Griffin MMM is a practical, production-ready Bayesian Media Mix Modelling (MMM) platform designed specifically for small to medium marketing agencies. Built on top of the PyMC framework, Griffin MMM offers marketers unparalleled insights into campaign performance, enabling:
- Precise ROI Measurement: Understand the true value of each media channel using Bayesian inference.
- Optimized Budget Allocation: Simulate and discover the best ways to allocate your marketing spend for maximum returns.
- Reliable and Transparent Insights: Gain stable, interpretable results, even with sparse or incomplete data.
Ready to get started? Run the demo notebook in Google Colab or contact us at [email protected].
| Feature | Griffin MMM Highlights |
|---|---|
| Bayesian Modelling | Robust modelling techniques for replicable results. |
| Intuitive Visualizations | Transform data into actionable insights using standardised csv and png outputs. |
| Scalable and Flexible | Model unlimited channels with Pro access. |
| Built for Agencies | Simple configuration and setup, no advanced coding needed. |
Once your data is ready, navigate to the /demo folder, and open the demo notebook in Google Colab (look for the .ipynb file).
- Install dependencies with one click in the Colab environment.
- Use the preloaded sample data and demo configuration files (config.yaml) or load your own.
- Explore features such as budget optimization, channel contribution analysis, and predictive insights.
- Review results to make data-driven decisions.
- (Optional) Use PowerBI or similar for further visualisations as needed.
Need help? Contact us or raise a GitHub issue.
Griffin MMM offers significant advantages over traditional solutions, including:
Methodological Rigor:
- Grounded in Bayesian principles for accurate and stable outcomes.
- Includes state-of-the-art methods for handling adstock, saturation, and seasonality.
Ease of Use:
- Designed for agencies without extensive technical teams.
- Fully functional in Google Colab with minimal setup.
Alternative to Robyn:
- Offers more robust outputs and interpretable results compared to open-source solutions like Meta's Robyn.
Griffin MMM is currently in early release until 31 March 2025. During this period:
- Early adopters enjoy discounted subscription rates.
- Feedback from beta users helps shape the future of Griffin MMM.
| Feature | Demo Version | Pro Version |
|---|---|---|
| Price | Free | Annual subscription |
| Channel Limit | 4 | Unlimited |
| Features | Core features only | Advanced tools, priority support |
| Support | Community-based (GitHub, email) | Priority and dedicated support |
Contact [email protected] to upgrade.
Comprehensive setup instructions, API references, and use cases are included in the /docs folder.
- GitHub Issues: Raise problems or questions directly on the GitHub issue tracker.
- Email: Contact us at [email protected] for personalized assistance.
Griffin MMM operates under two key legal agreements:
-
Griffin Subscription Agreement: This document outlines the subscription terms for Pro users, including pricing, usage rights, and support provisions.
-
Griffin End-User License Agreement (EULA): This agreement governs the use of Griffin MMM for both demo and Pro versions. All users must review and accept the EULA before using the software.
-
Demo Version: Free to evaluate, with support for up to 4 media channels.
-
Pro Version: Includes full feature access, unlimited media channels, and priority support.
For detailed licensing terms, refer to the respective documents linked above or visit the /docs/LEGAL directory of the demo repository (https://github.com/griffin-analytics/griffin-mmm-demo/).