2 min read

Meridian: Google’s latest Marketing Mix Model (MMM)

After the successful launch of Light Weight MMM Model, Google launched its latest advanced MMM model called Meridian. It is a pioneering open-source MMM that helps tackle critical challenges in measurement. With Meridian, you gain complete control over your marketing data and insights. This empowers you to make informed decisions within your organization, leading to smarter choices and improved outcomes.

Just in case, What is MMM ? Marketing mix modeling (MMM) applies statistical analysis to gauge the impact of marketing initiatives on revenue and other performance metrics. It aids in budget allocation and optimization. MMM utilizes aggregate data to assess the effectiveness of various marketing channels while considering external factors that influence outcomes. Importantly, MMM respects user privacy and does not utilize cookie or individual-level data.
Meridian, an MMM (Marketing Mix Modeling) framework, empowers advertisers to establish and operate their own in-house models. With Meridian, advertisers can gain insights into crucial marketing performance questions, such as:
  • The contribution of various marketing channels to revenue or other key performance indicators (KPIs)
  • The return on investment (ROI) achieved through marketing efforts
  • Informed optimization of future marketing budget allocation

Meridian: Beyond the Basics

Meridian is a full-fledged customizable MMM framework that is built on Bayeseian inference and therefore goes far beyond basic capabilities.  It offers:

  • Geo-Level Precision: Analyze marketing effectiveness down to local or regional levels, uncovering granular insights that drive targeted strategies.
  • Massive Data Handling: Meridian effortlessly tackles large datasets, providing scalability that matches the complexity of your marketing initiatives.
  • Prior Knowledge Integration: Leverage your unique business understanding by incorporating past experiments, industry expertise, or even gut feelings to refine your model.
  • Sophisticated Modeling: Meridian accounts for saturation effects, lagged impacts, reach and frequency data, and lower funnel channel modeling, giving you a comprehensive and accurate picture of your marketing performance.
Just in Case, What is Bayesian method? Meridian uses a Bayesian regression model, which combines prior knowledge with signals learned from data to estimate media effects and quantify uncertainty. Prior knowledge is incorporated into the model using prior distributions, which can be informed by experiment data, industry experience, or previous media mix models.

What data is needed to get started?

Meridian requires three main data arrays as model inputs (KPI, controls, and media). For channels with reach and frequency data available by geo and time period, the reach and frequency data can optionally be used instead of a single media metric. A unit value is required for KPI and media data so that units can be converted to a currency value for ROI calculations.

Take the Next Step

Currently Meridian is available as Apply Only and would likely remain for the next 4-6 months.  If you want to read more about it, please visit the public announcement on Google official blog here. I will share information as soon as this gets available beyond Apply Only.

Disclaimer: This article reflects the personal views and opinions of the author and does not constitute, nor should it be construed as, the official position or policy of my Employer. My Employer has not reviewed, or endorsed the content of this article.