Once again, HeyGen has updated their AI avatar capabilities (and there’s much more to come soon… stay tuned!) But I’m going to focus on just two major updates, HeyGen Avatar IV and Product Placement for influencers and marketers.
Starting with just a simple product image and a script, you can get a stock influencer to do a quick pitch video while holding your product for use on social media. I’m going to show you the steps needed to produce a simple video straight out of HeyGen, as well as a couple alternative workflows to get a high-quality result utilizing a few more steps and some other software tools.
COME LEARN FOR FREE!
I’m giving a FREE online workshop demoing these workflows on September 18th – REGISTER NOW! https://luma.com/mcv4um0u
HeyGen Product Placement
Option 1: Stock Talent & Voice
This is the stock method to produce a quick “influencer style” product video showcasing your handheld product.
Starting on the HeyGen Home page, select the “Product Placement” Tool
This prompts you to either select an existing product with a sample photo from their library or upload one of your own.
I just looked for a can of “something” online (and then wrote my script around it in ChatGPT) and placed the image in the tool.
I selected a stock model from their library that I thought woudl work for this test.
A few minuites later, I was presented with four different blended options to choose from… but notice that the text on the can is distorted and illegible when zoomed in.
I then entered my script text that I modified to emphasize differnt words and phrasing.
Even though I was able to select a good voice through the API for ElevenLabs, the voice controls really are lacking and quite robotic in their raw form. Here’s the rendered video straight out of HeyGen:
Option 2: Stock Talent & ElevenLabs Voice
This is my preferred option, since you can have so much more control over the voices as well as the reactions of the talent, to be more natural – based on the inflections and energy in the voice.
The workflow is the same until you get to the VO script in HeyGen. Then you just upload your pre-recorded audio track. Here, I used a voice in ElevenLabs that worked well with the stock model in HeyGen
And here’s the resulting video, this time choosing a different angle that was offered by HeyGen:
Option 3: Generate Product Image Only
This workflow allows you to use any person’s photo or an AI generated talent and your product to produce only the image of the talent holding the product, which you can then modify in Photoshop or outpaint in Midjourney and bring it back into HeyGen to animate and add the voice to.
Starting with a model I generated with Midjourney, I cleaned up the image in Photoshop and then tried to use it in the HeyGen Product Placement tool.
Unfortunately, my model got completely distorted and the deafulat result is someone’s living room or bedroom background and no way to change it. So I went a different direction this time.
So I had to find another way to get the product into her hand and approach this differently.
I discovered that Krea AI will give me what I needed, after several attempts with other tools, including Adobe Firefly in Photoshop.
I brought the edited image into HeyGen using the Image to Video Avatar IV tool and used the same audio file I had previously created in ElevenLabs with some simple instructions. The results were exactly what I was looking for!
Here’s the finished video straight out of HeyGen:
As you can see, I now have my talent exactly how I want with the green screen background to place her in any location or setting.
Using Photo to Video Avatar IV
This process continues from where we left off with the Option 3 workflow listed above.
Selecting Photo to Video with Avatar IV option from the HeyGen Home page and upload your edited image of your talent and also the recorded voice and SFX from ElevenLabs.
I was tired of the can of Bloody Mary mix so I tried another project using a power tool and another humorous script assisted by ChatGPT. This took a lot longer than I had originally thought because getting the prompts to work correctly for the B-roll shots gave me everything but what I wanted! LOL 🤪 (Wait ’til I release a “bloopers” reel!)
I had ChatGPT assist in the script writing, then went over to ElevenLabs to record my audio and SFX.
I created my model in Midjourney and brought it into HeyGen along with a photo of the drill and used their Product Placement tool to combine them together in the same image – then took that into Photoshop to clean it up and replace the background with greenscreen.
I saved the best image from the Product Placement and brought it into Photoshop to clean it up and replace the background.
I started by using Generative Expand to get the framing right, the did a simple Image Selection to knock out the background
Putting a solid green screen layer underneath, the combined image was what I needed for my avatar.
I brought everything back into HeyGen to use their Photo to Video Avatar IV tool and uploaded the cleaned up image and used my VO recording from ElevenLabs to generate the main video.
I did one pass of B-roll using just text prompts but it failed miserably, so I made a track in ElevenLabs of just “Hmmmmmmm” and “Ahhhhhhhhh” to get the model to move and react. I got some takes with it that were useable, but mostly I used other tools to generate most of my B-roll (when it actually worked!)
All the other B-roll was generated with various tools like Midjourney, Runway, Hailou 2 and the graphic was created with Gemini 2.5 Flash.
Video Editing to take it to the next level…
Of course when you have video editing skills, you can take all of your AI generated content for the your project and edit a finished video for any purpose, with content you’ve generated yourself.
Everything was traditionally edited in Adobe Premiere Pro, along with the music track I created in Suno AI.
NOTE: I know this is pretty lowbrow sexist objectification humor… but we all know that sex sells products typically marketed to men (think tool calendars?). But it’s really all in good fun with extra entendres! 😉
Go check out HeyGen’s new Avatar IV tools yourself and have some fun!