{"categories":[{"label":"Scientific/Engineering","url":"https://skillfed.io/packages/category/scientific-engineering/6"},{"label":"Mathematics","url":"https://skillfed.io/packages/category/scientific-engineering-mathematics/3"}],"enrichment":{"capability":"PyMC-Marketing provides Bayesian statistical models for marketing analytics, including Marketing Mix Modeling (MMM), Customer Lifetime Value (CLV), Customer Choice, Bass Diffusion, and Predicted Incrementality by Experimentation (PIE).","skillfed_tags":["bayesian-inference","marketing-analytics","causal-modeling"],"use_cases":["Quantify the incremental impact of each marketing channel on sales or conversions using Bayesian MMM with carryover and saturation effects.","Estimate customer lifetime value distributions to inform acquisition budgets and retention strategies.","Analyze discrete choice behavior (e.g., MaxDiff surveys, product preferences) using Bayesian hierarchical choice models.","Test incrementality of marketing experiments with PIE (Predicted Incrementality by Experimentation) framework.","Model product adoption and diffusion curves using Bass Diffusion models for new product forecasting.","Incorporate causal domain knowledge via DAGs to identify confounders and improve model validity."],"what_it_does":"PyMC-Marketing is a Bayesian statistical modeling library for marketing analytics built on top of PyMC, ArviZ, and related tools. It provides domain-specific implementations of Marketing Mix Models, Customer Lifetime Value estimation, discrete choice models, and experimentation frameworks that allow marketers and analysts to quantify the causal impact of marketing channels and customer behaviors.\n\nThe package is designed for practitioners who want to move beyond frequentist attribution methods and leverage Bayesian inference to incorporate prior knowledge, handle uncertainty explicitly, and make probabilistic predictions. It includes features like adstock transformations to model carryover effects, saturation functions to capture diminishing returns, time-varying intercepts and media efficiency, and causal identification tools via directed acyclic graphs. Models can be customized with domain-specific priors and likelihoods, and inference can be run on CPUs or GPUs via PyMC's backend options.","worth_installing":"Yes, if you need Bayesian marketing analytics and are comfortable with the PyMC ecosystem. The package is actively maintained, has no known vulnerabilities, uses a permissive license, and offers specialized tools for MMM and CLV that are difficult to assemble from scratch. Requires Python >= 3.12 and a substantial dependency stack; best suited for teams with statistical modeling experience or access to consulting support."},"id":"pymc-marketing","links":{"html":"https://skillfed.io/packages/pymc-marketing","md":"https://skillfed.io/packages/pymc-marketing.md","pypi":"https://pypi.org/project/pymc-marketing/"},"maintenance":{"status":"active"},"meta":{"latest_release":"2026-08-07","license_spdx":null,"license_treatment":"permissive","name":"pymc-marketing","python_support":"supports_current","summary":"Marketing Statistical Models in PyMC"},"popularity":{"monthly_downloads":184703,"position":10026,"tier":"top_15000"},"security":{"n_vulnerabilities":0},"version":"1.0.0"}