Overhead hands demo the product start-to-finish, cooking-show style. No faces, no actors, all clarity.
VIDEO (Seedance 2.0, 12s, vertical 9:16, overhead top-down, no faces) — a pair of hands demonstrates [PRODUCT — e.g. "a manual coffee grinder"] on a clean light-wood table, cooking-show style. HOOK 0.0–1.5s: hands slide the product into center frame and tap it once — decisive and fast, no empty-table dead time. 1.5–5.0s: [STEP 1 — e.g. "pour beans in and crank the handle twice"], framed tight enough that the mechanism fills half the frame. 5.0–8.5s: [STEP 2 — e.g. "twist open the base and pour the grounds into a filter"], with one quick punch-in cut on the payoff detail. 8.5–12s: the finished result is slid toward camera and released with an open-palm done gesture; hold the end frame 1.5s. Camera: locked overhead rig, 35mm feel, table filling frame edge to edge, cuts only on action beats. Light: bright, soft, even top light, no harsh shadows. Grade: clean bright appetizing warmth. Hands: natural skin texture, five fingers, confident tutorial pacing. No morphing hands, no floating objects, no in-gen text, label sharp and unwarped. VO (ElevenLabs, friendly and brisk — or captions instead): "Step one — [ACTION]. Step two — [ACTION]. That's it." AUDIO: real contact foley on every beat — taps, twists, pours — over a quiet kitchen room tone.