Product Placement Video


Goal	A short, realistic "creator holding the product" video for TikTok / Reels / Shorts
Feature used	AI Auto-Generate, plus Edit with AI and Remove background on the avatar
You'll need	A product photo (a clean shot on any background works) and a line or two of script
Time	About 5 minutes to set up, then a few minutes for the video to render
Best for	Skincare, supplements, drinks, gadgets, apps-in-hand — anything a person can hold up to a camera

How it works (at a glance)

Write a prompt describing the scene and what the avatar says
Pick an avatar (ideally one with a transparent background)
Remove the avatar's background if it has one
Edit the avatar with AI so it's holding your product
Choose a voice (or let the AI pick)
Generate

The trick is step 4: instead of hoping the video AI figures out your product, you give the avatar the product first, then hand that avatar to the video generator.

Before you start

You can find the video tools in your Media Library → + Add Media → Generate AI Video. Click the AI Auto-Generate tab.
Have your product photo ready — either already in your Media Library, or on your device to upload.
Optional but helpful: set up your Brand Voice and Brand Kit so drafted prompts sound like you.

Step-by-step

Step 1 — Turn off the default Layout & Rules

At the top of AI Auto-Generate you'll see the Layout & Rules card. Its default text splits the screen (background scenes on top, avatar on the bottom), which will fight with a "camera on the dashboard" shot.

For this use case, untick Apply on that card. It only skips the rules for this one video — your saved default in Settings is untouched.

Prefer to keep some rules? Instead of unticking, edit the text to keep only what still applies — for example "The avatar must not overshow its teeth while talking."

Step 2 — Write your prompt

In Describe your video, describe the scene, the camera, the product, and the exact words the avatar says. For example:

Place the avatar in a car with a close up shot as if the camera is mounted on the dashboard. Have the avatar say the words "I've been using this for a week now, and I'm already noticing a difference". Make sure the avatar is holding the product in view of the camera. Add natural head movements, just like a creator making a TikTok video.

What makes this prompt work:

Part	Why it's there
"in a car… camera mounted on the dashboard"	Sets the location and the camera angle
"close up shot"	Keeps the avatar and product large in frame
say the words "…"	Puts your exact script in quotes so it's spoken word-for-word
"holding the product in view of the camera"	Tells the video AI the product should be visible, not hidden
"natural head movements, like a creator making a TikTok"	Avoids the stiff, frozen-face look

Tip: stuck on wording? Click + Draft with AI, enter a topic, and it writes a full prompt for you. See Draft with AI.

Step 3 — Choose your avatar

Tick Pin a specific avatar under Options. You now have five choices:

Option	Use it when
HeyGen Avatars	You want a ready-made, professionally filmed presenter (best choice — look for one with a plain or transparent background)
My Avatars	You've prepared an avatar before and want to reuse it
Brand Face	You want your brand's designated face
From Library	The avatar photo is already in your Media Library
Upload Photo	You want to use a new photo — a clear, front-facing portrait works best

Pick one and give it a moment to load. You'll see a thumbnail with Edit with AI and Remove background underneath.

Step 4 — Remove the background (if it has one)

A busy background can bleed into the final video and confuse the AI when it places the avatar in a car. If your avatar's photo has a background:

Click Remove background.
Adobe Express opens with the avatar's photo — let it remove the background.
Click Save image.

The cut-out becomes your new avatar straight away, and the original stays safe in My Avatars.

Already transparent or plain? Skip this step.

Step 5 — Put the product in their hand

Now the important bit. Click ✨ Edit with AI and:

Click + Add reference graphics (or + Upload) and add your product photo.
In the box, describe the change. For example:

Place the product in the person's hand as if they were showing it to the camera.
Click Generate and wait a few seconds for the preview.
Happy with it? Click Accept. Not quite right? Adjust your wording and Generate again, or Discard.

Accepting saves it as a new avatar — holding your product — and it becomes your pinned avatar automatically. This uses a small number of AI credits per generation.

Better wording for better results:

Name the product and how it's held — "holding the bottle at chest height, label facing the camera"
Keep the person's face and clothes unchanged — "keep the person exactly the same"
One product per edit; attach up to 4 reference graphics if you need more detail (e.g. the front and back of the pack)

Step 6 — (Optional) Add the product to Reference files

For extra help, add the same product photo under Reference files using + From Library or + Upload Photo/Video. HeyGen uses these as creative reference, so the product's look stays consistent through the video.

Step 7 — Choose a voice

Under Options:

Leave "Pin a specific voice" unticked and HeyGen's AI picks a voice that fits the video, or
Tick it to choose your own — your Saved voices show first, or switch to All voices and filter by language and gender. Tap the play icon to preview any voice.

Have a brand voice? See Voice settings to save, clone or design one.

Step 8 — Generate

(Optional) Give the video a Title.
Click Generate Video.

That's it. You can close the panel and carry on working — you'll get a notification when it's ready, and it lands in your Media Library automatically. Missed it? Open My Videos at the top of the panel and use Check status.

What you should get

A vertical, creator-style clip: the avatar filmed close up in a car, from a dashboard-height angle, speaking your exact line while holding your product to the camera with natural head movement.

Success checklist

The avatar says your line word-for-word
The product is visible and recognisable in their hand
The setting matches your prompt (car interior, dashboard-camera angle)
Head and mouth movement look natural

Variations to try

Change	How
Different setting	Swap "in a car" for "in a bathroom mirror", "in a kitchen", "at a gym", "walking down a street" in your prompt
Different platform shape	Set Orientation to Portrait (TikTok/Reels/Shorts) or Landscape (YouTube)
Longer testimonial	Extend the quoted script, or use + Draft with AI with Include a full scene-by-scene script ticked
Match your brand	Tick Apply my Brand Kit and Use Brand Glossary so colors and product-name pronunciation are right
Reuse the same presenter	Your product-holding avatar is saved in My Avatars — pin it again next time and skip Steps 3–5
Different products, same presenter	Open the saved avatar, click Edit with AI, attach the new product, and Accept

Troubleshooting

Problem	What to try
The video has a split screen	Layout & Rules is still on — untick Apply (Step 1)
The product isn't in the avatar's hand after Edit with AI	Regenerate with more specific wording ("holding the [product] up in their right hand, label toward the camera"); use a cleaner, well-lit product photo
The product is missing or wrong in the final video	Add the product photo to Reference files (Step 6), and repeat the product in your main prompt
The avatar's face changed in Edit with AI	Add "keep the person's face and appearance exactly the same" to the edit prompt
Avatar won't prepare after removing the background	Try Edit with AI with "place the person on a plain light-grey background" instead, or pick a different avatar photo
Video never appeared	Open My Videos and click Check status on it
Mouth movement looks unnatural	Add "keep mouth movements minimal, natural and in sync with the audio" to your prompt

AI Auto-Generate — every option in the generator
Scene Builder — build the same idea scene by scene, with full control
Settings & Brand Setup — Brand Voice, Brand Kit, saved voices, avatars, and Layout & Rules