Google and world's largest ad group announce landmark AI collaboration

The collaboration between WPP, the world's biggest advertising group, and Google is about utilizing Gemini AI to help produce ads for some of the biggest brands in the world. This includes ad narration, voiceover script generation, and product image creation. The partnership aims to drive a step-change in marketing efficiency and effectiveness by combining Google's expertise in data analytics, generative AI technology, and cybersecurity with WPP's marketing capabilities and global creative reach.

Gemini AI can be used for ad production in several ways, as highlighted by the collaboration between WPP and Google. These include:

• Ad narration: Utilizing Gemini 1.5 Pro to produce customizable video narration scripts.
• Voiceover script generation: Scripts generated by Gemini 1.5 Pro are sent to Eleven Labs to create the voice for narrating videos.
• Product image creation: Gemini 1.5 Pro and Universal Scene Description 3D file formats are used to create hyper-realistic 3D product images that align with a brand’s style guidelines.

The benefits of WPP using Google's gen AI tools include:

• Enhanced creativity: With WPP Open Creative Studio, there's the development of richer and more dynamic user interfaces, leading to more creative and on-brand content.
• Smarter content optimization: Improving predictive capabilities for marketing content success before campaign activation.
• Deeper insights into target audiences: The integration enables WPP clients to gain deeper insights, accurately predict and explain content effectiveness, and optimize campaigns.
• Continuous, adaptive optimization: The integration of Gemini models allows for enhancement of traditional marketing tasks and integration of the end-to-end marketing process.

WPP is a UK-based advertising company with a global revenue of $18.9 billion in 2023. It employs approximately 114,000 staff members worldwide and is the parent company of renowned ad firms like Ogilvy, Wunderman Thompson, and VMLY&R.