London-based Kohort has closed a $7 million (€5.9 million) Series A funding round to build AI agents that handle user acquisition for mobile game studios. The Raine Group, a New York-based merchant bank focused on technology, media, and telecommunications, led the round after also backing Kohort’s seed stage.
What Kohort Is Building
Kohort, founded in 2018, operates an ML-based predictive analytics platform for the gaming and consumer sectors. The Series A funds three agent capabilities that sit on top of that prediction layer.
The first is campaign optimization through Ktrl, the company’s flagship product. Ktrl generates network-specific bidding strategies and targets for every campaign a studio runs, supporting ROAS, CPI, and CPE/CPA campaign types. It integrates directly with ad networks to push each campaign toward 100% return on ad spend.
The second is on-demand research. Studios can query any UA trend and get results cross-referenced against their own historical data, benchmarked against roughly $1 billion in annual spend flowing through Kohort’s platform. According to the company, this replicates the investigative work of a senior UA manager.
The third is automated reporting: tailored decks generated for every function from CEO to product to live operations, so teams operate from a single data source.
All three agents run on predictive models trained on $6 billion in historical UA spend across hundreds of games. Kohort claims 95% accuracy on daily campaign-specific predictions, with client-specific models training in under 20 minutes after connecting to an existing data warehouse.
Why UA Is an Agent Problem
“We are building agents, not just a Claude wrapper,” CEO Dan Marcus told EU-Startups. “The best UA teams operate more like high-frequency traders than marketers, and they need agents that act on real context, not vague signals.”
The high-frequency trading comparison is apt. Mobile game UA involves continuous bid adjustments across multiple ad networks, each with its own algorithm that reacts differently to signal changes. There is also a persistent tension between optimizing for short-term ROAS and finding high-lifetime-value users (the “whales” who sustain free-to-play economies). Most current tools optimize naively without accounting for these dynamics, according to Kohort.
The Raine Group’s John Salter, who co-founded the firm and led the investment, pointed to expansion beyond pure optimization. “Kohort has a clear pathway to become a category-defining platform in UA Agents,” Salter said in the company’s press release via FinSMEs, noting “the potential for a new product around user acquisition financing.”
A Vertical Agent Play in an Enterprise-Heavy Market
Most agentic AI coverage focuses on horizontal enterprise use cases: code generation, customer support, workflow automation. Kohort represents a different pattern: a vertical agent built for a specific, expensive operational problem in a $100 billion mobile gaming market.
The company’s bet is that domain-specific prediction models, not general-purpose LLMs, provide the context agents need to make decisions that hold up in production. Whether that thesis scales beyond gaming UA into broader consumer acquisition remains the open question as Kohort deploys the new capital.