PyMC-Marketing

PyMC-Marketing is a free, open-source Python library for Bayesian marketing analytics, including marketing mix modeling (MMM) and customer lifetime value models. Its MMM API supports custom priors and likelihoods, adstock and saturation transformations, time-varying effects, diagnostics, and predictions beyond the observed data. Models can be calibrated with empirical experiments such as lift tests, and built-in budget optimization allocates spend across channels to maximize ROI. Bayesian uncertainty estimates include posterior distributions and credible intervals. The library also includes customer choice, Bass diffusion, and alpha-stage predicted incrementality by experimentation models, alongside predictive CLV, cohort analysis, and customer segmentation. It is licensed under Apache 2.0 and can be installed with pip; a conda environment is not required. Current documentation requires Python 3.12 or later, PyMC 6.0 or later, and ArviZ 1.2 or later. Inference options include the BlackJax, NumPyro, and Nutpie NUTS samplers, and PyMC backends allow GPU acceleration. PyMC Labs and PyMC core developers maintain the library. Learning materials include a guided path, example notebooks, and an API reference; a free 30-minute strategy session is also offered.

PyMC-Marketing provides a broad set of Bayesian marketing models and MMM tools in a free Python library. Check the documented software requirements before adopting it.

PyMC-Marketing suits analysts and teams using Python for Bayesian marketing models, budget allocation, and customer lifetime value work. It may fit users comfortable with the listed Python and library requirements.

PyMC-Marketing is a free, open-source Python library for Bayesian marketing analytics. It is best suited to analysts building their own modeling workflows; its breadth and uncertainty estimates are valuable, but the Python and dependency requirements make it a poor fit for teams seeking a ready-made dashboard.

Overview

The library covers marketing mix modeling (MMM), customer lifetime value, customer choice and Bass diffusion, alongside an alpha-stage model for predicted incrementality by experimentation. Bayesian posterior distributions and credible intervals put uncertainty around predictions, which can help teams weigh decisions beyond a single estimate. That flexibility comes with the responsibility of building and evaluating a technical workflow.

PyMC-Marketing is maintained by PyMC Labs and PyMC core developers. PyMC Labs describes itself as a Bayesian AI consultancy founded by members of the PyMC community.

Key features

• Adaptable MMM: Custom priors and likelihoods, adstock and saturation transformations, time-varying effects, diagnostics and out-of-sample predictions support analyses tailored to a team's questions. This is a strength for experienced modelers, but requires informed choices rather than a fixed, turnkey setup.
• Budget optimization: Built-in optimization allocates spend across channels to maximize ROI, linking MMM work to budget decisions.
• Experiment calibration: Models can be calibrated with empirical experiments, including lift tests, giving teams a way to incorporate experimental evidence.
• Multiple inference options: The MMM API supports NUTS samplers including BlackJax, NumPyro and Nutpie; PyMC backends also allow GPU acceleration. These options suit technical teams with differing inference needs, but do not remove the need to manage the Python environment.
• Learning and support: A guided learning path, example notebooks and API reference provide a route into the library. PyMC Labs also offers a free 30-minute strategy session and a contact option.

Pricing

PyMC-Marketing costs free and is open source under the Apache 2.0 license. There is no paid tier to unlock additional library features, making it attractive for teams able to supply their own technical expertise. The trade-off is that this is a Python library, not a packaged marketing service.

Platforms

PyMC-Marketing supports Linux and self-hosted use. Installation is available through pip install pymc-marketing, and the documentation says a conda environment is not required. Current documentation requires Python 3.12 or later, PyMC 6.0 or later and ArviZ 1.2 or later; teams should check these requirements against their environment before adopting it.

Who it's for

This is a strong fit for analysts and marketing science teams comfortable with Python and Bayesian modeling who want adaptable MMM, customer lifetime value analysis or related models without a software license fee. It is less suitable for marketers looking for a graphical, nontechnical workflow or teams unable to meet the current software requirements.

Pros and cons

• Pro: Broad model coverage includes MMM, customer lifetime value, customer choice and Bass diffusion, so teams can use one library across several marketing analytics needs.
• Pro: Bayesian uncertainty estimates and experiment calibration support more nuanced decision-making than a single forecast alone.
• Pro: The Apache 2.0 license and free price avoid a library subscription charge.
• Con: Python and specific dependency versions are required, narrowing its fit for teams without a compatible technical environment.
• Con: The alpha-stage status of predicted incrementality by experimentation makes that model a less mature part of the toolkit.

Alternatives

Google Meridian is another free, open-source option for teams that value source-code and methodology transparency and want its broader listed platform support, including web, macOS and API alongside Linux and self-hosting.

Robyn is a free alternative for teams comparing open marketing measurement tools across Linux, macOS, Windows, API and self-hosted platforms.

Revup Marketing may suit teams seeking a web-based service with a free plan for one plan, one user and one team; its higher tiers expand those limits to five users and one team, or ten users and five teams, with pricing not stated.

Brandsystems is a web-based, paid marketing resource management option for teams seeking that category rather than a Python modeling library.

Fintastic offers a paid enterprise subscription with unlimited users and data and support included; it may suit teams seeking those terms rather than a free self-hosted library.

PYXIS Planning is a paid web-based alternative.

Revlo Marketing is another paid alternative.

LiftLab is a paid enterprise marketing measurement platform; teams seeking a demo-led service can contact it for pricing, with subscription terms and fees set in an Order Form.

For broader category comparisons, see Marketing Performance Management Software and Customer Lifetime Value Software.

Verdict

Choose PyMC-Marketing if your team can work in Python and wants flexible Bayesian marketing models, experiment calibration and uncertainty-aware predictions without a license cost. Look elsewhere if you need a ready-to-use graphical product or cannot meet its current Python and dependency requirements.