Open the box of something you just bought, and show how it looks and how it's used, in a vertical video of about 30 seconds.
In Shot 1, put a caption on the first frame, over the lid about to open, to raise the question 'what's inside?' In Shorts, viewers decide within the first few seconds whether to swipe away.
To recreate it: Write the hook prompt so the lid is only lifted and the inside isn't shown.
In Shot 1, look straight down at a black box on a white desk. The simple background leaves only the box and hands to look at, and it fits neatly in a vertical frame.
To recreate it: Include 'plain matte box, no text or logos' so the AI doesn't draw garbled text on the box.
In Shot 2, show hands peeling the protective film as if looking down through your own eyes. The viewer feels the hands-on pleasure of tearing it off themselves.
To recreate it: Fingers peeling film tend to smear, so generate 2-3 drafts at 480p and choose.
In Shot 3, slowly approach the corner and hinge of the case at very close range. When the texture is visible, even a small product looks valuable.
To recreate it: Write a dark gray background with 'thin rim light tracing its outline' so only the outline glows.
In Shot 4, place the product on a white turntable and turn it slowly. The camera stays put and only the product rotates, so every side shows in turn.
To recreate it: Lock the camera with 'locked off' and write that only the object rotates, so the background shakes less.
The earlier shots show only parts such as the box, film and corner, and the full product is revealed for the first time in the Shot 4 turntable. Leaving curiosity open and resolving it at the end keeps people watching to the end.
To recreate it: Use a prompt that shows the whole product in the turntable shot only, and keep the other shots deliberately partial.
Lay a riser that grows louder in the 1s before the reveal, and land a 'thud' hit on the first frame of the turntable shot. The sound pins down the moment of the reveal.
To recreate it: Don't leave sound effects to the AI. Place them exactly on the frame where the cut changes in editing.
Cinematic multi-shot sequence, 11 seconds total, vertical 9:16. Shot 1 (2s): static top-down flat lay of hands lifting the lid of a plain black box on a white desk; Shot 2 (2s): first-person POV of hands peeling protective film off a white earbuds case; Shot 3 (2s): macro slow push-in on the case's curved edge and hinge, thin rim light; Shot 4 (3s): the open case rotating on a white turntable, locked-off camera, product low in frame with empty space above for a caption; Shot 5 (2s): handheld selfie-angle close-up of a young woman putting an earbud in and smiling. Consistent unbranded product, bright high-key light, clean natural colors, no logos, no text, satisfying mood.
The box and hands seen from above build curiosity that "it's about to open." Hide the whole product until the end, then show it all at once in the rotating shot. Bright lighting and rim light make even a small product look sharp and valuable.
Vertical 9:16 top-down flat lay of a plain matte black box on a clean white desk, two hands slowly lifting the lid, static overhead camera, bright high-key lighting with soft shadows, crisp clean look, mood of anticipation, no text or logos.
Vertical 9:16 POV shot looking down at hands peeling a thin protective film off a sleek white earbuds case, static camera, bright soft daylight, glossy surface detail, shallow depth of field, satisfying tactile mood, plain unbranded product with no logos.
Vertical 9:16 macro shot of the curved edge and hinge of the earbuds case, slow push-in, a thin rim light tracing its outline against a dark gray background, extreme detail of the matte texture, shallow depth of field, premium clean mood, no logos.
Vertical 9:16 shot of the open earbuds case slowly rotating on a white turntable, camera locked off, seamless spin, bright high-key lighting, soft reflections on the glossy surface, product centered low with empty space above for a caption, clean commercial look, no logos.
Vertical 9:16 selfie-angle medium close-up of a young woman putting an earbud in and smiling as music starts, handheld phone at arm's length with subtle natural shake, soft window light from the side, bright natural colors, friendly review mood.
Put a hook caption on screen in the first 1–3 seconds, from the very first frame, just before the lid opens. Join the packaging and texture inserts with hard cuts of 1–2 seconds each. Show the whole product for the first time in the rotating shot, as an open loop → payoff: lay a riser in the 1 second before the reveal and land a hit on the first frame of the rotation. Keep captions inside the safe area, away from the bottom 20% and the button column on the right.
Generate each shot separately in 9:16 from the start. For a real product, make the start frame from a product photo to lock its shape. AI changes or garbles logos and text, so leave those spots blank and add the original image in editing. Fingers peeling off film are easily smeared, so make 2–3 480p drafts and pick one. For the rotating shot, lock the camera and rotate only the object, and the background wobbles less.