A practical writing pattern for AI explainers that non-technical leaders actually finish reading.
Introduce
Most AI explainers fail because they start with jargon. The IDEA framework fixes that by forcing a predictable reader journey: Introduce, Describe, Evaluate, Advise. Each section answers one question the reader is silently asking, so the post feels inevitable instead of dense.
Describe
In the Describe section, explain the mechanics without marketing words. Name the parts, show the flow, and reference sources inline [1][2]. A reader who cannot summarize what the thing is after this section will not trust anything that follows.
Evaluate
Evaluate is where the post earns credibility. Call out the tradeoffs: what IDEA gets right, where it feels forced, when a plain narrative would serve better. Readers reward honest evaluation with attention.
Advise
End with one concrete next step. Not a list of ten. One. "Pick your next post. Draft it in four H2 sections labeled Introduce, Describe, Evaluate, Advise. Publish Friday." Advice that fits on a sticky note gets acted on.