TextIntermediate

Diagnose a video that loses people at twenty seconds

Max Submitted by Max Added today
You are an editor who reads retention graphs for a living. My video underperforms and I know where, not why. Video: {{length, format, platform}}. Retention data: {{drop-off timestamps with % remaining at each}}. Timestamped transcript: {{paste}}. What is on screen in the first 30 seconds: {{describe shot by shot}}.

Produce:
1. THE DROP MAP — table: Timestamp | % remaining | What is happening in audio | What is happening on screen | Most likely reason people leave.
2. THE ONE REAL CAUSE — the drop costing the most views, and why you picked it over the others.
3. RECUT — the exact edit for that section: which lines to delete (quote them), what to move earlier, where to add a cut or a visual change. Give timecodes.
4. NEW OPENING — rewrite the first 15 seconds using only sentences that already exist in my transcript, marking each line's timestamp.
5. LEAVE ALONE — the drops that are normal for this format and not worth fixing, with the reason.
6. THE TEST — what the retention curve should look like if the recut worked.

Rules: no advice that needs footage I did not describe. Do not tell me to "add more energy". If the data is too thin to separate two causes, say so.

How to use it

Bring the retention export, not your impression of where it sags — without timestamps and percentages step 2 is a coin flip. It cannot see your footage, so its read of the visual column is only as good as your shot description, and a drop caused by a bad thumbnail-to-content mismatch will not show up here at all.

Compatible popular AI tools

These tools are mapped to this prompt based on their capabilities.