Skip to content
Social Rocket

The Trade Desk puts greater focus on agentic AI and measurement with Kokai Zuma

The Trade Desk is expanding Kokai Zuma with specialized AI agents, new reporting features and simplified conversion lift measurement.

From Social Rocket editorial team

This text was created with AI assistance and reviewed editorially before publication. All sources are linked below.

Illustrative image: analytics charts for evaluating Meta ad campaigns
Photo: Luke Chesser / Unsplash

The Trade Desk is steering its programmatic platform Kokai with the Zuma version more strongly toward agentic AI and measurement. The move comes in the company's tenth year on the stock exchange and shows where the independent DSP wants to take its technology.

Koa gets specialized agents

Zuma adds a new Koa Assistant and specialized agents for campaign creation, audiences, troubleshooting and performance analysis. The AI is meant to support media teams with recurring tasks and turn optimization suggestions into concrete workflows more quickly.

This means agentic AI is increasingly becoming part of the actual media buying interface. That fits our overview of AI in media buying and the growing automation of ad spend.

Measurement is tied more closely to optimization

Alongside the AI interface, The Trade Desk is also expanding Conversion Lift, reporting and the Report Builder. The goal is to assess campaign impact faster and derive optimizations directly from measurement signals.

For its own platform analysis covering 62 campaigns, The Trade Desk reports better CPA results on average with the updated Koa optimization. Since the data comes from the provider itself, it should be read as a product benchmark rather than a general performance guarantee.

The direction resembles developments in the creative space: Kroger is testing AI creative scoring with conversion data, while Runway is bringing creative production and campaign data together in an agent workflow.

Why this matters for marketers

Programmatic platforms are increasingly competing on how much operational work their AI takes over and how traceable the resulting decisions remain. For advertisers, not only the degree of automation and speed matter, but also control, measurement logic and transparency.

Zuma shows exactly this shift: AI is moving from the recommendation layer to the operating layer, while measurement moves closer to the optimization process.