Overview

Developed by Meta Marketing Science, Robyn combines media-effect estimation, model calibration, comparison and budget allocation. It is designed for granular datasets with many independent variables, making it a better fit for established measurement teams than advertisers with sparse data or little modeling experience.

Robyn’s privacy design avoids dependence on personally identifiable information, individual-level log data, cookies or pixel data. That makes it relevant to teams working with aggregated advertising data, although the package still demands disciplined input preparation: paid media variables and spend vectors must have the same length and media order.

Key features

Robyn estimates media-channel efficiency and effectiveness, adstock rates and saturation curves. Geometric, Weibull CDF and Weibull PDF transformations give analysts several ways to represent how advertising effects persist. Ridge regression regularizes multicollinearity and helps limit overfitting, while Nevergrad evolutionary algorithms automate hyperparameter optimization. These methods are useful for complex media mixes, but automation does not eliminate the need to interpret results carefully.

For time-series structure, Robyn uses Prophet to decompose trend, seasonality and holiday patterns. It can calibrate models against ground-truth approaches including geo-based tests, Facebook Lift and MTA, and supports calibration using causal experiments such as randomized controlled and geo experiments. This gives teams a way to connect modeled results with observed evidence rather than relying on model output alone.

A gradient-based constrained nonlinear solver reallocates budgets to maximize outcomes, supporting budget planning and scenario analysis. Robyn also generates model one-pagers to make comparisons more intuitive. Those features help measurement teams move from estimation toward planning, though the package remains a technical modeling tool rather than a turnkey advertising dashboard.

Pricing

Robyn is free and open source under the MIT license, with a free plan and no paid tier described. That makes it a practical option for teams able to provide their own data and technical resources without a software subscription. The trade-off is that support is community-oriented: users can turn to the public Robyn MMM Users Facebook Group and GitHub issues.

Platforms

Robyn is available across Linux, macOS and Windows, with API and self-hosted options. The stable R version is on CRAN, while a development version is on GitHub. A Python version is documented as a beta rewrite translated from the R package; it may have bugs or translation issues and requires the Robyn R package to be installed first. Teams standardizing on Python should weigh that dependency before adopting it.

Who it's for

Robyn suits digital and direct-response advertisers with rich, granular data and analysts who want control over modeling, calibration and budget scenarios. It is especially attractive when privacy constraints rule out individual-level tracking data. It is a weaker fit for teams seeking a polished, low-effort commercial platform, or those without the capacity to prepare aligned media inputs and interpret statistical models.

Pros and cons

Pros: Free, MIT-licensed software makes a substantial MMM workflow accessible without subscription costs.
Pros: Adstock choices, automated optimization, experiment-based calibration and budget allocation cover several stages of measurement and planning.
Pros: It does not depend on PII, individual-level logs, cookies or pixel data, which suits privacy-conscious measurement programs.
Cons: The package is aimed at granular, data-rich use cases, limiting its fit for advertisers with sparse inputs.
Cons: Python remains beta, can encounter bugs and depends on the R package, so it is not a clean Python-only deployment.
Cons: Media variables and spend vectors must align in length and order, adding a concrete data-preparation burden.

Alternatives

Marketing Performance Management Software is a broader category to explore when the need extends beyond marketing mix modeling.

Google Meridian is another free, open-source option; choose it when source-code transparency and user auditability supported by methodology papers are priorities. Fintastic offers a paid enterprise subscription with unlimited users and data, which may suit teams seeking a commercial service. LiftLab is a paid enterprise marketing measurement platform with subscription terms set in an Order Form. Measured is positioned for enterprise brands. Planful tailors its subscription to company size, user count and selected modules.

Etropo Budget Buddy is a paid budgeting option with monthly Essential and Pro plans; consider it when team-member and line-item limits are a better fit than a modeling package. Keen is another paid option. Aprimo offers paid packages that include digital asset management capabilities.

Verdict

Choose Robyn if your team has rich advertising data, can work in R and wants a free, privacy-conscious route from media modeling through calibration to budget planning. Its breadth and open license are strong reasons to adopt it; teams that need a stable Python-native workflow or a turnkey commercial platform should look elsewhere.

	rex by AureliaX	Robyn
Free plan	No	No
Free trial	No	No
Paid from	—	—
Open source	No	No
Platforms	api, Web	api, Linux, macOS, self-hosted, Windows
Budget planning	Yes	Yes
Forecasting	Yes	Yes
Scenario planning	Yes	Yes
ROI reporting	Yes	Yes
Free plan	—	Yes

rex by AureliaX vs Robyn

rex by AureliaX

Robyn