Here's what Meridian is, what the integration changes, and what it means whether or not your company pays for GA360.

First, What Is Meridian?

Meridian is Google's open-source marketing mix modeling (MMM) framework. Like any MMM, it uses aggregated historical data, such as weekly spend by channel and weekly sales, to estimate how much each channel actually contributes to business results. It needs no cookies or user-level tracking, and it can include offline media like TV and radio alongside digital channels.

What sets Meridian apart is its Bayesian approach. Rather than producing a single "best" answer, it produces a range of plausible results, and it lets you feed in prior knowledge, such as results from past lift tests, to make the estimates more reliable. It also supports geo-level modeling, using regional data rather than one national total. That gives the model far more variation to learn from, which usually means sharper results.

Meridian has developed quickly:

What the GA360 Integration Does

According to Google, Meridian in GA360 is designed to help advertisers do three things:

  1. Unify data in one place. Bring first-party and cross-channel data together, so the model can draw on what GA360 already holds instead of requiring a separate data pipeline.
  2. Measure causal performance. Identify which marketing activities actually drive business results, not just which ones appear in the customer journey.
  3. Plan with predictive scenarios. Test budget allocations, such as shifting spend from one channel to another, before committing real money.

The biggest practical change is where MMM lives. Until now, MMM has usually been a specialist project. A data science team or consultancy gathered data, built a model in Python or R, and delivered results in a slide deck once or twice a year. Putting it inside GA360 places MMM next to the reports, audiences and attribution data analytics teams already use every day.

Google has also signaled where this may go next. At the same event, it introduced Qualified Future Conversions (QFCs) in Google Ads, a Gemini-powered signal that links today's ad spend to future sales. Google has said it eventually plans to bring QFC signals into Meridian to refine model accuracy, though no timeline has been given.

Why It Matters for Analytics Teams

MMM becomes part of the analytics job. Digital analysts have mostly worked with user-level data: sessions, events and conversions. With MMM inside GA360, they'll increasingly be expected to understand adstock, saturation, incrementality and model diagnostics. That's a real skills shift, and an opportunity for analysts who get ahead of it.

Attribution and MMM will sit side by side. GA360 users will see data-driven attribution and MMM results in the same platform, and those numbers will often disagree. Analysts will need to explain why. Attribution shows who touched a conversion; MMM estimates what caused it.

Data quality becomes even more important. An MMM is only as good as its inputs. Consistent conversion tracking, clean channel groupings, accurate cost data and reliable revenue figures all directly affect model results. Years of messy GA data become years of messy model inputs.

It raises the bar for neutrality questions. Some marketers will reasonably ask whether a Google-built model inside a Google product treats Google's own channels fairly. Meridian's open-source code helps here, since anyone can inspect exactly how it works. Still, good practice is to validate key results with independent incrementality experiments.

Data quality becomes a model problem, not just a reporting problem. If your GA4 revenue numbers, conversion events or channel groupings are wrong, those errors don't just distort a dashboard — they get baked into a model that's now driving budget decisions.

What It Costs

The open-source Meridian library is free. Anyone can download it and run it in Python, for example in a Google Colab notebook.

The GA360 integration is a different story. Google Analytics 360 is Google's paid enterprise tier, so the built-in Meridian experience comes with a GA360 contract. Google hasn't published a detailed rollout date, so teams interested in the integration should check timing with their Google account team.

If You Use Free GA4

You don't need GA360 to benefit from Meridian. A practical path looks like this:

Even if you never adopt GA360, the integration is worth watching. It's a strong sign that the industry now treats MMM as standard practice rather than a niche technique.

The Bottom Line

Meridian's move into GA360 marks a shift in what an analytics platform is for. The question is no longer only "where did our traffic come from?" but "which marketing actually grew the business, and where should the next dollar go?" Teams that build MMM skills and clean data foundations now will be ready, whichever version of Meridian they end up using.

Whether you're exporting GA4 data into open-source Meridian or waiting on the GA360 integration, the same prerequisite applies: the model can only be as accurate as the GA4 property underneath it. Audit your GA4 setup before you invest time building on top of it.

In the final post of this series, we'll walk through how to get started with MMM: the data you need, the tools to consider and your first steps.

Sources: PPC Land: Meridian lands inside Analytics 360 · B&T: Google to bring Meridian into Google Analytics 360 · AdNews: Google launches open-source MMM Meridian

Travis Gunn
Travis Gunn
Founder of GA4 Health Check. Working with Google Analytics since 2013, with over 250 clients audited across almost every industry vertical. 100% Job Success on Upwork for over a decade.