InMarket launches Maestrø beta for AI media optimisation

InMarket has launched Maestrø in beta, describing it as an agentic media optimisation platform designed to tie campaign decisions to real-world consumer outcomes.

The product targets brands and agencies increasing their use of artificial intelligence in media buying and campaign management. Maestrø acts as an orchestration layer, feeding measured results back into campaigns while they are still running rather than after they end.

At the centre of the launch is the argument that AI-driven media decisions should be based on observed outcomes rather than platform-specific signals or modelled proxies. The platform uses a decade of store visit and purchase data from more than 200 million opted-in consumers and returns sub-weekly, user-level incrementality data during a campaign.

That means AI agents, or human teams using the same data, can adjust audiences, creative choices and media allocation while a campaign is live. Those actions are linked to measured store visits and item-level purchases associated with ad exposure.

Three models

Maestrø is being offered in three forms. The first, Outcomes, is a managed service in which InMarket runs audiences, creative and optimisation against a guaranteed incremental return on ad spend, or iROAS.

The second, Optimisation, provides AI-generated recommendations that still require approval by human users. The third, Open Access, is built on Model Context Protocol and is intended to stream measurement data into AI assistants, business intelligence platforms and custom tools.

The launch reflects a broader shift in advertising technology as marketers try to make faster decisions with growing use of automated systems. A key issue in that transition is whether AI tools can show that media changes are producing incremental commercial results, rather than simply reacting to signals generated within individual platforms.

InMarket presents Maestrø as a response to that problem. Its existing measurement products have long been used to show whether campaigns drove visits, spending and purchases, but the new system changes the timing by returning those answers during the campaign window.

Todd Dipaola, Founder & CEO at InMarket, set out the company's argument for the new product.

"Brands rightfully believe that AI can drive growth from smarter media buying. To do so, the system must be rooted in truth and be as fast as the changes," said Todd Dipaola, Founder & CEO at InMarket. "We built Maestrø to solve that challenge as a universal accountability layer for AI-powered media - an orchestrator that ensures every agent and human change is backed by real-world, real-time causation, not last quarter's MMM model."

Measurement focus

The pitch centres on causal measurement rather than correlation. InMarket said each action in the system is linked to observed outcomes at user level, and that the data can be returned quickly enough to support mid-campaign decisions.

That matters because many advertising measurement systems still rely on post-campaign analysis, media mix modelling or platform-owned data sets. Those approaches can help identify patterns, but they may not give campaign teams a timely view of whether budget changes are generating additional sales or visits while the campaign is active.

By contrast, Maestrø is intended to close what InMarket describes as an actionability gap between measurement and execution. Instead of leaving outcome analysis until after a campaign has finished, the system feeds those signals back to the tools where teams and AI agents are already working.

The emphasis on cross-platform measurement also reflects advertiser concern about depending too heavily on metrics provided by media owners. As more buying and optimisation functions move into automated systems, scrutiny is likely to increase over the source of the data those systems use to make decisions.

Open Access may be the clearest sign of that direction. Using Model Context Protocol, Maestrø can make its measurement available outside its own interface, including within third-party AI assistants and analytics environments.

The goal is to connect media investment more directly to store visits and purchases instead of relying on proxy measures. InMarket describes the platform as accountable to outcomes rather than media ownership, with sub-weekly incremental data intended to support campaign changes before spending is complete.

InMarket said Maestrø returns "sub-weekly, user-level incrementality in-flight so AI agents can act on real outcomes in real-time during the campaign rather than after the fact."