Agent skill · marketing growth · aradotso
pymc-marketing-mmm-clv
Bayesian marketing analytics with PyMC for Media Mix Modeling (MMM), Customer Lifetime Value (CLV), and BTYD models
Why this skill is useful
Provides domain-specific Bayesian modeling techniques for marketing analytics that the AI wouldn't reliably generate on its own.
What it needs
Requires conda, docker, pymc-marketing installed locally. About 6k tokens when loaded. Last updated 2026-08-05. 8 stars on the source repository.
What this skill does
PyMC-Marketing: Bayesian Marketing Analytics Skill by ara.so — Marketing Skills collection. PyMC-Marketing is a Bayesian marketing analytics toolbox built on PyMC. It provides production-ready implementations for Media Mix Modeling (MMM), Customer Lifetime Value (CLV), and Buy-Till-You-Die (BTYD) models with full probabilistic inference capabilities. Installation Basic Installation Docker Installation The project provides Docker support for Jupyter-based workflows: Core Capabilities 1. Media Mix Modeling (MMM) MMM helps quantify the impact of marketing channels on business outcomes with: Adstock transformations: Geometric, delayed, Weibull Saturation effects: Logistic, Michaelis-Menten, Tanh Time-varying effects: Intercept and media contribution dynamics Budget optimization: ROI-maximizing allocation across channels Experiment calibration: Lift test integration 2. Customer Lifetime Value (CLV) Probabilistic models for customer value prediction: Beta-Geometric/NBD: For non-contractual settings Pareto/NBD: Customer dropout modeling Gamma-Gamma: Monetary value modeling 3. Customer Choice Analysis (CSA) Discrete choice modeling for understanding customer preferences. Media Mix Modeling (MMM) Basic MMM Setup Available Adstock Functions Available Saturation Functions Time-Varying Effects Model Diagnostics and Visualization Budget Optimization ROAS and Channel Efficiency Lift Test Calibration Out-of-Sample Prediction Alternative NUTS Samplers Customer Lifetime Value (CLV) Beta-Geometric/NBD Model Pareto/NBD Model Gamma-Gamma Model (Monetary Value) Model Saving and Loading Configuration Patterns Custom Priors Sampler Configuration Troubleshooting Convergence Issues Long Fitting Times Memory Issues Data Scaling Issues Common Workflows End-to-End MMM Analysis CLV Prediction Pipeline Environment Variables Additional Resources Documentation: https://www.pymc-marketing.io/ Examples: https://www.pymc-marketing.io/en/stable/notebooks/ Discourse: https://discourse.pymc.io/ GitHub D …
How to use it
Reference it in AdaL, Claude Code, Cursor or any coding agent — nothing to install:
@skills aradotso/pymc-marketing-mmm-clv