Fragrance Ads Too Hard to Make? Let an AI Agent Break Down Viral Shots for You

For e-commerce in the fragrance category, the biggest pain point is that "scent cannot be digitized." No matter how beautifully the bottle is shot, the screen can't convey the smell—and consumers' desire for "premium feel" and "atmosphere" is hard to communicate through static images. Fragrance ads must therefore translate scent into visuals, letting viewers "see" what the bottle smells like—but concept, set building, shooting, and post-production are each expensive, and producing more versions for testing is even harder.

A recent Bilibili sharing session aimed at cross-border merchants introduced a workflow that uses an AI Agent to assist fragrance video creation. The core idea is "viral replication + product images" to quickly generate presentable ad assets.

The "Expensive" and "Slow" Nature of Fragrance Ads

Fragrance is a typical experiential product—users place orders based on imagination rather than specifications. The traditional approach requires completing the full chain of "creative script—set building—model shooting—post-production color grading," which is long and costly. For small and medium merchants, this means each ad campaign can only bet on one or two versions, unable to batch-test assets and iterate quickly like fast-moving consumer goods.

The video says outright that this kind of creative work is "expensive across the board," and producing more versions to test is "even harder."

How the Agent Breaks Down Viral Shots

The key capability of this solution is "viral replication": merchants upload a reference video and their own product images, and the Agent automatically breaks down the reference video's shot structure—shot size, camera movement, rhythm, transitions—then reorganizes the visuals around the merchant's product.

This is essentially automating "shot-by-shot analysis" (known in film as lapian). In the past, a director or editor had to analyze frame by frame why a viral video grabs attention; now the Agent breaks cinematic language into reusable modules, and merchants only need to provide product assets to get a finished video with a similar structure.

"Shot-by-shot analysis" (lapian) is a professional term in film and television referring to frame-by-frame or shot-by-shot analysis of a film's audiovisual language, including shot size (the switching logic of long shot, medium shot, close-up, extreme close-up), camera movement (push in, pull out, pan, track, follow), editing rhythm (shot duration and cut frequency), and transition techniques. Traditional shot-by-shot analysis relies on the director's or editor's experiential judgment, with the goal of extracting reusable patterns for "why this content grabs attention." When an AI Agent automates this process, it means breaking unstructured video into structured shot parameters and mapping them onto new product assets—reproducing a viral video's narrative skeleton without manual frame-by-frame analysis.

On-Camera Talent: The UJC Ads Workflow

If the content needs a real person on camera to narrate usage scenarios, the UJC Ads workflow is an option. It organizes product selling points into a more immersive video script, forming a complete closed loop from refining creative ideas and confirming visuals to final video generation.

For categories like fragrance that need a "sense of scene," on-camera talent can add emotional value beyond the visuals—such as spraying in a morning bedroom or reapplying fragrance before a date. These scenes are hard to accomplish with product close-ups alone.

"Sense of scene" in fragrance marketing refers to evoking consumers' usage associations through specific life situations, rather than merely displaying the product's appearance. Purchase decisions for experiential products (such as perfume, aromatherapy, and skincare) rely heavily on mental simulation—consumers need to "rehearse" in their minds the image of themselves using the product. On-camera talent works because it provides viewers with a relatable role and situational anchor (such as a morning bedroom or reapplying fragrance before a date), translating abstract scent descriptions into concrete behavioral narratives, thereby lowering the imagination barrier and increasing willingness to convert.

Practical Significance for Cross-Border Merchants

The value of this workflow isn't in "replacing shooting," but in lowering the barrier to asset testing. Merchants can first use AI to quickly produce multiple versions of creative directions, verify which type of cinematic language converts better, and then concentrate budget on validated approaches. For cross-border teams that need localized content across multiple overseas markets, this "run volume first, then select the best" model fits advertising logic better than one-time heavy investment.

As for the tool's specific pricing and generation quality, the original material didn't elaborate—actual results still need to be tested and verified by merchants themselves.