Skip to content

Measurement (MMM & MTA)

The platform measures marketing effectiveness with two complementary approaches: Marketing Mix Modeling (MMM), a top-down statistical view of how channel spend drives outcomes, and Multi-Touch Attribution (MTA), a bottom-up view of individual user journeys. Content about them spans methodology, dbt implementation, API access, and MCP agent tooling — this page is the map.

Methodology

  • MMM Methodology

    Long-form reference on MMM as a method: response curves, adstock, Bayesian estimation, validation checkpoints, and structural considerations.

    MMM Methodology

  • MTA Methodology

    Multi-touch attribution for on-chain user journeys: touchpoint definitions, attribution windows, and model comparison.

    MTA Methodology

Implementation

  • Measurement Stack (dbt)

    How the MTA + MMM pipeline is implemented in dbt-cerebro: sources, staging, feature tables, and output marts.

    Measurement Stack

Data & API

  • MMM Models

    The dbt model catalog for the marketing-mix module.

    models/mmm

  • MTA Models

    The dbt model catalog for the attribution module.

    models/mta

  • MMM API Endpoints

    REST endpoints serving MMM outputs.

    api/catalog/mmm

  • MTA API Endpoints

    REST endpoints serving MTA outputs.

    api/catalog/mta

Agent Tooling (MCP)

  • MMM Standard Operating Procedure

    The MCP server's MMM workflow: gates, phases, and outputs.

    MCP MMM

  • MMM User Guide

    Prompt recipes and end-to-end examples for running MMM through an AI assistant.

    MMM User Guide

  • Measurement Flow

    How MTA, MMM, and the unified measurement view fit together in agent workflows.

    Measurement Flow