Key Takeaways
• Pencil, a unit of Brandtech Group, integrates Google’s Veo 2 to generate video content in under 15 hours.
• The system combines Google’s Gemini, Imagen 3, and Cloud frameworks into a unified creative pipeline.
• Japan Airlines used the AI platform to produce in-flight videos that promote Japanese tourism.
• Over 50 pilot projects have demonstrated up to 90% efficiency gains in content production.
• Human oversight remains essential to prevent AI errors and ensure brand-safe content.
Generative AI is accelerating change in the advertising industry, offering solutions that streamline traditionally time-intensive creative processes.
One of the most notable recent advancements is the collaboration between Pencil, a creative technology division of the Brandtech Group, and Google, aimed at optimizing high-quality content production through the integration of Veo 2, an AI-driven video generation model.
This partnership marks a significant shift in how AI can be deployed at scale in agency settings—reducing production cycles from weeks to hours and delivering content tailored for platforms ranging from mobile to in-flight entertainment.
Technology Stack: Veo 2 and Its Supporting Models
The Veo 2 integration draws on Google’s broader AI infrastructure to deliver a full-spectrum creative suite. Rather than operating in isolation, Veo 2 is embedded within a multi-model environment that includes:
• Gemini: Google’s large language model, used for scripting and content logic.
• Imagen 3: A text-to-image model that helps generate creative visuals.
• Google Cloud’s ABCD framework: Used for asset performance analysis and optimization.
Together, these tools allow agencies to move from a prompt to a polished, brand-ready video in a fraction of the time previously required—without losing fidelity or creative control.
Real-World Deployment: Japan Airlines Case Study
A prominent example of the system in action is a campaign developed for Japan Airlines (JAL). Pencil and sibling agency Jellyfish created a series of AI-generated travel videos for in-flight entertainment, promoting various destinations across Japan.
“The quality of marketing on lots of channels, like little screens on the back of a seat on an airline can be made better because now there are resources to make that marketing better,” said Will Hanschell, CEO and co-founder of Pencil. “That has an implication for the brand, yes, but also for the consumer, because now they’re seeing content that’s slightly more relevant to the time of year or the flight they’re on.”
“We are always thinking of ways to help travelers curate memorable itineraries and learn more about traditional and unique travel experiences throughout Japan,” said Minako Kent, JAL’s managing director of global marketing. “The advances in text-to-video through Google’s Gemini models allow us to create content inspired by the incredible diversity of Japan’s destinations, offering a new level of engagement for our customers.”
Speed to Market: 15 Hours from Concept to Delivery
The most compelling metric behind this partnership is the production timeline. From ideation to execution, Pencil’s team completed the Japan Airlines campaign in under 15 hours.
This efficiency was made possible by combining AI generation with human quality control—a hybrid process designed to ensure creative integrity and accuracy.
“It’s, in many ways, the first commercially-ready video generation model that comes with the enterprise-grade backing of Google’s indemnity for marketers, who are using it to generate content,” said Hanschell. “And it’s ready to be used to make ads that are going to go live in the real world.”
While the AI handles the heavy lifting, the process does not eliminate human input. Final videos undergo quality checks by human creatives to ensure factual accuracy, emotional resonance, and brand alignment.
Performance Metrics: Efficiency Gains in the Pilot Phase
A source close to the project revealed internal data from over 50 pilot projects, conducted in collaboration with Google marketing teams. The pilots highlight substantial improvements in production efficiency:
• 80,000 content generations produced during trials
• 25,000 final assets selected and used in campaigns
• 20–30% time and cost savings for premium, customized creative work
• 80–90% efficiency gains in scaled executions like programmatic video and digital ads
“AI is helping agencies revolutionize marketing for their clients,” said Lisa O’Malley, senior director of industry products and solutions at Google Cloud, who showcased Jellyfish’s JAL campaign at Google Cloud Next 25.
These early results suggest a compelling case for broader adoption, especially for brands managing multi-market or high-volume content strategies.
Balancing Speed with Oversight
Despite the automation potential, Pencil’s leadership underscores the importance of human-in-the-loop safeguards. While hallucinations in large models are decreasing, they remain a known risk, particularly in marketing contexts where factual and emotional accuracy is paramount.
Hanschell noted that no client-facing content is shipped without final human review, ensuring that each asset maintains the integrity, tone, and brand consistency expected in commercial campaigns.
December 17, 2024: Veo 2 and Imagen 3 Set New Standards for High-Quality Video and Image Generation! March 18, 2025: xAI Expands as Musk Acquires GenAI Video Start-Up HotShot! March 26, 2025: OpenAI Unveils AI Image Tool in ChatGPT
For more news and insights, visit AI News on our website.
