Adobe Advertising’s Greg Collison Publishes Framework for Practical AI in Advertising

Article helps marketers distinguish predictive, generative, and agentic AI to drive stronger performance, creative experimentation, and operational efficiency.

Adobe Advertising announced the publication of a new CMO Alliance article by Greg Collison, Head of Product, Adobe Advertising, titled “A Framework for Practical AI in Advertising.” The article explores how AI is reshaping advertising platforms and why marketers need a clearer framework for understanding the distinct roles that different forms of AI play in campaign performance.

“AI is becoming central to how marketers plan, activate and optimize campaigns, but the industry needs to be more precise about what different AI capabilities actually do,” Collison says. “Predictive AI, generative AI and agentic AI each create value in different ways. The real opportunity is not adopting AI for its own sake, but combining these capabilities to reduce complexity, improve creative experimentation and drive stronger outcomes.”

Marketing Technology News: MarTech Interview With Jay H. Lee, Chief Marketing and Growth Officer @ Five9

As AI adoption accelerates across the industry, Collison argues that advertisers must look beyond headline-grabbing features and evaluate how the three core categories of AI can work together to improve business outcomes. He defines each form of AI simply:

• Predictive AI, which underpins performance by analyzing signals and optimizing bidding
• Generative AI, which expands creative possibilities by enabling faster asset creation and variation
• Agentic AI, which simplifies workflows by helping marketers set up campaigns, interpret results, and automate routine tasks.

The article emphasizes that while agentic AI is likely to become one of the most visible areas of innovation in advertising platforms, workflow automation alone does not guarantee stronger performance. Instead, Collison writes that marketers should first assess the strength of a platform’s underlying predictive systems and data foundation, since automation can only optimize within the limits of the performance engine beneath it.

The byline encourages marketers to ask more practical questions of their advertising partners, including whether they are fully using organic site traffic to train predictive models, testing multiple DSPs to understand performance differences, experimenting with generative AI to create more ad variants, and working with partners that are innovating with agentic experiences to reduce manual work.

Collison concludes that practical AI in advertising is not about chasing every new tool in isolation. Rather, it is about using AI across performance, creative, and workflow automation to help marketers eliminate busy work while maximizing return on media investment.