How to turn a product photo into a video ad with AI

Step by step

1. Step 1

   Start with a clear photo you have permission to use

   Use your own product image or one you are licensed to modify. Choose a sharp, well-lit photo with the whole product visible and no busy background. A front three-quarter view usually shows shape better than a flat side view. Do not use another seller's listing photos without permission. Small labels and fine print may change in the video, so do not rely on them to communicate a claim.

2. Step 2

   Choose one action and one buying reason

   Decide what the viewer should understand after 10 seconds: warm hands on a cold commute, frothy coffee at home, or a shirt that can be worn two ways. Build the clip around one visible action that supports that idea. A simple sequence — hold the product, use it once, show the result — is easier for the generator to keep consistent than a multi-scene story with several people and props.

3. Step 3

   Upload the image as a reference, not as a text attachment

   In VIVID, choose reference-to-video and upload one product photo. Up to six reference images are supported, but start with one: several angles can help show a product, while unrelated lifestyle images can confuse which object should appear. Keep the reference image and the requested product identical in color, model and accessories.

4. Step 4

   Describe what must stay fixed and what should move

   Name the product, its important visual features and the action. A useful template is: "Use the product in the reference image. Preserve its shape, color, materials and visible details. Vertical 9:16 handheld smartphone UGC video. A person holds it at chest height and says [short hook]. Cut to one close-up of [one realistic action]. Natural room sound, soft background music." Do not ask for new packaging text, extra features or a testimonial you cannot substantiate.

5. Step 5

   Make a short draft before spending on variations

   Start with a 5–10 second, 9:16 clip for Reels, Stories or TikTok. Use a lower-resolution draft to check the idea, then render the keeper at 720p. In VIVID, the prompt enhancer can expand a rough idea into a shot description, but keep the product-preservation instruction in the final prompt. One generation lasts up to 15 seconds; a longer ad needs separate clips joined in an editor.

6. Step 6

   Inspect the product before using the ad

   Watch the entire clip with sound, then pause on the first frame, the close-up and the last frame. Compare the product with the reference photo: silhouette, color, number of parts, controls, packaging and logo. Check hands and physical interactions too. Reject clips that change the product, make an unsupported claim or show it doing something it cannot do. A visually attractive video is not usable if it misrepresents what you sell.

7. Step 7

   Test one hook at a time and disclose AI

   Once the product stays consistent, make three to five versions with the same middle and ending but different opening lines or first shots. Change only one variable so the result tells you something. Send each ad to the matching product page. If the video looks realistic and is AI-generated, use the advertising platform's AI-content disclosure and follow its current ad policies.

Examples made this way

These public examples are useful shot and framing references. They were generated from text prompts rather than reference photos, so they illustrate the style of the clips, not exact product matching.

Tips for testing

• One product, one person and one action usually produce a more consistent clip than a crowded scene.
• Keep the product large enough to see in the reference image and in the requested shot; tiny details are the first thing to drift.
• Do not spend the whole budget on the first prompt. Reserve enough for retries and for several hook variations.
• Save the prompt that preserves the product best, then reuse it with a new hook instead of rewriting every shot.

What it costs

With VIVID you pay per generation: a 5-second 480p clip is about $0.045–$0.10, a 15-second 720p clip about $0.45–$1.00, and an ad image about $0.015–$0.033. Credit packs from $1, one-time, no subscription. New accounts get 10 free credits — enough for about 3 short clips or 10 images — and failed generations are refunded automatically. Live prices are on the pricing page.

Need the script first? The free UGC ad script generator writes five hooks, a timed 15/30-second script and a video prompt from your product details, no sign-up needed.

FAQ