Video Hook Bank

The hook is where most retention is lost. YouTube's average audience retention at 30 seconds sits around 60-70%, meaning up to four in ten viewers leave before the half-minute mark. The hook is what stops them. Most creators waste this window on greetings, logos, and "welcome back to my channel" — a retention killer. This skill fills it with precision. Twenty hooks per request, written word-for-word, ready to film.

The 20 are spread across 10 mechanisms — Result First, Contrarian Claim, Specific Number, Personal Failure, Time Pressure, Question, Demonstration, Story Entry Point, Us vs Them, and Cold Open — with two hooks each. The mechanisms map to video formats: tutorials lean on Result First and Demonstration; essays on Contrarian Claim and Question; vlogs on Story and Personal Failure; reviews on Specific Number. The skill picks the weighting that fits the format you're shooting, so you don't get hooks that work for a different kind of video.

Every hook gets scored on three dimensions out of five — Thumb-Stop Power (does it stop a scrolling viewer in the first second), Promise Clarity (does the viewer know what they'll get by staying), and Authenticity Fit (does it sound natural for the creator's actual voice). The top five by total score are flagged "film these first." Each top hook also gets the 10-second bridge written for it — because a hook disconnected from the content feels baited, and the transition is part of the hook's job.

The rules are tight. No lies, ever — clickbait that doesn't pay off tanks long-term retention and destroys trust. Emotion before information in the first three seconds. Word-for-word writing, not vague suggestions ("something like, you know, I tried this thing" is not a hook). And the skill flags which mechanisms to avoid for sensitive topics — health journeys, loss, trauma — where Contrarian Claim and Us vs Them read exploitative but Story Entry Point and Personal Failure can land authentically. You also get a suggested A/B pairing using two different mechanisms, so the test result tells you what your audience actually responds to instead of just which sentence sounded better.