How Unilever Is Using AI to Rethink Asset Creation

To fully capitalise on the efficiencies AI offers, Unilever recognised that rethinking their marketing workflow was just as important as adopting the technology itself. Rather than relying on traditional content production methods, they opted for a "super shoot" - capturing hundreds of modular assets in one go. This provided the raw material for AI-driven content generation, allowing them to create bespoke assets for specific audience segments, product benefits, platforms, and formats. By working with their in-house AI studio, BAIS, and using Pencil’s AI capabilities in collaboration with Oliver, Unilever explored how AI could fundamentally reshape the creative process.

The Challenge

Unilever wanted to change the way they produced marketing assets, moving away from a linear, manual-intensive process to one that leveraged AI for scale and efficiency. The goal was to see how AI could remix content from a single, large-scale shoot to create targeted assets for different audiences, products, and platforms. By integrating AI into their workflow, Unilever aimed to unlock significant time and cost savings while ensuring their content remained high quality and fit for purpose. This shift required not just new tools, but a rethinking of traditional marketing workflows.

The Creative

For TRESemmé, Unilever started with 12 benefit-specific modules and five key audience segments, including professionals, fashionistas, and home stylists. Using AI, these initial assets were transformed into over 200 edits for TikTok, Meta, and YouTube, ensuring tailored creative for each audience and platform.

With Clear, Unilever captured content featuring different lifestyle-based media cohorts, from e-gamers and dancers to basketball players. AI then expanded these into 18 unique creative cohorts with copy and visuals adapted to audiences ranging from fitness buffs to social networkers, ensuring hyper-relevant messaging across multiple touchpoints.

For Elidor in Turkey, the challenge was elevating flat 2D product imagery into premium e-commerce assets. AI-generated 32 elegant visuals featuring soft shadows, refined lighting, and a high-end aesthetic, making the product feel more desirable in an online shopping environment.

Dove Mini Dream in the U.S. leveraged AI to generate 36 bespoke Pinterest Pins—both static and video—tailored to four different audience segments. By remixing 2D product packs, demo videos, and talent imagery, the AI created visually engaging assets suited for awareness and conversion phases.

Vaseline’s campaign targeted their "Competitive Buyer" audience through a mix of Pinterest, display, and online video (OLV) assets. AI helped create 10 digital pieces with fresh opening and closing cards, subtle animations, and messaging designed to build trust and drive action across various platforms.

The Results

The AI-driven workflow proved to be significantly more efficient than traditional methods:

• TRESemmé: Scaled from 12 modules to over 200 AI-generated edits, enabling faster and more targeted content creation. [Do we have any data on time / cost for TRESemmé?]
• Elidor: Achieved a 65.1% cost saving, reducing the cost per asset by ₺1017.81.
• Dove Mini Dream: Saw a 55.4% cost saving (€776.61 per asset) and a 45.3% time saving, cutting production time by 2.6 hours per asset.
• Vaseline Void: Delivered a 43.6% cost saving (€1370.60 per asset) and a 52.3% time saving, saving nearly 5 hours per asset.

By shifting towards an AI-assisted approach, Unilever not only reduced costs and production time but also built a more scalable and adaptable content creation model. This pilot demonstrated that in order to make use of the opportunity this new technology has given us, we also need to rethink how marketing teams operate, and be open to changing the entire workflow.