# facebookexperimental/Robyn

Robyn is an experimental, AI/ML-powered and open sourced Marketing Mix Modeling (MMM) package from Meta Marketing Science. Our mission is to democratise modeling knowledge, inspire the industry through innovation, reduce human bias in the modeling process & build a strong open source marketing science community.

Repository: https://github.com/facebookexperimental/Robyn
Canonical: https://ross.abutalabs.com/products/facebookexperimental-robyn
Homepage: https://facebookexperimental.github.io/Robyn/
Language: Jupyter Notebook
License: MIT
License Family: permissive
Topics: marketing-mix-modeling, marketing-mix-modelling, mmm, marketing-science, econometrics, adstocking, cost-response-curve, budget-allocation, hyperparameter-optimization, evolutionary-algorithm, ridge-regression, gradient-based-optimisation
Last push: 2026-01-26T13:18:56+00:00

## Health v2 (maintenance only)
Score: 52/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 64, release rhythm 8, longevity 100
- inputs: {"age_days": 2316, "days_push": 219, "days_rel": 622, "gap_med": null, "n_releases_24m": 1}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1508, forks 430 (observed 2026-08-28T04:04:55.621396+00:00)

## What it is
Robyn is Meta Marketing Science's open-source, semi-automated Marketing Mix Modeling (MMM) package available in R and Python. It uses ridge regression, evolutionary hyperparameter optimization (Nevergrad), Prophet-based time-series decomposition, and gradient-based budget allocation to measure media channel effectiveness in a privacy-safe way.

## Use cases
- build a marketing mix model to measure ad channel ROI
- allocate marketing budget across media channels optimally
- model adstock and saturation curves for digital campaigns
- run privacy-safe marketing measurement without cookies or PII
- decompose sales trends, seasonality, and holiday effects
- calibrate MMM against lift tests and ground-truth experiments

## When to choose
- you have granular time-series marketing spend and outcome data
- you need a free, open-source, privacy-friendly alternative to paid MMM vendors
- you want semi-automated modeling with reduced human bias
- you work in R or Python and want community-supported marketing science tooling

## When to avoid
- you need individual-level or real-time attribution rather than aggregate modeling
- your dataset is small or lacks sufficient historical media spend data
- you need a fully supported commercial product with vendor guarantees
- you require the Python version to be fully production-hardened (it is a beta LLM translation)

## Facets
- artifact type: library
- maturity: active
- function: machine-learning, data-science, analytics
- domain: data-science, analytics
- platform: python, cross-platform
- tags: marketing-mix-modeling, mmm, adstock, budget-allocation, ridge-regression, nevergrad, prophet, marketing-science, econometrics, hyperparameter-optimization, optimization, marketing, r

## Member repositories
- facebookexperimental/Robyn (main) score 52

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:55.621396+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T04:32:32.677935+00:00, confidence not recorded.
  - readme: https://github.com/facebookexperimental/Robyn (fetched 2026-08-28T04:04:55.621396+00:00, sha d30cbf6f83af)
  - homepage: https://facebookexperimental.github.io/Robyn/ (fetched 2026-08-29T11:36:53.298019+00:00, sha 18fa3acb8f83)
- Data as of 2026-08-30T08:39:29.467469+00:00.