Getting found · Independent · 24 Oct 2026

google/meridian

Meridian is an MMM framework that enables advertisers to set up and run their own in-house models.

← Getting found

In simple English

This tool helps you figure out which advertising channels actually bring in money. You feed it your past spending data and sales numbers. It runs math models to show your return on investment. Without using tracking cookies or personal user data, it keeps privacy safe. It is free to download, but you need a strong computer with a graphics card to run the calculations.

What it is

Meridian is an open-source marketing mix modeling framework designed by Google. It uses statistical calculations called Bayesian causal inference to estimate how different ad campaigns affect sales.

What it is good for

Solo founders and marketers who want to build their own in-house analytics models instead of paying expensive agencies for media planning.

Stars1.6k
LicenceApache-2.0
LatestNo tagged releasechecked 24 Oct 2026
Last code change2026-10-24
Written inPython
Setup effort
4 / 5
Payoff
4 / 5
Good
  • Built and maintained by Google with detailed documentation and tutorial notebooks.
  • Processes large amounts of regional sales data without relying on user tracking cookies.
  • Provides built-in tools for planning future advertising budgets and scenario testing.
Watch for
  • Requires Python 3.11 to 3.13 and a dedicated graphics card with at least 16 GB of memory for reasonable speeds.
  • Heavy mathematical calculations mean training models can take a significant amount of computer time.
  • Code contributions from outside developers are rarely merged because the project links to internal Google systems.
Replaces

expensive marketing agency reports and manual spreadsheet calculations for ad spend

Works well with

Google Colab for running test notebooks and interactive budget planning

Similar repositories
  • facebookexperimental/Robyn1.5k stars

    An open-source marketing mix modeling package by Meta that also uses evolutionary algorithms for budget optimization.

  • pymc-labs/pymc-marketing1.3k stars

    A Bayesian marketing and customer analytics library built on top of PyMC for custom attribution models.

Also reviewed on Repo Radar
  • GoogleChrome/lighthouse-ci7.1k stars

    Runs the standard page audit on every code change and fails the change if the page got slower.

  • harlan-zw/nuxt-seo1.5k stars

    Seven search-engine and answer-engine modules for Nuxt sites, installed as one package.

  • Swetrix/swetrix1.2k stars

    Website analytics with no cookies, plus performance monitoring and error tracking, which you can host yourself.

Install
python -m venv venv
source venv/bin/activate
pip install --upgrade google-meridian

Independent: reviewed for Repo Radar outside the daily editions. Every independent repository

On Stack Radar: Marketing & Growth › Conversion & Growth Analytics › Attribution.

Get the next edition in your inbox

A dozen repositories, opened and checked. The licence read, the last release dated, and the ones that did not make it named with the reason. It is the half most lists leave out.

No tracking pixels. One click to leave. The archive stays free either way.

We use your address to send the edition and nothing else. Confirm by email, leave in one click. How we handle it.