Data Book
The prepared dataset used to train the model.
Model Results
Channel contributions, response curves, and diagnostics.
Set up MMM
Bucket, IAM, labels, and artifact paths.
Media Mix Modeling
Media Mix Modeling
Channel contributions and response curves from a trained Meridian model.
Media Mix Modeling (MMM) estimates how marketing channels contribute to outcomes, accounting for seasonality, baseline, and diminishing returns.
Your Python pipeline trains Google Meridian and writes results to GCS. Interact displays those artifacts — it does not train the model.
You can configure multiple labels (for example one per market). Each label has its own GCS folders and, optionally, its own Data Book table.