Newton Research: AI Agents for Media

Welcome to The Revisionists, a series featuring innovators in the TV ecosystem. In this episode, we talk with John Hoctor, a serial entrepreneur and media disruptor. He is the co-founder of Newton Research, a tech and services company. Newton Research is developing and implementing marketing analytics AI agents in the media world.

Jason Damata: What is Newton Research?

John Hoctor: Newton Research has created Newton. It is a platform with AI agents that are highly specialized. These agents integrate into each step of the media process, from planning and targeting to buying, improving, and reporting. These agents are specifically trained on marketing and media workflows, datasets, and decision patterns. They are unlike chatbots or co-pilots.

Jason Damata: How did the concept develop?

John Hoctor: The idea comes from decades spent in the industry. My co-founder, Matt, and I have worked together for 26 years, most recently at Data Plus Math, focusing on measurement and reporting across linear and CTV. Over the years, we noticed companies wanted more data scientists and analysts. We realized that AI could make sophisticated measurement more accessible, assisting teams in managing data complexity, expectations, and campaign dynamics.

Jason Damata: How does Newton stand out with so much AI investment?

John Hoctor: In the AI age, what makes something stand out is changing. Common tools such as Gemini, ChatGPT, or Claude are good at general tasks, but have problems with industry analytics. If you ask Claude to create a media mix model, it might give you code that is impractical and fragile. Newton is specialized. Newton agents are trained to mirror real-world campaign dynamics, providing marketers with useable analytics.

Jason Damata: How is Newton made so specialized?

John Hoctor: Newton is constantly being taught new methodologies and skills. It’s like giving it a master’s degree in marketing science or a media analytics handbook. When you ask Newton to set up an incrementality test, it will understand the request. It has examples to draw from and understands how to use the correct methodology. Generic chatbots trained on the open web do not.

Jason Damata: How is customer data secured?

John Hoctor: Newton was created for large agencies, brands, and publishers who cannot risk sharing proprietary information. Newton functions as a containerized application where data resides, with no data leaving the customer's environment. Customers are able to teach Newton their own notebooks, methods, or scripts. This training remains in their local environment and never goes back into our knowledge base or algorithms.

Jason Damata: Does this make Newton have business advantages?

John Hoctor: Yes. Customers aren’t turning over data. They are training an extension of their team that is always available. Newton helps overworked analytics teams become more productive and doesn’t replace people. Often, they have used Claude or ChatGPT, but these tools don’t act how they need them to. Newton agents already “think like they think” and are able to be trained further, helping teams add data and analytics into more of their workflow.

Jason Damata: How will Newton impact television?

John Hoctor: Measurement tells you the past. Newton agents allow predictive media modeling, where analytics can be used in real time to shape what will happen. For TV, campaigns are measured while in progress, with insights going directly into optimizing the next decision. Customers are using Newton in this way, changing from retrospective reporting to predictive decision-making.

Jason Damata: Will there be a “battle of the bots” in media optimization?

John Hoctor: I think that bots will cooperate more. Protocols that are emerging, like MCP and agent-to-agent, could allow intelligent agents to work together. If done right, the industry could build an ecosystem that benefits everyone without repeating walled gardens and monopolies.