'3 mistakes' in 40 seconds: the educational format that converts
Three mistakes the viewer is making right now, each with the fix shown in a shot, ranked so the last one is the one they did not know about. Prompt returns the script, text plan and shot list.
Write a '3 mistakes' script about {topic} for {audience}. The format works because every viewer checks themselves against the list; educational videos convert because the fix is the product or the method. Rules for the mistakes: - Mistake 1 is the one everyone knows they make (recognition). - Mistake 2 is the one they suspect (confirmation). - Mistake 3 is the one they have never heard (the reason to save the video). - Each mistake is a thing the viewer can check in under a minute. No vague ones ('not being consistent'). Structure (40 s, 9:16): - 0-3 s HOOK: 'Three {topic} mistakes, and number three is the one nobody tells you.' or a sharper variant. Text on screen: '3 mistakes' with the topic word. - 3-14 s MISTAKE 1: name it (3 s), show it (4 s), show the fix (4 s). - 14-25 s MISTAKE 2: same shape. - 25-37 s MISTAKE 3: same shape, slightly longer on the fix. - 37-40 s CLOSE: 'Save this for the next time you {action}.' A save ask, not a download ask. The download ask goes in one video out of several, not here. Word budget: 100-110 words at 165 wpm. On-screen text: a numbered title card per mistake (1/3, 2/3, 3/3) with the mistake in 4 words max, and the fix in 4 words max when it appears. Return: a table (time, section, spoken line, on-screen text, what is in the shot), then the three mistakes as a one-line list for the caption. Topic: {topic} Audience: {audience} What the fix involves: {the method or the product feature, if any}
In Claude Code, paste the prompt as a message.
Example
Topic: product photos for an online shop. Audience: someone selling handmade things from home.
| Time | Section | Spoken line | On-screen text | Shot |
|---|---|---|---|---|
| 0-3 s | Hook | Three product-photo mistakes, and the third one is why your listing looks cheap. | 3 PHOTO MISTAKES | Face, phone in hand |
| 3-6 s | M1 name | One. Shooting under the kitchen light. | 1/3 OVERHEAD LIGHT | Ceiling lamp |
| 6-10 s | M1 show | Yellow cast, shadow under everything. | Mug under the lamp | |
| 10-14 s | M1 fix | Window, side on, lamp off. | FIX: WINDOW, SIDE ON | Same mug by the window |
| 14-17 s | M2 name | Two. Zooming in. | 2/3 DIGITAL ZOOM | Fingers pinching |
| 17-21 s | M2 show | The phone throws away pixels; the edge goes soft. | Blurry edge close-up | |
| 21-25 s | M2 fix | Step closer, crop later. | FIX: STEP CLOSER | Hand moving the phone |
| 25-28 s | M3 name | Three. The background. | 3/3 THE BACKGROUND | Cluttered counter |
| 28-32 s | M3 show | Your eye reads the mess before it reads the product. | Same | |
| 32-37 s | M3 fix | Cut it out and put it on white; the app does that in one tap, on the phone. | FIX: CUT OUT, WHITE | Thumb taps, background drops |
| 37-40 s | Close | Save this for your next listing. | SAVE FOR NEXT TIME | Face |
Caption list: overhead light, digital zoom, cluttered background.
Tips
- The fix for mistake 3 is where the product lives, if there is one. One mention, shown not pitched.
- 'Save this' outperforms 'follow' on an educational video because it is the thing the viewer actually wants to do with it.
| works with | Claude Code: Runs as a skill or a prompt in the terminal; can read your files and write the result. · Claude: Paste into claude.ai or the Claude app. · ChatGPT: Paste into ChatGPT; works with any GPT model. · Any LLM: Model-agnostic: Gemini, Mistral, a local model. |
| cost | Free. No API, no key. |
| status | Unverified: written from the technique, not yet run end to end by us. |
| license | CC BY 4.0 for the text here; skills are MIT. |
| updated | 2026-11-01 |
| for agents | three-mistakes-format.md |
Related
Tutorial in 30 seconds: result first, then the steps at reading speed
The compressed how-to that gets saved: the finished thing in frame 1, then 3-5 numbered steps each shown and captioned. The prompt budgets the words so the steps fit at 165 wpm.
Countdown listicle: 5 items in 45 seconds, best one last, numbers on screen
The ranked list format, with the rule that makes it hold: each item gets shorter as the count drops and number one gets the longest shot. The prompt returns script, text plan and the caption list.
Myth vs fact: three beliefs checked in 35 seconds, each with a demonstration
The format where a common belief is stated, then tested on camera, then labelled. Works because the viewer picks a side in the first second; the prompt insists on a visible test per myth.
'Watch me use it': a 30-second product demo with no pitch and one real task
The demo that converts because it is a task, not a feature tour: one thing the viewer wants done, done on camera, in real time where possible. The prompt forbids adjectives and plans the text.
Before/after: the 20-second reveal with the 'after' shown first
The before/after format with the twist that keeps it from being skipped: open on the finished result, then show how bad it was, then the change. The prompt returns a timed script and text plan.
Founder explainer: 'I built this because' in 40 seconds, to camera, one problem
The founder-to-camera format that converts because it names the specific moment the problem became unbearable, then shows the product once. The prompt bans origin-story clichés and budgets the words.