AI tells
The phrases, rhythms and structures that mark text as machine-written. A customer who spots one stops reading the argument and starts reading the author.
AI tells are the fingerprints a model leaves on text: words nobody uses in speech (leverage, robust, delve, seamless), every sentence the same length, the three-part balanced conclusion, the em dash doing structural work in every paragraph, the summary at the end that repeats the middle. None of them is wrong exactly. Together they mark a document as produced rather than written, and a customer who notices stops reading the argument and starts wondering who, if anyone, thought about it.
For a key account manager this is a commercial problem, not a stylistic one, because the proposition is the one document the customer reads closely. The Value Matters tools carry an explicit rule in their instructions: no corporate padding and no AI tells. The practical test is to read the draft aloud. Anything you would not say to the customer across a table, cut.
Set out in full in: Brooks, AI Glossary of Terms (Cranfield KAM Best Practice, 2026); Value Matters editorial style guide.
Related terms
Human in the loop
Adopt, but keep your hands on the wheel. The machine amplifies judgement. It does not supply it.
System prompt
The standing instructions a model runs on before any user speaks: its role, rules, tone and knowledge base. In the method's tools, the schema is the system prompt.
Prompt
The instruction you give an AI model. The single biggest factor in the quality of what comes back. Also, in the book, a structured question you ask yourself.
See the method run on one real account in three days: the One Page Proposition.