# How to Make Dropshipping Video Ads With AI (9:16 UGC Workflow) Build a UGC-style creator with text to image, pair it with your product photo in Video Studio, add a voice line and a lip sync prompt, and render a 9:16 MP4 with sound for TikTok, Reels and Shorts. _Andreas Edesberg · 2026-11-01 · Content Creation_ --- The supplier gives you one thing: a product photo on a white background. The feed wants something else entirely, a person holding the product, talking to the camera, filling the whole screen vertically. Closing that gap used to mean paying a creator, shipping them a sample, and waiting a week for a file you have not seen yet. You can close it from the photo you already have. The pipeline below is the whole of it: an [AI UGC video generator](/ai-ugc-video-generator) workflow assembled out of the two studios already in your account, [text to image](/text-to-image) for the creator and [Sloyd Video Studio](/ai-video-generator) for the clip. Video and image generation are part of Sloyd Plus and Sloyd Pro. This guide covers the four steps, the hook and script structure that makes a 9:16 ad watchable, and what a clip costs in credits before you commit a budget to it. ## The Four-Step Workflow for Dropshipping Video Ads Be clear about what Sloyd does and does not give you. There is no dedicated UGC mode and no actor library to cast from, so a dropshipping ad video is a workflow you run across two studios rather than a single preset you pick. That is good news for a dropshipper: the creator you invent is yours, consistent across every SKU you sell, and never books out. Run it in this order. ### Step 1: Create the Creator With Text to Image Open [text to image](/text-to-image) and describe the person you want fronting the ad. Treat it as casting, not as prompting: the age, the setting, the lighting, and the framing all carry more weight than adjectives about quality. Try: _"A woman in her late twenties in a sunlit apartment kitchen, holding a pastel pink portable blender bottle up beside her face, soft window light, natural skin texture, looking straight into the lens, handheld front-camera selfie angle, waist-up vertical framing, no phone or on-screen interface visible."_ Put a product in the hand already, so the portrait has the pose the ad needs. The exact product still comes from your supplier photo in step 2. For a grooming store, cast differently: _"A man in his early thirties in a small bright bathroom, morning light from a frosted window, holding a silicone scalp massager brush toward the lens, natural skin texture, talking straight to the lens, handheld front-camera selfie angle, waist-up vertical framing, no phone, mirror or on-screen interface visible."_ ![Two creator portraits side by side: a woman in her late twenties in a sunlit kitchen holding a pink portable blender, and a man in his early thirties in a bright bathroom holding a green silicone scalp massager](/blog-assets/how-to-make-dropshipping-video-ads-with-ai/ugc-creators.avif) Two things make the rest of the workflow easier: - **Generate the portrait vertically framed and waist-up.** A 9:16 ad needs the face high in the frame and room for the hands to come in with the product. If you later open the clip on that portrait as a start image, the clip takes its shape from the image, so a vertical portrait is what gets you a vertical ad. - **Keep one creator per store.** Save the image you like and reuse it across SKUs so the ads read as one person running one account rather than a rotating cast of strangers. A clear, front-facing face matters most. Profiles and heavily shadowed faces make the talking shot in step 3 harder, because there is less of the mouth for the model to work with. ### Step 2: Bring the Creator and the Product Into Video Studio Now you have two images: your creator and the supplier photo. ![A pink portable blender bottle filled with a berry smoothie on a plain white background](/blog-assets/how-to-make-dropshipping-video-ads-with-ai/supplier-product-photo.avif) Both go into [Sloyd Video Studio](/ai-video-generator). You have two ways in, and they behave differently later, so pick one now: - **References only.** Attach both images in the References slot and let the model steer the look while it builds the shot. You keep the aspect ratio and the resolution settings, so this is the path for a clip that has to be 9:16 and may need the top resolution setting. - **A start image plus references.** Drop the creator portrait into the Start image slot so the clip opens on exactly the face you cast, with the product photo still attached as a reference. The clip then inherits the portrait's shape instead of taking an aspect ratio from you, and the resolution picker narrows to 480p and 768p. A single request can carry up to 9 reference images, 3 reference videos and 3 audio references, 12 references in total, so there is room for the creator, several angles of the product, and the packaging shot in one go. Then describe the beat, naming the product the way a shopper would see it: _"The woman holds the pink portable blender up beside her face, turns it toward the camera, then lowers it and keeps talking, natural kitchen light, the blender stays in focus."_ Prompts run up to 2,500 characters, so there is room to describe the room and the product finish. One action per clip still beats a paragraph of instructions, because every extra demand competes with your product for the model's attention. ### Step 3: Add the Voice Line and a Lip Sync Prompt An ad is a script before it is a video. Write the line first, generate or record it as audio, and attach it as the audio reference. An audio reference has to travel with a visual, so it goes in alongside the creator image, a reference video, or a start image rather than on its own. Then say what the mouth should do, because lip sync here is prompt-driven rather than a toggle: _"She speaks the line from the audio with accurate lip sync, warm and matter-of-fact, a small nod on the last word."_ Name the sync explicitly and name the mood. The longer version of this half of the workflow, including what makes audio clean enough to sync well, is in our guide to [lip sync character videos](/blog/lip-sync-character-videos-with-ai). ### Step 4: Render the 9:16 Clip and Download the MP4 Settle the vertical shape before you generate, not after, and how you do that depends on the choice you made in step 2. On the references-only path, set the aspect ratio to 9:16 yourself; Video Studio also offers 16:9, 1:1, 21:9, 4:3 and 3:4 when the same ad has to run in a feed or on a product page. If you opened the clip on a start image, the aspect ratio setting is greyed out and the clip follows that image, so the vertical portrait from step 1 is doing the job instead. Either way, cropping a landscape render down to vertical afterwards cuts the product out of frame or pushes the face off center, and re-rendering costs less than repairing a bad crop. Pick the resolution setting next. On the references-only path the options are 480p, 768p and 2K; a clip built on a start image plus references tops out at the 768p setting. The choice is a budget decision as much as a quality one, as the credit table below shows. Clips run up to 15 seconds per generation, which is the right length for one hook and one claim. Generate, then watch it back with the sound on. Clips are generated with sound on the paid plans, and that first playback is where you hear whether the voice line and the lip sync landed the way you wrote them. When the take is right, download the MP4 and upload it to TikTok, Reels or Shorts. ## How to Write a Hook and Script for a 9:16 Ad The workflow gives you a clip. The script is what makes it an ad. The UGC ads that hold attention in a vertical feed tend to run on three beats, and all three fit inside a single short generation. ### The first seconds have one job Someone is scrolling. The opening has to be a reason to stop, not an introduction. Lead with the product in the hand, the problem on the face, or a line that sounds like the end of a sentence rather than the start of a pitch. Lines that work as openers: - _"I have been using this for a week and I am not going back."_ - _"If your hair looks like this by noon, this is for you."_ - _"Okay, this is the thing I keep getting asked about."_ Avoid an opener that spends its first second on your brand name. Nobody has agreed to listen yet. ### The middle is a problem and a solution One problem, one solution, nothing else. Name the irritation the way a customer would say it out loud, then show the product solving it. This is the beat where the supplier photo earns its place: the label turns toward the camera, the lid opens, the hand demonstrates. Keep it to one claim. A clip that promises three benefits lands as none. ### The CTA is a specific instruction Finish with the one action you want. "Link in bio" is fine when it names what is at the end of the link. Better: _"It is in my bio, the bundle is the one to get."_ Say it while the product is still in frame so the last thing on screen is the thing you are selling. A practical rhythm for a vertical ad: the hook in the opening seconds, the problem and the product in the middle, the CTA on the last beat with the product held up. If the script does not fit in one clip, that is two clips and a cut in your editor, not a longer prompt. ## What Dropshipping Ad Videos Cost in Credits Credits are charged per second of finished video, and the rate follows the resolution setting you chose in step 4. | Resolution setting | Credits per second | A 10-second clip | | ------------------ | ------------------ | ---------------- | | 480p | 1 | 10 credits | | 768p | 2 | 20 credits | | 2K | 4 | 40 credits | The 2K row applies to a clip built from references only. A clip that opens on a start image is offered 480p and 768p, so its ceiling is 20 credits for a 10-second take. The practical read for anyone testing dropshipping ads at volume: draft at 480p. When you are iterating on a hook, you are judging the script and the framing, and the cheapest setting answers both questions at a quarter of what the 2K setting charges. Once a hook earns its place, re-render the winner at a higher setting and run that one as the paid creative. Video generation is part of Sloyd Plus and Sloyd Pro, so a subscription comes before the first clip. Compare the tiers on the [pricing page](/pricing). ## Mistakes to Avoid With Dropshipping Ad Videos - **Rendering landscape and cropping later.** Decide the vertical shape up front, with the aspect ratio on the references-only path or with a vertical start image. The vertical frame is the format, not a post-production step. - **Recasting the creator every time.** One saved portrait across every SKU builds a face people recognize in the feed. A new person per ad builds nothing. - **Writing the script after the clip.** The voice line is an input to step 3, so the words come first. - **Asking one clip to carry three claims.** One problem, one solution, one CTA. - **Testing hooks at the top resolution setting.** Draft cheap, then re-render the winner. - **Leaving the product photo out of the references.** Without it the model invents a product that resembles yours instead of showing the one you ship. ## FAQ ### How do I make a dropshipping video ad with AI? Create a UGC-style creator in text to image, bring that portrait and your product photo into Video Studio as references or with the portrait as the start image, attach a voice line as the audio reference with a lip sync prompt, render it vertical, and generate. Download the MP4 and post it. ### Does Sloyd have a UGC mode or an actor library? No. There is no dedicated UGC mode and no cast of stock actors. You invent the creator yourself in text to image, which is why the same face can front every ad in your store. ### What does a dropshipping ad video cost in credits? Credits are charged per second of finished video: 1 credit per second at the 480p setting, 2 at 768p and 4 at 2K. A 10-second draft at 480p is 10 credits. ### Which plan do I need? Video and image generation are part of Sloyd Plus and Sloyd Pro. Compare the tiers on the [pricing page](/pricing). ### What aspect ratio and file format do I get? Choose 9:16 for TikTok, Reels and Shorts. Video Studio also offers 16:9, 1:1, 21:9, 4:3 and 3:4, and generated clips download as MP4. A clip that opens on a start image takes its shape from that image instead, which is why the creator portrait is worth generating vertically. ### Do the clips have sound? Clips are generated with sound on the paid plans, and the lip sync follows the audio you attached as a reference. Play the first take back with the sound on to check the voice line before you spend credits on a higher resolution setting. ### How long can one ad clip be? Up to 15 seconds per generation. That covers one hook, one claim and one CTA. A longer ad is several clips cut together in your editor. ## Make the First Ad From the Photo You Already Have A dropshipping ad stops being a production once the creator, the product and the voice all come out of the same account. Cast the creator in [text to image](/text-to-image), bring the product photo in beside it in [Sloyd Video Studio](/ai-video-generator), attach the voice line, render it vertical, and download the MP4. See what else video and 3D do for a storefront on the [e-commerce page](/use-case/e-commerce), and for the product-only version of the same idea, read [how to make a 3D product video for e-commerce](/blog/3d-product-videos-for-ecommerce). Ready to shoot? [Open Video Studio](https://app.sloyd.ai/video) and compare plans on the [pricing page](/pricing). --- Source: https://www.sloyd.ai/blog/how-to-make-dropshipping-video-ads-with-ai