The hook, named and taken apart

What the breakdown contains

• The hook, verbatim and classified. The opening seconds quoted exactly, labelled as one of seven types — question, contrarian, benefit-promise, conflict, identity, result-first, or other — plus one sentence on why it holds.
• A structural timeline. Three to six chronological segments, each with a time range, the job that segment is doing, and a summary of what is said in it.
• Pain points the script targets, listed.
• Selling points or value claims it makes, listed.
• The call to action, quoted verbatim — or marked empty, which is itself informative.
• Pacing notes — how often a new beat lands and where the script drags.
• A skeleton: the same script as a four-to-six-step fill-in-the-blank template you can write your own version into.

Why naming the hook type is the useful part

Creators describe hooks with adjectives — punchy, relatable, strong. Adjectives do not transfer. Categories do. Once you can see that eleven of the top fifteen videos in a niche open with a contrarian claim and none of them open with a question, you know something concrete about what that audience responds to.

Run the breakdown across a batch and you get a distribution instead of an anecdote. That distribution is the actual finding: not that one video worked, but that one shape keeps working while another keeps failing.

The skeleton is the deliverable

Everything above is diagnosis. The skeleton is what you take into a document. It strips the specific product, niche and personality out of a proven script and leaves the load-bearing steps: what the first line has to accomplish, what has to be established before the pitch, where the objection gets handled, what the last line asks for.

Filled in with your own subject matter, that is an original script built on a structure with evidence behind it. If you would rather have a first draft written for you, the rewriter does that step directly.

Cost and limits

Breakdowns do not consume minutes. You pay for extraction — the audio becoming text — and each extracted video includes one breakdown and one rewrite at no additional cost. The result is stored with the item, so reopening it later does not re-run anything.

One breakdown per video is the limit. It is generated on demand rather than automatically, because on a batch of a hundred you usually only want the eight that outperformed.

Read the output as a draft, not a verdict

The breakdown comes from a language model reading a transcript. It is a strong reader of structure and a mediocre judge of causation: it can tell you what a script does, but it cannot know whether the video worked because of the hook or because the creator has two million followers. Treat the pattern across many videos as the signal, and any single breakdown as one opinion.

Questions people ask