Google Meridian Gains AI Chatbot and Upper-Funnel Signal Ingestion as Meta Retreats from MMM
The Development
Google has shipped three material updates to its open-source Meridian marketing mix modelling tool. A natural-language chatbot interface now lets marketers query Meridian for model-building guidance, audit data quality, and troubleshoot configuration errors. GeoX, Meridian's geotesting module for segmenting campaigns by geography and feeding regional lift findings back into models, has exited beta and is now generally available. Most significantly, Meridian can now ingest upper-funnel signals — including Branded Google Query Volume — alongside lower-funnel conversion data, enabling attribution of hard-to-measure channels like TV and DOOH. Separately, Google is integrating its Data Manager into Google Analytics and DV360, and adopting IAB Tech Lab's ECAPI standard for broader interoperability. These updates arrive as multiple agency and measurement vendor sources confirm Meta is no longer actively pushing its competing Robyn open-source MMM.
Our Take
Google is doing something structurally clever here. By lowering Meridian's technical floor with a chatbot interface while raising its analytical ceiling with upper-funnel signal ingestion, it is making the tool attractive to both sophisticated measurement teams and mid-market advertisers who previously couldn't resource a credible MMM practice. The timing relative to Meta's Robyn retreat is not coincidental. Google is expanding its measurement surface area — query volume, brand signals, geo-lift — all of which feed proprietary data back into an ecosystem Google controls. The open-source framing creates the optics of neutrality, but the data dependencies run deep. Measurement teams should go in clear-eyed about what they are trading.
What Changed
Meridian can now bridge upper- and lower-funnel signal ingestion within a single open-source MMM framework, while its AI chatbot lowers the technical barrier to model setup and auditing. This is the first widely accessible MMM tool to combine conversational AI guidance with brand-signal-aware attribution in a no-licence model.
Marketing Impact
Marketing science and measurement functions gain the most immediate leverage. Upper-funnel signal ingestion directly addresses the attribution gap for TV and DOOH investment, while the chatbot reduces the analyst hours required to validate model integrity — compressing the MMM iteration cycle from weeks to days.
Competitive Implication
In-house measurement teams and independent MMM vendors operating Robyn-based practices are now exposed. Google's accelerating Meridian roadmap, combined with Meta's withdrawal from the open-source MMM race, creates a gravitational pull toward Google's stack that smaller vendors will struggle to counter without equivalent upper-funnel data assets.
Strategic Outlook
Expect Meridian adoption to accelerate through Q4 2026 as agencies and brands that were sitting out open-source MMM re-evaluate their tooling. Independent measurement vendors will be forced to either differentiate on data neutrality or integrate Meridian as a foundation. Google's next logical move is tighter Meridian integration with Performance Max reporting.
The Exploit
Action Item
Marketing science leads running TV or DOOH budgets above $5M should activate Meridian's Branded Query Volume ingestion this quarter and rerun existing models — the delta between old and new attribution outputs is the case for reallocation that CFOs will actually approve.