The End of the "Holding Company": How Agency Models Are Changing in the Age of AI

The traditional agency holding company, built on the promise of massive scale and bulk media buying, is facing a turning point. As AI handles repetitive tasks, the value of the "holdco" is shifting from volume-based leverage to high-level orchestration and specialized counsel. Agencies are no longer just selling ad placement; they are selling the ability to navigate fragmented ecosystems and manage the mix of data, technology, and human strategy. For leaders and marketers, this shift creates a clear advantage: those who move away from siloed, manual-heavy structures toward "capability companies" will survive, while those clinging to traditional scale will be outpaced by the platforms they once negotiated against.

The Shift from "Holding" to "Capability"

The conventional wisdom of the agency world has long been built on the "Costco model": buying media in bulk to secure better rates. But as AI automates the tactical execution of media buying, that leverage is thinning. Ralph Pardo, CEO of Omnicom Media Group North America, argues that the term "holdco" is now outdated.

This transformation is a structural necessity rather than a rebranding exercise. Agencies are moving toward becoming "capability companies," which are specialized entities designed to stitch together fragmented digital touchpoints. This is a move away from the "do-it-all" agency toward a model that prioritizes strategic orchestration.

"I guess maybe one thing let's start to challenge immediately this concept of the holdco, right? We're not holding anything. We are actually capability companies is the better way I would describe it."

-- Ralph Pardo

This transition is difficult because it requires dismantling the siloed structures that agencies have spent decades building. By moving away from separate digital, social, and PR teams, agencies are attempting to create "M-shaped" talent: individuals who possess deep expertise in one trade but maintain secondary fluencies in data and content. This creates a more resilient, albeit more expensive, human capital base.

The Hidden Cost of "Intelligent Scale"

While AI removes the grunt work, it introduces new costs, both financial and operational. Pardo notes that tokens are not free and querying LLMs requires a new kind of investment. This creates a downstream effect where the commercial relationship between agency and client must evolve.

The move toward paying for outcomes rather than paying for people is a direct response to the diminishing returns of manual labor. However, this creates a friction point: procurement teams are used to buying hours, not algorithmic outcomes. The agencies that thrive will be those that successfully educate their clients on this new cost structure, treating transparency as a competitive advantage rather than a liability.

"The fact of prompting, all right, and being able to ask great questions. This is also a skill set now that we are training for. That's different than managing an active performance dashboard."

-- Ralph Pardo

The system is responding to this by forcing a tighter integration between client data and agency strategy. Because data often cannot leave secure environments, agencies are moving into the role of clean room architects, enabling clients to maintain ownership of their data while using the agency to orchestrate the agentic ecosystems that sit on top of it.

Why "Doing Less" Creates More Value

The most non-obvious insight is that as platforms like Meta and Google build more all-in-one automated tools, the agency role becomes more important, not less. The paradox is that when barriers to entry are lowered by AI, competition increases, and differentiation becomes the only true driver of growth.

Clients are not looking for someone to run the dashboard; they are looking for someone to interpret the signal. The agency value is shifting from the how (placing the ad) to the where and why (the strategy of the ecosystem). This requires a shift in mindset from media buying to counseling. The agencies that win over the next 18 months will be the ones that stop acting as vendors of inventory and start acting as the gatekeepers of the client brand identity in an LLM-dominated world.

Key Action Items

• Audit for Silos (Immediate): Identify where your internal structures, such as separate digital versus traditional teams, are preventing a holistic view of the customer. Begin consolidating these into cross-functional units.
• Invest in Prompt Literacy (Next Quarter): Shift training budgets away from dashboard management and toward query-based thinking. Your team needs to know how to interrogate platforms, not just monitor them.
• Redefine Commercial Terms (Next 6-12 Months): Move conversations with procurement toward outcome-based pricing. Account for the non-human costs of AI, such as token usage, compute, and clean room maintenance, now so they are not surprises later.
• Prioritize "M-Shaped" Talent (Ongoing): Stop hiring specialists. Look for individuals who can bridge the gap between media, data, and content. These individuals will be your future business leaders.
• Focus on Data Ownership (12-18 Months): Help clients build their own data foundations rather than relying on third-party platforms. This creates a sticky relationship where the agency is the architect of the client proprietary ecosystem, not just a middleman for inventory.