Win the first second, then keep delivering.
Start with a truthful visual hook, keep every scene useful, control pacing for mobile viewing and deliver the promised payoff before the final CTA.
Act as an expert short-form video strategist, vertical-video director, editor and AI video prompt specialist. TASK: Create a high-retention short-form video prompt optimized for: Facebook Reels Instagram Reels TikTok YouTube Shorts The video should be fast enough to maintain attention, but still clear, credible and visually coherent. PLATFORM: [Facebook Reel / Instagram Reel / TikTok / YouTube Short.] TOPIC: [What is the video about?] GOAL: [Educate / entertain / demonstrate / promote / explain / inspire / convert / other.] TARGET AUDIENCE: [Who is this for?] TOTAL DURATION: [For example: 10 sec / 15 sec / 20 sec / 30 sec / 45 sec / 60 sec.] ASPECT RATIO: [Default: 9:16.] MAIN HOOK: [What should make the viewer stop scrolling?] CORE MESSAGE: [What should the viewer understand or remember?] PAYOFF: [What does the viewer gain by watching until the end?] CTA: [Follow / save / comment / share / learn more / shop / none.] SUBJECT: [Person / food / product / process / screen / environment / other.] VISUAL STYLE: [Photorealistic / premium / creator-style / cinematic / educational / commercial / casual / other.] VOICE-OVER: [Paste script or describe tone and content.] ON-SCREEN TEXT: [Exact wording or none.] MUSIC: [Optional.] SOUND EFFECTS: [Optional.] BRAND: [Optional.] REFERENCE REQUIREMENTS: [Character / product / food / style details that must remain consistent.] MANDATORY MOMENTS: [List anything that must appear.] ELEMENTS TO AVOID: [Optional.] SPECIAL REQUIREMENTS: [Any additional instructions.] SHORT-FORM VIDEO REQUIREMENTS: 1. OPTIMIZE FOR VERTICAL VIEWING Default to 9:16 unless another format is specified. 2. DESIGN FOR MOBILE FIRST Important subjects, text and actions must remain easy to understand on a phone screen. 3. KEEP CRITICAL CONTENT INSIDE SAFE AREAS Avoid placing essential content too close to: Top UI Bottom UI Left edge Right edge 4. ESTABLISH THE SUBJECT IMMEDIATELY The viewer should understand what the video is about within the opening moments. 5. CREATE A STRONG FIRST-SECOND HOOK The opening should create: Curiosity Recognition A useful promise A visual surprise A compelling result A clear problem 6. DO NOT USE MISLEADING HOOKS The opening must match the actual video. 7. SHOW VALUE EARLY Do not spend several seconds on logos, introductions or unnecessary setup. 8. AVOID SLOW INTRODUCTIONS Do not begin with: Long establishing shots Lengthy greetings Logo animations Generic cinematic filler unless specifically requested. 9. KEEP THE OPENING VISUALLY CLEAR Avoid clutter that makes the hook hard to understand. 10. USE ONE PRIMARY HOOK Do not stack multiple unrelated hooks. 11. BUILD A RETENTION ARC A useful structure is: Hook Progression Value Payoff CTA 12. EACH MOMENT SHOULD EARN ITS SCREEN TIME Remove shots that do not add: Information Emotion Proof Progress Entertainment Clarity 13. KEEP SCENES SHORT Short-form content usually benefits from concise visual beats. 14. DO NOT CUT SO FAST THAT INFORMATION BECOMES UNREADABLE 15. MATCH CUT SPEED TO CONTENT Educational content may require more viewing time. Entertainment may support faster pacing. 16. VARY VISUAL SCALE Where useful, alternate between: Wide Medium Close-up Detail 17. DO NOT CHANGE SHOT SIZE RANDOMLY Each framing change should reveal something new. 18. CREATE PATTERN CHANGES Use purposeful shifts in: Shot size Camera angle Subject action Visual detail Text Sound to renew attention. 19. DO NOT RELY ON RANDOM EFFECTS FOR RETENTION 20. USE MOTION EARLY When appropriate, start with visible action rather than a static setup. 21. SUBJECT MOTION SHOULD BE CLEAR Avoid subtle actions that are difficult to notice on mobile. 22. PRIORITIZE LARGE READABLE DETAILS Tiny objects may require close-up shots. 23. HUMAN PRESENTER MODE Use: Direct eye contact Natural expression Clear gestures Readable body language 24. DO NOT OVER-ACT EXPRESSIONS Keep reactions believable. 25. PRESERVE CHARACTER IDENTITY Maintain: Face Hair Age Skin tone Clothing Accessories Body proportions 26. DO NOT ALLOW FACE DRIFT BETWEEN SHOTS 27. FOOD REEL MODE Prioritize: Finished result Texture Cut Pour Steam Mixing Serving Close-up detail 28. FOOD HOOKS SHOULD BE APPETITE-DRIVEN Show the finished dish or most visually satisfying moment early when appropriate. 29. PRESERVE FOOD CONTINUITY Do not change: Dish shape Layers Colour Toppings Serving vessel without reason. 30. PRODUCT REEL MODE Show the product clearly in the opening or very early. 31. PRESERVE PRODUCT IDENTITY Maintain: Shape Colour Logo Packaging Features Materials 32. DO NOT INVENT PRODUCT FEATURES 33. PROCESS / TUTORIAL MODE Prioritize meaningful steps. 34. DO NOT SHOW EVERY MICRO-STEP Only show what helps understanding. 35. USE STEP COMPRESSION Combine repetitive actions when possible. 36. BEFORE / AFTER MODE Only use when both states are truthful and supported. 37. DO NOT FABRICATE RESULTS 38. EDUCATIONAL MODE Use: Question Explanation Example Takeaway 39. TIP VIDEO MODE Lead with the problem or practical result. 40. LIST MODE Keep item count manageable for the duration. 41. STORY MODE Use: Immediate situation Progress Change Resolution 42. COMMERCIAL MODE Show: Problem or desire Product Use Benefit CTA 43. KEEP THE CTA SHORT Do not turn the ending into a long sales pitch. 44. MATCH CTA TO THE GOAL Examples: Save this Follow for more Try it Learn more Shop now Comment below 45. DO NOT USE FALSE URGENCY 46. DO NOT USE FALSE SCARCITY 47. DO NOT CLAIM GUARANTEED RESULTS 48. DO NOT INVENT SOCIAL PROOF No fake: Views Likes Reviews Ratings Testimonials Follower counts 49. DEFINE SCENE DURATION Every scene should have a realistic time allocation. 50. DO NOT OVERLOAD A 2-SECOND SHOT Avoid placing: Long narration Multiple actions Large text Complex camera movement inside a very short moment. 51. TIME VOICE-OVER REALISTICALLY Spoken content must fit naturally within the duration. 52. SIMPLIFY WORDING FOR SHORT-FORM Prefer conversational language and concise sentences. 53. ALIGN VOICE-OVER WITH VISUAL ACTION 54. DO NOT SHOW SOMETHING UNRELATED TO THE NARRATION 55. USE VISUAL PROOF WHERE POSSIBLE If explaining an action, show it. 56. ALLOW SILENT MOMENTS Not every second needs narration. 57. USE SOUND EFFECTS TO EMPHASIZE REAL ACTION Examples: Cut Pour Click Crunch Sizzle Door close Packaging open 58. DO NOT OVERLOAD SOUND EFFECTS 59. MUSIC SHOULD SUPPORT PACING Not compete with voice-over. 60. ON-SCREEN TEXT SHOULD BE SHORT Keep text readable at mobile size. 61. USE HIGH TEXT CONTRAST 62. LIMIT TEXT PER SCREEN 63. KEEP IMPORTANT TEXT LONG ENOUGH TO READ 64. DO NOT GENERATE RANDOM TEXT 65. USE EXACT PROVIDED WORDING 66. RECOMMEND MANUAL TEXT OVERLAY WHEN ACCURACY IS CRITICAL 67. CAPTION SAFE-AREA MODE Leave space where platform captions may appear. 68. PRESERVE SUBJECT VISIBILITY BEHIND TEXT 69. KEEP CTA AWAY FROM PLATFORM UI 70. DEFINE FIRST FRAME The opening image should make sense even before motion begins. 71. DEFINE FINAL FRAME The ending should provide: Payoff Result Hero image CTA Resolution 72. LOOP MODE When useful, design the ending so it can transition naturally back into the beginning. 73. DO NOT FORCE A LOOP WHEN IT DAMAGES THE STORY 74. CAMERA MOVEMENT SHOULD BE SIMPLE Use one dominant move per short scene. 75. USE STATIC CAMERA WHEN SUBJECT ACTION IS STRONG ENOUGH 76. USE PUSH-IN FOR EMPHASIS 77. USE TRACKING FOR ACTION 78. USE CLOSE-UPS FOR DETAIL 79. USE OVERHEAD FOR PROCESS CLARITY 80. AVOID EXCESSIVE ORBITS 81. DO NOT ADD RANDOM WHIP PANS 82. TRANSITIONS SHOULD SUPPORT PACING Default to clean cuts. 83. USE MATCH CUTS WHEN THEY IMPROVE FLOW 84. USE SPEED RAMPS ONLY WHEN APPROPRIATE 85. DO NOT USE TRANSITIONS AS FILLER 86. PRESERVE SCREEN DIRECTION 87. PRESERVE OBJECT STATE If something changes during the video, later shots should reflect it. 88. PRESERVE PROP CONTINUITY 89. PRESERVE ENVIRONMENT CONTINUITY When scenes occur in the same location. 90. KEEP LIGHTING STYLE COHERENT 91. CREATE A VISUAL PRIORITY FOR EVERY SHOT The viewer should know where to look. 92. AVOID CLUTTERED BACKGROUNDS 93. KEEP MOBILE-SIZE SUBJECT SEPARATION STRONG 94. CREATE TARGETED NEGATIVE CONSTRAINTS Examples: No random text No face drift No duplicate hands No product redesign No object teleportation No warped food No background morphing 95. CREATE A RETENTION CHECK Ask whether each scene gives the viewer a reason to continue. 96. CREATE A MOBILE READABILITY CHECK Review the composition at small-screen scale. 97. CREATE A CLAIM CHECK Verify any: Numbers Benefits Results Offers Prices before publishing. 98. CREATE A PLATFORM CHECK Make sure the video structure fits the selected short-form platform. 99. CREATE AN AUDIO TIMING CHECK Ensure voice-over, music and sounds fit the final duration. 100. FINAL QUALITY CHECK Before finalizing verify: - platform is clear - aspect ratio is correct - first frame is strong - hook appears immediately - hook accurately represents the video - subject is mobile-readable - safe zones are respected - pacing matches the content - scenes are not overloaded - every scene adds value - visual variety has a purpose - character identity is consistent - product identity is consistent - food identity is consistent - object states remain logical - voice-over fits the runtime - narration matches the visuals - text is concise - text remains readable - no random text is generated - CTA is brief - CTA matches the video goal - no fake urgency is used - no fake scarcity is used - no unsupported result is shown - final payoff is clear - the video feels complete even at short duration OUTPUT FORMAT: 1. SHORT-FORM STRATEGY Summarize: Platform Audience Hook Core message Payoff CTA 2. RETENTION STRUCTURE Provide: Opening hook Development Pattern changes Payoff Ending 3. TIMELINE For each scene provide: Scene number Timestamp Duration Purpose Subject Action Shot size Camera movement Voice-over Visible text Sound Transition 4. FIRST-FRAME DESIGN Describe the exact opening composition. 5. FULL SHORT-FORM VIDEO PROMPT Provide a polished model-ready prompt. 6. VOICE-OVER TIMING Map each spoken line to the appropriate visual moment. 7. TEXT & SAFE-ZONE PLAN Specify: Exact text Placement Reading time Safe-area considerations 8. AUDIO PLAN Specify: Music Ambient sound Sound effects 9. NEGATIVE / PROTECTION INSTRUCTIONS List the most relevant short-form generation failures to prevent. 10. FINAL FRAME / CTA Describe the final visual and CTA treatment. 11. MOBILE RETENTION CHECK Confirm: Hook clarity Small-screen readability Pacing Payoff CTA IMPORTANT: - Optimize for mobile-first vertical viewing. - Deliver value immediately. - Keep the first-second hook truthful. - Remove unnecessary filler. - Match action, narration and text to the available duration. - Use purposeful visual changes rather than random effects. - Protect character, food and product continuity. - Keep CTA short. - Do not fabricate results, social proof, urgency or scarcity.
Use the prompt effectively.
Start with the scroll-stopping moment
Choose the strongest truthful visual, question, result or action for the opening instead of wasting the first seconds on introductions.
Build a retention timeline
Break the total runtime into short visual beats where each scene adds new information, progress, proof or payoff.
Design for a phone screen
Keep the subject large, protect caption and interface safe areas, and limit on-screen wording so everything remains readable vertically.
End with payoff before CTA
Give viewers the result or useful takeaway they were promised before asking them to follow, save, comment, learn more or take another action.
Turn a brownie recipe into a concise vertical Reel.
Platform: Facebook Reel.
Duration: 20 seconds.
Aspect ratio: 9:16.
Topic: Homemade chocolate brownies.
Audience: Home bakers.
Goal: Show the result and summarize the preparation.
Hook: Break open a brownie to reveal the moist fudgy centre.
CTA: Save this recipe.
0–2 SEC — HOOK: Extreme close-up of one brownie being gently broken apart to reveal the moist fudgy centre. Keep the brownie large in frame and immediately readable on mobile.
2–6 SEC — MIXTURE: Show melted chocolate mixture followed by a clean close-up of the batter being folded. Use short purposeful cuts.
6–10 SEC — PAN: Show batter spreading into the prepared baking pan, then a close detail of chocolate pieces on top.
10–14 SEC — RESULT TRANSITION: Move directly from the baked tray to neat brownie squares without unnecessary oven filler.
14–18 SEC — PAYOFF: Return to the strongest fudgy interior close-up, using a gentle push-in.
18–20 SEC — CTA: Finished brownie hero shot with the short text 'Save this recipe'. Keep the text inside the central mobile safe area.
Retention comes from useful progression, not random effects.
The hook should reveal, not deceive
Strong retention does not require clickbait. Show the most interesting real moment early and then deliver the information or result it promises.
Shorter is not always faster
A close-up texture reveal may need two or three seconds to register. Fast cutting only helps when viewers still have enough time to understand each shot.
Check the video at phone size
A composition that looks clear on a desktop can become unreadable in a vertical feed. Review subject size, text, captions and CTA placement at realistic mobile scale.