Simba documentation

Further Reading — Papers, Articles, and External Resources

A curated collection of resources for deeper learning about marketing mix modeling, Bayesian statistics, and the methodologies behind Simba.

Marketing Mix Modeling

Foundational Concepts

• Marketing Mix Modeling: A Practical Guide — Overview of how Bayesian approaches are transforming the MMM landscape (see PyMC-Marketing docs for worked examples)
• The Resurgence of Marketing Mix Modeling — Why MMM is experiencing a renaissance driven by privacy changes and cookie deprecation
• Lift Test Integration in PyMC-Marketing — How lift test observations are integrated as likelihood terms in Bayesian MMM

Industry Reports

• Google’s guide to MMM and Meridian — Google’s perspective on how marketing mix modeling fits into modern measurement
• Meta’s Robyn documentation — Meta’s open-source MMM approach (ridge regression, for comparison with Bayesian methods)

Bayesian Statistics

Introductory Resources

• “Bayesian Data Analysis” by Gelman et al. — The definitive textbook on Bayesian statistics (advanced but comprehensive)
• “Statistical Rethinking” by Richard McElreath — An accessible introduction to Bayesian modeling with practical examples
• “Think Bayes” by Allen Downey — A beginner-friendly introduction to Bayesian statistics using Python

PyMC Resources

• PyMC Documentation — Official documentation for the PyMC probabilistic programming library
• PyMC-Marketing Documentation — Documentation for the marketing-specific library that Simba is built on
• PyMC Examples Gallery — Worked examples of Bayesian modeling with PyMC

Media Effectiveness

Saturation and Diminishing Returns

• Understanding Media Saturation — How diminishing returns affect media investment decisions (see PyMC-Marketing saturation docs)

Adstock and Carryover

• Advertising Adstock — Broadbent (1979) — The foundational concept of how advertising effects persist and decay over time

Privacy and the Future of Measurement

• The deprecation of third-party cookies, iOS App Tracking Transparency (ATT), GDPR, and similar regulations are making user-level attribution increasingly unreliable. MMM’s reliance on aggregate time-series data makes it inherently privacy-compliant.

Open-Source MMM Tools

• PyMC-Marketing — The open-source Bayesian marketing analytics library (Simba’s foundation)
• Robyn by Meta — Meta’s open-source MMM solution (R-based, ridge regression approach)
• Meridian by Google — Google’s open-source Bayesian MMM tool

Conferences and Communities

• PyMCon — The PyMC community conference featuring talks on Bayesian statistics and applications
• Marketing Science Conference — INFORMS conference on quantitative marketing research
• Measure Camp — Unconference for analytics and measurement professionals

---

Have a resource suggestion? Let us know.

*See also: Glossary

PyMC-Marketing & Simba

Core Concepts*

This site is open source. Improve this page.