playbook-creative-ideas

Define three fixed bullet slots in the seller's own words, hand-write three exemplar bullet sets for real companies, then run one AI call per row that fills each slot with a detail from the company evidence. Fixed slots measured 21/23 usable versus 2/5 for free-form ideation.

Require the model to output an evidence quote alongside each bullet, then assert that quote appears as a substring of the input evidence text. A bullet whose evidence is not in the input was invented and gets blanked, with no second model call needed.

Use gpt-4o-mini inside Clay at roughly $0.27 per 1,000 rows. A nano-class reasoning model with reasoning unset burns hidden reasoning tokens, costs about 19x more, and frequently returns blank content; outside Clay, a nano model at minimal reasoning costs $0.12 per 1,000 rows.

Generic bullets come from free-form ideation where the model picks its own subjects, or from gating the evidence waterfall on emptiness instead of length. Gate each evidence source on length under 200 characters so thin boilerplate does not pass through.

The row abstains with empty strings, and any empty bullet excludes the row from the ideas campaign entirely. It is routed to a separate non-ideas campaign, because a 3-bullet email with 2 bullets is a different email that a sequencer cannot pin to a variant.

Skip it when one personalized sentence atop a fixed email is enough, when the hook is an event, or when the operator cannot name three deliverable slots. An idea the operator did not ask for is one the seller cannot deliver on the call.