Observed Signal · Mar 20, 2026 · Industry Analysis · Source: Digiday · Impact: 3/5 · Sentiment: Negative
Agencies’ AI Operating Systems Face Commoditization Risk
Agencies have invested in proprietary AI-powered operating systems (WPP Open, Omnicom’s Omni, Stagwell’s The Machine, Havas Converged, PMG’s Alli, Horizon’s Blu, etc.) to centralize data, planning and creative tools and increase client stickiness. A Gartner research note warned that roughly half of agency-built AI platforms could fail by 2029, citing consolidation, advertisers building bespoke AI or buying from hyperscalers, and shrinking differentiation. Agency and holding-company leaders (e.g., Omnicom’s John Wren, WPP’s Stephan Pretorius) argue scale and negotiated enterprise agreements give them an advantage, while vendors such as Brandtech’s Pencil emphasize transparency about underlying models (Anthropic, OpenAI) and token-costing. The ANA survey flagged principal-media buying is widespread but governance lags. The briefing frames a growing industry debate over AI token economics, vendor transparency and whether hyperscalers will disintermediate agency platforms.
Gartner’s projection and industry commentary signal a potential shift in how agencies compete on proprietary AI platforms, with implications for agency stickiness, vendor economics, and hyperscaler disintermediation.
Track Omnicom Signals & Market Shifts in Real-Time
Polaris7 autonomous intelligence agents track regulatory filings, primary sources, executive changes, and deal flow 24/7. Create your free Explorer workspace to monitor these entities.
Key Takeaways & Evidence Grounding
- A Gartner research note projected that half of agency-built AI platforms would fail (go kaput) by 2029.
- Major holding companies and agency groups have launched proprietary platforms: WPP (WPP Open), Omnicom (Omni), Stagwell (The Machine), Havas (Converged), PMG (Alli), Horizon (Blu), and Meet The People (MTP Intelligence).
- Omnicom CEO John Wren asserted Omni benefits from holding-company scale and negotiated access to AI developers.
- Brandtech’s Pencil uses models from Anthropic and OpenAI and allows clients to port negotiated token agreements into its system, per CEO Will Hanschell.
- ANA survey of 119 U.S. marketers: 58% said their agencies use principal media buying; 37% reported increased use over the last year; 76% cited reduced costs as the primary benefit; only 57% have guidelines and 63% include contract language addressing the practice.
Connected Companies & Entities
5 Entities mappedOntology Mapping & Concepts
Related Market Signals & Shifts
Recent verified developments and strategic activity across this market segment.
Agency AI Platform Claims Face Increasing Scrutiny
A new analysis by 3C Ventures, timed ahead of the Cannes Lions Festival, finds that AI-related language in agency pitches has become standardized while underlying platform capabilities, architectures and maturity vary widely. The report highlights an "accountability gap": agencies frequently present AI-enabled platforms for planning and optimization without clear evidence separating automated output from human operator contributions. It warns about data ownership and platform lock-in as first‑party audiences, attribution models and campaign learnings become embedded in proprietary systems. The study also flags tensions between AI-driven efficiency gains and agency commercial models, noting platform fees can preserve or grow revenue even as labor declines. 3C Ventures recommends marketers demand contractual clarity, proofs of concept, portability commitments, disclosure of technology partners, and audits to verify where AI acts autonomously versus providing recommendations.
Agencies Rethink Ownership of Agentic AI and Data
At Programmatic IO in New York, agency leaders debated the future of data and AI tool ownership. Executives from S4 Capital, Horizon Media, Dentsu, and WPP agreed that agencies should avoid owning client data and AI models to maintain neutrality and avoid conflicts of interest. They highlighted a shift from FTE-based compensation to outcome-based pricing and licensing AI tools to clients. The panel also discussed AI's impact on DSPs, with WPP's McAndrew noting a move toward agent-to-agent interactions and increased supply-chain compression. Sorrell warned of Big Tech platforms like Meta becoming end-to-end threats, pushing agencies to become validators. The discussion reflects a broader industry trend away from data ownership, contrasting with Publicis's acquisition of LiveRamp.
Ad industry wrestles with AI costs and value
As agencies move AI beyond pilots into daily workflows, computing and token usage costs are rising and forcing new governance and pricing models. Agencies and holding companies are experimenting with token caps, pooled access, subscriptions, and output-based fees, but struggle to measure the actual business impact of AI versus token and compute spend. Some firms (PMG, S4 Capital/Monks, Dept, Cheil, Publicis) are testing different approaches to manage or absorb AI costs while clients and procurement often expect lower fees from automation. The article highlights the unresolved measurement problem — tracking token spend is straightforward, attributing business value to that spend is not.
Track Real-Time Market Signals & Shifts
Set up custom watchlists to receive automated, evidence-grounded executive digests whenever material signals or shifts occur across your tracked landscape.