Input: product photo, platform, and one benefit
Output: 9:16, 1:1, and 16:9 cuts of one approved clip
Aspect: 9:16, 1:1, 16:9
Build log
1. Write the platform hook
TikTok Shop needs motion in the first second; Xiaohongshu needs a context moment; listing previews need the product and offer. I keep one benefit per platform, not a feature list.
2. Build the master cut
I generate 9:16 first because it contains the most information. This is the only render I iterate until the product, hook, and CTA pass.
3. Crop without rerendering
After the master passes, I export 1:1 and 16:9 from the same clip in the external editor. If an important label falls outside 1:1, I move the safe area instead of changing the prompt.
4. Set duration by platform
A 5-second clip is a clean test. Longer shots can work, but Kling v3 Omni snaps 5 or 10 seconds and Veo 3.1 snaps 4, 6, or 8 seconds, so check the bucket before you promise a platform duration.
5. Save channel presets
I save the ratio, hook text, model, and duration in the asset library. The next SKU reuses the same channel template.
Complete execution record
Task boundary and source gate
Treat Channel product video AI as a deliverable, not a definition. The job receives product photo, platform, and one benefit and owes 9:16, 1:1, and 16:9 cuts of one approved clip in 9:16, 1:1, 16:9. Write the rejection reason first, then turn it into the first hard constraint.
Do not start until the source photo has a visible product code, a note on material and color, the approved claim, and a named target module. Check packaging damage, sticker state, lighting direction, and background rights before generation. Do not invent certifications, competitor claims, price windows, or a model using the product in an unsafe way. If the seller cannot prove it, the image does not carry it.
The imgmov route
Open Asset Library first, attach the master SKU as the locked reference, then build the module or shot around that reference. Keep offer copy, platform text, and CTA outside the visual prompt. In imgmov, the approved image becomes the source of truth and generation supplies only the requested scene, angle, or motion.
Keep platform requirements outside the model prompt. The prompt can describe a scene; a named checklist records the crop, claim, consent, and console check that make it deliverable.
Lock the variables that cannot drift
Lock product silhouette, label spelling, logo position, material, primary color, accessory count, contact shadow, and the platform safe area. Scene, props, and motion may change; these fields may not.
Put the locked fields at the top of the prompt, not at the end: One master clip is cheaper than four new prompts. Lock the product and shot order, export 9:16 first, then crop approved footage to 1:1 or 16:9. At 5 seconds, Agnes Video is 10 credits; Seedance 2.5 is 90 credits at 480p or 195 credits at 720p. Save that prompt with the reference, ratio, and model so the next run starts from a decision instead of a guess.
Step-by-step build path
1) Write the platform hook: TikTok Shop needs motion in the first second; Xiaohongshu needs a context moment; listing previews need the product and offer. I keep one benefit per platform, not a feature list. Leave one checkable artifact from this step; do not start the next until it exists. 2) Build the master cut: I generate 9:16 first because it contains the most information. This is the only render I iterate until the product, hook, and CTA pass. Leave one checkable artifact from this step; do not start the next until it exists. 3) Crop without rerendering: After the master passes, I export 1:1 and 16:9 from the same clip in the external editor. If an important label falls outside 1:1, I move the safe area instead of changing the prompt. Leave one checkable artifact from this step; do not start the next until it exists. 4) Set duration by platform: A 5-second clip is a clean test. Longer shots can work, but Kling v3 Omni snaps 5 or 10 seconds and Veo 3.1 snaps 4, 6, or 8 seconds, so check the bucket before you promise a platform duration. Leave one checkable artifact from this step; do not start the next until it exists. 5) Save channel presets: I save the ratio, hook text, model, and duration in the asset library. The next SKU reuses the same channel template. Leave one checkable artifact from this step; do not start the next until it exists.
Change one named variable per rerun: prompt, reference, camera, duration, model, ratio, or export crop. If two variables change together, a better output cannot be reused because nobody knows which fix worked.
Review gates and evidence
Review at 100%. Read small text, count buttons and parts, compare color against the source, inspect the four corners, and overlay the platform crop. A thumbnail is only a concept check, not a marketplace check.
Keep source, approved wording, rejected version, correction, credit cost, and final crop next to the asset. The record exists so the next reviewer can reproduce the decision without a chat thread.
Failure diagnosis and retry ladder
Name the failure first: label redraw, color drift, shape drift, missing accessory, wrong contact shadow, or cropped CTA. Stop prompting for more letters, reduce the variable count, return to the reference, and only then upgrade the model.
Do not upgrade on hope. A 5-second Agnes Video probe is 10 credits. If the concept passes, Seedance 2.5 costs 90 credits at 480p or 195 at 720p for 5 seconds; Kling is 195 without audio or 245 with audio. Veo enters only when a 4/6/8-second cinematic bucket is genuinely worth 80/125/165 credits. Prove hook, subject, and rhythm first, then pay to clean up motion that already worked.
Versioning and handoff
Deliver by module and ratio, not one stretched master. Name files with SKU, module, crop, version, and approval state. Keep the rejected render and the correction line in the same folder.
The folder is boring on purpose: source, approved reference, generation settings, caption file, platform cuts, QA screenshots, and one correction line. A reusable asset is boring in the right way.
Why this is not a generic answer
The imgmov advantage is the chain: Asset Library preserves the subject reference, Canvas fixes shot order and first/last frames, the workspace routes Agnes to Seedance, Kling, or Veo only after proof, and your external editor adds captions and CTA without regenerating the media.
Do not invent certifications, competitor claims, price windows, or a model using the product in an unsafe way. If the seller cannot prove it, the image does not carry it. If the request only asks what channel product video ai means, a search page is faster. This page is useful when someone must deliver 9:16, 1:1, and 16:9 cuts of one approved clip under real constraints.
Frequently asked questions
Which ratio should I render first?
9:16. It carries social hooks and can still crop to 1:1 if the safe area is correct.
The workflow is specific about its stopping point: it starts with product photo, platform, and one benefit and stops at 9:16, 1:1, and 16:9 cuts of one approved clip. In imgmov, upload the source, lock it as a reference, set 9:16, 1:1, 16:9, run one proof, and save the passing settings. If a reviewer cannot tell which version is approved, the workflow has failed even when the media looks good.
How do I avoid four expensive renders?
Approve one master clip, then export platform crops. Rerender only when the composition is impossible.
Do not upgrade on hope. A 5-second Agnes Video probe is 10 credits. If the concept passes, Seedance 2.5 costs 90 credits at 480p or 195 at 720p for 5 seconds; Kling is 195 without audio or 245 with audio. Veo enters only when a 4/6/8-second cinematic bucket is genuinely worth 80/125/165 credits. Prove hook, subject, and rhythm first, then pay to clean up motion that already worked.
What is the cheapest first test?
A 5-second Agnes Video clip is 10 credits. Guest previews are watermarked at 1K.
The check is not “does it look AI-nice?” It is: Review at 100%. Read small text, count buttons and parts, compare color against the source, inspect the four corners, and overlay the platform crop. A thumbnail is only a concept check, not a marketplace check. Then the source, approved copy, rejected version, correction, and final crop stay in the same handoff folder.
How long should a TikTok Shop clip be?
Start with 5-8 seconds. Keep the product in motion in the first second and one CTA at the end.
Name the failure first: label redraw, color drift, shape drift, missing accessory, wrong contact shadow, or cropped CTA. Stop prompting for more letters, reduce the variable count, return to the reference, and only then upgrade the model. Change one named variable, keep the old version, and record the credit cost. If the same failure repeats, fix the reference or scope rather than asking the prompt for forgiveness.
Should text be generated in the video?
No for prices and claims. Add those in your external editor so you can change them without regenerating motion.
The imgmov advantage is the chain: Asset Library preserves the subject reference, Canvas fixes shot order and first/last frames, the workspace routes Agnes to Seedance, Kling, or Veo only after proof, and your external editor adds captions and CTA without regenerating the media. The handoff is reusable only when the source, reference ID, prompt, model, cost, rejection reason, and approved cut are linked. That chain is what makes the next channel product video ai task faster.