Courses fail without extraction

The "sell a course" material that floods social media covers funnels, pricing, launches — everything except the one step that determines whether the course works. There is no knowledge extraction anywhere in it. Most experts think they know what they do, and they teach that, in vague terms, believing it's simple and transmissible. It isn't — experts teach the description, not the skill.

My inference from watching this market: that's why most courses coaches and consultants sell don't transfer skill. And it's likely why courses are getting harder to sell — students who paid for earlier ones are jaded, and the wariness spreads. The market is running on the failure of its own products.

There's a market-level mechanic underneath. Course quality is unobservable before purchase, and because sellers genuinely believe their description-based courses work, bad courses flood in alongside any good ones — nobody self-selects out. Burned buyers can't tell the difference at the point of sale, so they discount everyone's claims. That's a lemon market, and in a lemon market the winning move isn't louder marketing or lower prices; it's verifiable demonstration — students who can be observed applying the skill.

What a working course would require is the full pipeline: structured extraction of the expert's tacit knowledge, encoding into perceptual language a learner can use, and a way to teach learners to recognise their own correct execution — the double translation problem. None of that appears in current course-creation advice.

Working claim, held with appropriate looseness. What would test it: completion and skill-application data for expert-led online courses, and whether any course-creation curriculum anywhere mentions tacit knowledge.