AI-Powered Ad Platforms to Capture 27% of Spend by 2030

The automation divide

The advertising industry faces a stark bifurcation: AI-powered automated platforms are poised to capture 27% of all ad revenue by 2030—approximately $158 billion—while television advertising struggles to grow faster than inflation, according to projections from Madison and Wall.

This divergence reveals a fundamental shift in how brands allocate marketing budgets. Google's Performance Max and Meta's Advantage+ represent a new category of advertising infrastructure that uses machine learning to automatically optimize spending across channels based on performance data. The momentum behind these systems is driving projected global ad spending growth of 11% in 2025.

Meanwhile, television advertising revenue in North America is expected to grow just 3% in 2026 when political spending is excluded—a figure that barely keeps pace with inflation, despite billions invested in sports rights and considerable industry attention focused on connected TV's potential.

Why it matters

This gap exposes television's structural disadvantage in an increasingly automated advertising ecosystem. As brands demand unified platforms that can ingest first-party data and dynamically shift budgets based on outcomes, TV's fragmented infrastructure and inventory-control philosophy leave it unable to compete with the seamless automation offered by digital platforms. The result: a widening performance gap that threatens TV's share of marketing budgets regardless of audience reach.

The unified platform trend

The rise of what investor Corey Ferengul calls "Unified Media Platforms" further accelerates this shift. These systems promise brands a single login to manage spending allocations across every platform with an API, with machines handling optimization automatically.

Madison and Wall describes this as part of a "broader transformation" where "more businesses are gaining access to sophisticated advertising products, while increasingly automated platforms reduce the complexity and cost required to participate."

Television remains conspicuously absent from this evolution.

TV's complexity problem

Justin Rosen, who leads Measurement Product at IQVIA and previously held roles at Comcast, Turner, and Ampersand, confirms the challenge. "In the case of targeted TV, I certainly know this to be true," Rosen said. "It is extremely complex. The workflows are difficult. Even in something like a more digital native environment, there's still issues around workflow and speed. AI will fix that."

The question is when and how. While every TV seller offers self-serve buying options and data-driven solutions, the industry's ruling philosophy remains inventory control and fragmentation. Cross-platform automated optimization—the kind Madison and Wall projects will dominate advertising—requires cooperation among competing media companies that TV has yet to demonstrate.

The data quality imperative

Rosen emphasizes that automation alone won't solve television's challenges. Brands must "ensure that the data inputs that are going into those AI approaches are rock solid," he warns. "AI is not gonna fix a weak signal. It's gonna make whatever's underneath move faster, so you think you're getting to a better outcome. But if that input right off the bat was weak, you risk getting into trouble."

He calls this the "doom loop of AI" and stresses the need for "the best quality, scaled, first-party data inputs from the start."

Categories like pharmaceutical advertising, with substantial consumer data and strict privacy requirements, would benefit enormously from TV automation—if the infrastructure existed to support it.

These details were first reported by Mike Shields in AI Watch.