What Newton Shared at CES: How Agentic AI Is Changing Analytics and Media Buying
CES is often where big ideas get introduced and this year, one theme cut through the noise: agentic AI is no longer theoretical. It’s moving from experimentation to execution, especially in media buying.
Newton Research demonstrated at CES the first real, end-to-end example of agentic AI planning, buying, and executing premium media across linear TV and streaming.
From Theory to Practice
For years, media teams have relied on disconnected tools to plan, buy, and measure campaigns. What we demonstrated at CES was a shift from AI as a helper to AI as an agent of action. Newton is built to connect planning, buying, execution, and learning into a single, continuous system, one that doesn’t stop at insight, but carries decisions all the way through to activation.
Planning and Execution With Unlimited Analytics
In the agentic buying demo seen below, Newton began with scenario planning powered by Marketing Mix Modeling (MMM). Whether teams bring their own MMM or build one directly in Newton, the system uses it as a living intelligence layer, continuously learning from performance and shaping future allocations.
In seconds, Newton analyzed prior-quarter performance, identified over- and under-performing channels, and produced an optimized budget scenario with projected lift and efficiency gains — work that typically takes weeks.
From that optimized scenario, Newton automatically generated a detailed media plan with publisher-specific buying briefs across premium linear TV, and streaming.
Each line item then became an executable action, allowing the client to seamlessly place all of the buys within the workflow.
Why This Matters for the Industry
This evolution is meaningful for agencies and brands alike. Agentic systems fundamentally change how work gets done:
- Operational bottlenecks disappear as agents take on execution-heavy tasks
- Speed increases dramatically, compressing weeks of manual effort into moments
- Strategy stays central, with humans setting goals, guardrails, and approvals, while agents handle negotiations, activation, and optimization
- Learning compounds, because every outcome immediately informs the next decision
Instead of handing work from planning tools to buying platforms to reporting dashboards, Newton’s agents stay connected across every step, learning from outcomes and immediately informing what happens next.
Premium Video as the Proof Point
At CES, we announced an industry first: the first end-to-end agentic AI media buy across premium linear TV and streaming, including live sports, in partnership with NBCUniversal, FreeWheel, and RPA.This announcement marked the first time ever AI agents automated buying live sports inventory on linear television.
We also announced another collaboration with RPA, Yahoo DSP and Locality. This collaboration was designed to streamline video investment planning and execution, integrating intelligent workflows and audience insights across platforms.
Reducing Friction Across the Ecosystem
One of the most important shifts we highlighted is agent-to-agent collaboration. Newton’s agents don’t operate in isolation, they communicate directly with publishers, platforms, and partners, reducing friction across the buying process. The result is smarter diversification, better analytics, faster execution, and clearer accountability, without adding operational overhead.
What Comes Next
CES reinforced what we’ve been building toward: a future where media operations are no longer constrained by manual processes or fragmented systems. Agentic AI unlocks scale, speed, and intelligence, without sacrificing control.
If you’d like to explore what an agentic system of action could unlock for your agency or clients, we’d love to continue the conversation.
Book a meeting with the Newton Research team today.
