Social Rocket
Google is further expanding its measurement and first-party data infrastructure. The company is integrating Data Manager more deeply into Google Analytics and Display & Video 360, making the Data Manager API more universally usable, and adding automatic diagnostics that are meant to detect data problems before they affect campaign performance. In parallel, Google's open-source marketing mix model Meridian receives new agentic functions for data checking and model building.
For marketing teams, this is more relevant than an ordinary Google Ads feature update: Google is trying to bring data activation, measurement and AI-supported campaign control closer together in the same workflow.
Data Manager becomes the central layer for first-party data
Google positions Data Manager as the central place to connect data from websites, apps, physical stores or CRM systems and make it available for measurement and campaign activation.
A new element is that Data Manager is integrated directly into Google Analytics and Display & Video 360. The API is based on the IAB Tech Lab's Event and Conversion API standard and is meant to serve as a unified connector for audience and measurement data across various Google advertising products.
Google is also expanding direct integrations, including Mailchimp, ActiveCampaign, Klaviyo and Google Drive. This further blurs the boundary between CRM, commerce and adtech data.
For teams that already work with multiple data and measurement providers, this fits into a broader consolidation trend. The recent acquisitions involving Nielsen, DoubleVerify and LiveRamp also show how closely data, measurement and activation are moving together.
Automatic diagnostics are meant to catch data errors earlier
Data Manager receives built-in Diagnostics that automatically detect and address problems in data connections. The practical benefit lies mainly in making faulty or incomplete signals visible earlier, before they influence bidding strategies, audiences or reporting.
Google is also introducing the Data Strength Uplift metric in Google Ads. It is meant to show how many additional conversions are captured or recovered through the respective first-party data setup.
Performance figures published by Google should be read as vendor claims. Google reports, for example, that advertisers who connect offline and app data in Data Manager see a higher incremental ROAS on average. Such values are not a general promise for individual accounts and depend heavily on setup, campaign type and data quality.
Meridian gets agentic help for marketing mix models
Meridian, Google's open-source marketing mix model, is also being expanded. New agentic functions are meant to check data quality, identify modelling errors and support users while they build models.
In addition, brand signals such as brand search volume can be included in models. This is intended to better reflect how upper-funnel investments, such as video or TV, later affect demand and revenue.
With Meridian GeoX, Google's library for geographic causal experiments is now generally available. Such tests can be used to measure incremental effects beyond classic platform attribution and then feed them into marketing mix models.
This is especially relevant because many marketing teams use more AI but still struggle to demonstrate the actual business effect cleanly. This same gap is also reflected in our overview of AI use and impact measurement in marketing.
What changes in practice for advertisers
The strategic trend is clear: first-party data is becoming the foundation for AI-supported campaign control and, at the same time, for its measurement. Google is therefore no longer treating data connections, diagnostics, conversion measurement and causal models as separate tools.
For larger teams, however, this also means that data quality and governance become more important. If the same signals are used for audience building, bidding, reporting and AI agents, faulty data can affect several areas at once.
The most sensible approach is therefore not "more data at any price", but a reliable data pipeline with clearly defined sources, consent logic and traceable measurement goals. As agents increasingly take over operational media tasks, such controls become even more important; our article on AI agents in media buying and new cost and audit processes addresses this.