What Holding Group Agencies Won't Tell You About Their AI Capabilities
The major holding groups — WPP, Publicis, IPG, Omnicom, Dentsu — have collectively made billions of dollars in announced AI investments over the past three years. They've branded their AI platforms, hired chief AI officers, and produced enough thought leadership content to fill several libraries. What happens inside the client engagement is frequently at a significant remove from the capability implied by these announcements. The AI tools that get positioned to clients during pitches are often centralized platforms built by a small innovation team, and the account teams actually running campaigns may have limited training on those tools, limited access, or limited incentive to use them given how their work is billed and evaluated.
This isn't a conspiracy — it's an organizational reality. Holding group agencies are large, distributed organizations with thousands of employees across dozens of markets. Deploying a genuinely new capability at scale across that organization requires change management, training, new process design, and often compensation model adjustments. These are hard, expensive, slow things to do in organizations that are also managing client relationships, quarterly earnings pressure, and significant operational complexity. The gap between announced capability and delivered capability is a natural consequence of organizational scale, not a deliberate deception.
The practical implication for clients is that the AI capabilities you should evaluate are not the ones in the pitch deck — they're the ones being used by the specific team that will actually run your account. Ask to speak with the people who will be day-to-day on the business and ask them directly: what AI tools do you use in your daily workflow? How has your process changed in the last 12 months because of AI? What are you doing with AI on your current clients that's producing measurable results? The answers to those questions will tell you far more than any platform demonstration from an innovation team that won't be in the room after the pitch.
Independent and mid-size agencies often have a genuine capability advantage over holding group agencies in AI adoption precisely because of their scale. An independent agency can run a new AI tool across its entire creative team in a single workshop. It can restructure its workflow around AI-assisted production in a quarter rather than a two-year transformation program. It can hire and incentivize based on AI fluency without navigating the political complexity of redefining roles in a 10,000-person organization. The result is that many independent agencies are running AI-enabled processes that are further evolved than what holding groups are doing on most accounts, without the resources to announce it at Cannes.
The clients best protected against the gap between AI hype and AI delivery are the ones who write explicit AI performance expectations into their contracts and scope documents. Not vague commitments to 'AI-enabled creative processes,' but specific, measurable capabilities: creative variant generation within defined timeframes, AI-assisted analytics reporting on a defined cadence, named tools in use for specific functions. When AI capability has to be delivered and measured rather than announced and implied, the honest conversation about what's actually ready starts happening in the right place — before the contract is signed rather than after.