[PR] I tried making a commercial with acorn beauty oil | For ad production | Pollo AI | GPT Image 2 | Seedance 2.0

*This article is a PR article.

Advertising videos, in the end, require money and people.

I had always thought so. You need a cameraman, a model, and an editor to make something that 'looks the part.' I thought it was a world a bit too far away for individuals or small teams.

That changed somewhat when I started using Pollo AI.

This time, using a fictional organic beauty oil brand called 'TOBUKI' as the subject, I recorded the process of creating everything from posters to keyframes and a 30-second commercial all at once. I hope this serves as a reference for brand managers and those involved in ad production.

🌿 What kind of tool is Pollo AI?

Pollo AI is a platform that allows you to use features necessary for content creation, such as image generation, video generation, and template utilization, all in one place.

I mainly used two features this time.

• Generate images from text or images GPT Image 2

• Generate videos from text or images Seedance 2.0

Even without special video editing skills, you can get a video just by preparing prompts and reference images. I believe that is the biggest feature of Pollo AI. The UI is simple, and despite the many features, it was easy to navigate.

🌰 I built the brand TOBUKI from scratch

Before making the video, I decided to design the brand itself.

Even for a fictional brand, without a worldview, the instructions for the video would be vague. What kind of product is it, what are the colors, and what is the atmosphere? I felt that if those weren't decided, I would get inconsistent results every time I generated something. I wanted to prevent that inconsistency in advance.

The theme was 'an acorn becomes a big tree'. A small seed sleeps in the soil and takes time to become a large tree. I layered that image with the concept of beauty oil 'nurturing the skin from within.' Rather than flashiness, I wanted to focus on something like trust in time and nature.

It took some time to decide on the brand name. I tried several directions, including English-based, French-based, and Japanese-mixed, and finally chose 'TOBUKI.' It is a coined word derived from 'mebuki' (sprouting). The roundness of the sound and the direction of the meaning matched perfectly.

The main ingredients consist of three components: acorn (oak)-derived oil as the base, plus rosehip and jojoba. The bottle has a rounded form reminiscent of an acorn, and I chose pastel cream and light green colors. This was the result of prioritizing cuteness that makes you want to pick it up over being stylish.

As the worldview solidified, I was a bit surprised myself that I started to feel, 'I want this to actually exist.' Even though it was supposed to be a fictional brand.

🖼️ Started with a poster

After the world-building was decided, the first thing I made was an infographic-style poster.

I wanted to jump straight into video, but generating video without reference material causes the visuals to drift between cuts. I decided to first solidify one poster as the "correct version" of this brand.

The composition features the bottle in the center, with the ingredient ratios (Oak Acorn Oil 45%, Rosehip Oil 30%, Jojoba Oil 25%) scattered on the left and right. I used botanical acorn icons for each ingredient, connecting them to the bottle with hairline strokes.

When I first added a donut chart, it looked much tackier than I had imagined. I removed it immediately and changed the layout to scatter ingredient cards on the sides, which made it look much more professional. This kind of "try it once, then fix it" approach might just be the rhythm of using generative tools.

As a preparatory step before moving into video, this process was more effective than I expected.

🎬 Gradually gathered the materials

Once the poster was done, the next step waskeyframes.

I divided the 30-second commercial into two 15-second parts and prepared a total of 8 keyframe images. I decided on the story first, then carefully articulated the intent of "what I want to show in this scene" for each cut before generating them.

For example, the opening scene of the "acorn in the soil" features a single acorn in dark soil with a ray of light shining down from above. The image is that the light awakens the acorn. When this cut came out, it had the intended stillness, which made it a great first keyframe.

On the other hand, the cuts featuring a person were difficult. There are more constraints than I thought when trying to create a character that is easy to empathize with while keeping their face out of view. So, I created a separate reference image that combined a back view, hands, and a silhouette with arms extended into a singlethree-view drawing. By attaching this image to the keyframe generation, I used it as a standard to maintain consistency in the character's clothing, hair, and silhouette.

Creating keyframes first was also a way to avoid wasting video generation credits. With a reference image, the range thatSeedancehas to fill in for motion is narrowed, making it easier to get results closer to my intent. I felt that whether or not you follow this procedure significantly changes the credit consumption required to reach the final product.

🎥 The video is finished

PART 1 is "From Reality to the Dream Forest".

It follows a flow where I pick up the bottle at the washbasin, drop oil onto my wrist, and the moment it touches, gold particles spread and the room dissolves into a forest. I wanted to express the image of the world changing the moment of contact through movement.

PART 2 is "From the Dream Forest to the Product".

I designed it to pull the camera back from the woman standing in the forest to show the scale of the entire forest, then land on the falling acorn, the large tree, the product shot, and finally the logo. To avoid monotony, I designed the focus to switch between the person, the space, the movement, the light, and the product for each cut.

To share my impressions first, it felt more "established as a film" than I had expected. I was a little surprised that the movement of the particles and the transformation of the light, in particular, were close to my instructions. When something I had "vaguely imagined" appeared on the screen, I actually paused and watched it again.

I edited the two parts together to finish it as a 30-second commercial. I used Microsoft Clipchamp for the editing.

✂️ Things I noticed during editing

When I connected them in Clipchamp, the first thing I felt was that the tones of the two videos were closer than I expected.

Since PART 1 transitions from warm to cool colors and PART 2 returns from cool to warm, they become symmetrical when connected. That was something I intended, but I didn't know if they would really match until I actually tried it.

When I added a 1-second crossfade, the cut became almost unnoticeable. As a flow of footage, it looks connected.

💼 Can this be used for work?

I will write honestly. I think the answer changes depending on your position.

To brand managers

I thought it could be used in the early stages of planning, especially for "image sharing" before getting internal approval. Even if it's not a finished product, you can show them visually, "I want to make a video with this kind of worldview." It's much easier to convey than a proposal document with just words.

However, to maintain the brand's tone and consistency in details, it takes time to design the reference images. If you don't do that carefully, you'll get inconsistent results from cut to cut. To be honest, the effort doesn't drop to zero.

To agencies and small teams

No filming required, no models needed. These two points alone significantly change the structure of production costs. For example, I felt that using it to quickly create and show a mock-up worldview video for a new client proposal is quite realistic.

However, I am still a bit cautious about the question of "is it quality that can be delivered as a finished product?" I think there are times when final confirmation and adjustments by human eyes are necessary for use in important brand campaigns.

To EC operations and product managers

If you have one product image, you can create moving assets. This produces content at a level that can be used directly for SNS ads and LP videos. Especially the way of creating teaser videos before shooting new products has significant benefits in terms of both cost and time.

This was the area where I felt it was "ready to use right now."

🌳 Conclusion

An acorn is just a small seed as it is.

It is only after being buried in the soil, rained on, exposed to light, and given time that it becomes a tree. During that process, it is hard to tell what is growing.

I was thinking about things like that while creating TOBUKI.

Pollo AI is a tool that I think shortens that process a little.

It doesn't mean everything becomes automatic. But I was able to get this far on my own. That is my most honest impression right now.

Afterword

Thank you for reading this far.

I intended to create a fictional brand, but before I knew it, it had become quite authentic. I think the TOBUKI bottle is cute, if I do say so myself. If acorn oil existed, I would probably buy it. Depending on the price, lol.

Pollo AI I hope the appeal of it has been conveyed even a little.

Following or liking this post would be encouraging!
I look forward to your continued support!

#PR #PolloAI #GenerativeAI #ImageGenerationAI #AIVideo #VideoProduction #VideoEditing #Marketing #Branding #ECSite #note #Creative #TriedIt #Design #Advertising #WorkSkills #BusinessEfficiency #OrganizingThoughts #RecentLearning #AdultLearning