Kantar turns advertising intelligence into action with Microsoft AI

Brands are creating more content than ever, across more channels, in more formats, and at greater speed with generative AI. How is it possible to optimize so much content before media investment is committed? For Kantar, a global brand growth, data, insights, and analytics company, that shift prompted a rethink about the strategic direction of the company. Historically, a global advertiser might bring Kantar a TV commercial or digital video campaign and ask, “Will this work?” Kantar’s job was to measure effectiveness and predict outcomes, based on decades of proprietary data. In the AI era, however, marketers increasingly need help answering a different question: “How can we make it better?”

Kantar recognized a new business opportunity. “It isn't enough just to be a trusted provider of content evaluation, but to provide greater actionability around it. How can we improve ads that score low? How can we create new versions of an ad? How can we personalize it, adapt it to different platforms, and then close the loop and show that those new versions actually improve the score?” explains Ashok Kalidas, Chief AI Scientist at Kantar.

Reimagining creative intelligence

Kantar’s customers needed to move from insight to action much faster, and Kantar stood in a unique position to meet that need. Working with Microsoft Frontier Company (Forward Deployed Engineers, or FDEs), Kantar transformed its LINK AI ad-scoring platform from a creative effectiveness scoring tool into a broader creative intelligence platform that connects Kantar’s proprietary intelligence, AI models, and creative workflows to help marketers evaluate, optimize, generate and improve content before it goes to market.

At scale, this pivot has enormous impact: A global company with 400+ brands may run tens of thousands of campaigns and millions of creative assets annually, supported by hundreds of millions in ad spend. Optimizing each asset with generative AI can meaningfully improve ROI and business performance.
The shift from simply evaluating and scoring ads for effectiveness to now optimizing and generating new ads in response to the scoring opened the door to broader commercial opportunity for Kantar. “The work we’ve done with Microsoft on LINK AI will help us grow our creative solutions portfolio and drive incremental revenue—not just replace or cannibalize existing revenue,” says Heath Greenfield, EVP of Solutions & Platforms at Kantar.

Connecting intelligence, AI, and media workflows

To support the transformation, Kantar and Microsoft co-engineered a cloud-native platform that would scale securely and reliably, built on Microsoft Azure. For connected intelligence and AI models, Microsoft Foundry provides the foundation for orchestration and governs the AI workflows that power LINK AI scoring as well as the new agentic platform with recommendation, optimization, and generation capabilities. Azure OpenAI in Foundry Models provides image generation, video generation, and reasoning models. Azure Kubernetes Service (AKS) orchestrates creative workflows with the scalable compute layer needed to run more than 20 AI models and automatically handle fluctuating demand for ad scoring, optimization, and content generation. Azure Service Bus manages the high-volume workflows that move assets through scoring and optimization pipelines, while Azure Cosmos DB and Azure Database for PostgreSQL store the creative-performance data, operational metadata, and customer information required to support decision-making at scale.

The collaboration also helped Kantar manage the organizational change behind the technology shift, accelerating the adoption of LLMs as key components of LINK AI to augment traditional ML models. By helping teams understand the value of the new architectures and using LLMs while adopting a scalable approach, Microsoft’s FDEs helped shorten time to value—not only implementing the reasoning agent platform, but helping Kantar put it to work. “The Microsoft FDE team brought strong technical talent, but just as importantly, they helped win people over to a new architectural approach,” says Jonathon Adler, Director, Enterprise Architect. “That support was really important in helping people see the value of the new architecture.”

Scaling creative intelligence with measurable impact

The transformation is already delivering results for Kantar and its customers. Re-architecting the LINK AI platform has improved batch-scoring performance by 20×, reduced platform operating costs by 4×, and increased the speed of single-video scoring by 6x. Customers now move from insight to action in minutes, not hours, and test and refine creative before launch, which helps validate direction earlier, reduces rework, and improves creative before budgets and media spend are committed. LINK AI has now scored more than 100,000 video advertisements in production.

In the generative AI era, Kantar is turning the question “How can we make it better?” into a continuous learning loop for agentic creative optimization: market feedback informs creative intelligence, AI-powered recommendations and controlled generation improve content, and people retain final authority over what goes to market. “The future capability I’m most excited to validate is real-time, in-flight optimization. The ability to use continuous market feedback to tweak and personalize ads on the fly,” says Kalidas. “With the foundation we’ve built with Microsoft, this vision is within reach.” That foundation brings together connected intelligence, scalable execution, and governed AI—helping Kantar move creative optimization from a pre-launch exercise toward a real-time capability.

Learn more about Kantar on Facebook, Instagram, LinkedIn, X, and YouTube.