Real-world conversation mining method for generating contrarian video ideas

Opens by addressing a specific pain state (no video ideas, tired of AI tools) to earn attention. Middle introduces a two-step ideation ritual: (1) identify a keyword in overheard conversation that triggers personal curiosity , the lean-in signal; (2) locate the point of mild disagreement within that conversation , the contrarian stance. Close frames the output as 'naturally contrarian' rather than performatively contrarian, distinguishing authentic friction from attention-bait. The whole structure is a repeatable decision tree, not a one-time tip.

No explicit CTA , relies on implicit follow / save signal.

- Open by naming the exact frustration your audience has RIGHT NOW ('tired of using AI for ideas') before offering the solution , this filters in the right viewer in the first 3 words - Frame your method as a named test or ritual ('This is called the test') , a label makes the technique memorable and shareable - Use a hyper-specific third-party example (chiropractor + hyperbaric chamber certification) to make an abstract process feel concrete and real - Build your ideation prompt as a two-question sequence: first 'what makes you lean in?' then 'what makes you go huh?' , two-step structures are easier to remember and apply than single-step advice - Distinguish your contrarian angle as belief-based, not attention-based , explicitly say 'I actually believe this' to pre-empt the 'you're just doing it for views' objection - Anchor the content idea to social discomfort ('slightly uncomfortable to say out loud') , this is a reliable proxy for content that will generate comments and saves - Position the method as offline and embodied (parties, real conversations) to contrast with every other creator telling people to use tools , the contrast itself is the hook