new dialect of the internet. Conversion copy that works in the coming decade gets written in a way no other agent writes and most humans don't think. I check that with a measurement instead of my own ear.
Who this is for
For any business whose brand voice is worth protecting, and who'd rather catch drift with a measurement than argue about taste.
Get a quoteSee it live
The method behind this result
- Content StrategyContent strategy can start with the answers editors already have. A concept map shows which pages need new links and which reader questions still need an article.
- Standard Model of Effective MediaThe Standard Model of Effective Media connects familiar details to difficult subjects. Creators test the connection against what audiences understand and do.