Most small shops already own the raw material. A phone photo of the product on a clean surface is enough to produce a moving clip, and the gap between the still and the clip is motion rather than budget. That reframing matters, because it turns an expensive-sounding task into an editing habit.
A photo that animates well usually has three qualities: an uncluttered background, breathing room around the subject, and light that falls from one direction. Tight crops fight you, because any camera movement immediately pushes the product against the edge of the frame. Rejected e-commerce shots, the ones with a distracting cable in the corner, are often good animation sources once you reframe them.
The words that survive translation into footage are the dull ones: push in, pull back, drift left, tilt up, slow orbit. A sentence like "the bottle rotates and the label catches the light" gives the generator a subject, a motion and a lighting event. A sentence like "the bottle explodes into sparkling particles and reassembles" gives it a demand it will probably meet badly, and you will spend three attempts discovering that.
Keep the product readable. If the label faces away from the lens for half the clip, the shot has failed commercially even if it looks handsome. State the orientation: "front label facing camera throughout".
Write one prompt template per product category and swap only the subject words. Skincare: a frosted glass bottle on a pale stone ledge, front label facing camera, slow push-in, soft overhead light. Footwear: a canvas sneaker on a concrete step, slow drift right, low side light. Templates keep a catalogue visually coherent, and they make a batch of clips feel like one campaign instead of twelve unrelated experiments.
Some product stories are really reveals. Lid on, lid off. Box closed, box open, contents visible. Supplying a first frame and a last frame lets you ask for the transition between two states you have already photographed, which is far more controllable than describing the change in prose.
Watch the clip at the size it will actually appear. A subtle drift that reads beautifully full-screen can vanish in a feed thumbnail, and a slow reveal can lose its punch in a three-second autoplay. Test on a phone, not a monitor.
If you want to try the workflow without a subscription conversation, this image to video AI generator runs in the browser and needs no account, which makes it cheap to test a template against ten products before committing to a look.