As MMM Rides Again, Google Finds Its Place In The Conversation With Meridian

Meridian is the name of Google's open-source MMM (marketing mix model) that was launched last month. It is designed to attribute ad spend channel by channel over time and is becoming increasingly important due to changes in the privacy and regulatory environment.

MMMs (marketing mix models) are becoming more important for advertisers because of the growing signal loss in user-level data and the changes in the privacy and regulatory environment. As IP address identity data becomes less reliable, MMMs offer a way to measure the effectiveness of ad campaigns without relying on device IDs or user-level information, making them suitable for the current privacy-focused landscape.

Meridian differs from Meta's Robyn in terms of accessibility as it uses Python, a coding language that is familiar to both marketers and developers. In contrast, Robyn uses the R language, which is more commonly used by statisticians and data scientists. This makes Meridian technically more accessible to a broader range of users.

Google's Meridian MMM model includes unique features such as the ability to gather reach and frequency data from YouTube and incorporate Google search data. These features provide exclusive ways to attribute credit to certain channels, including Google's own, and are considered indicators of organic interest in a product or business.

When implementing open-source MMM models like Meridian, advertisers face challenges such as the need for data analysts and rigorous oversight. Despite being open-source, these models require expert setup and maintenance to function effectively. Additionally, many external factors must be manually incorporated by the advertiser, such as seasonal shopping patterns, price fluctuations, competitor pricing, and macroeconomic factors, to avoid skewed outputs.