Attribution & marketing mix modeling · Meta
Robyn
Open-source semi-automated marketing mix modeling package from Meta, built in R for data scientists to run themselves.
Robyn is an open-source R package, originally built inside Meta, for semi-automated marketing mix modeling: it fits a model relating historical spend across channels to a business outcome (revenue, sign-ups), using ridge regression and evolutionary (genetic) algorithm-driven hyperparameter search to find plausible saturation and adstock curves, then outputs budget-allocation recommendations. It is a code library, not a hosted product — there is no dashboard, account or support line. Running it in production requires someone fluent in R (or the community Python port), several years of weekly spend and outcome history, and ongoing judgment to validate outputs against business knowledge or holdout experiments, which is exactly the burden that paid MMM vendors like Recast exist to remove.
At a glance
| Vendor | Meta |
|---|---|
| Pricing model | Open source + paid options |
| Free tier | Yes |
| Deployment | Self-hosted |
| Open source | Yes (MIT) |
| Best for | Teams with an in-house data scientist and multi-year weekly spend history who want full control over their MMM rather than a managed SaaS. |
Pricing
Free, open-source R package; the only cost is the compute and the data scientist's time to run and maintain it.
Pricing has not been verified yet — see the vendor's site.
Features
- Ridge regression with adstock and saturation transformations
- Genetic-algorithm-driven hyperparameter search
- Budget allocation optimizer output
- Support for multiple media and non-media variables
- Model diagnostics and one-pager output plots
- Community Python port available
Profile last reviewed September 21, 2026