# google/lightweight_mmm

LightweightMMM 🦇 is a lightweight Bayesian Marketing Mix Modeling (MMM) library that allows users to easily train MMMs and obtain channel attribution information.

Repository: https://github.com/google/lightweight_mmm
Canonical: https://ross.abutalabs.com/products/lightweight_mmm
Homepage: https://lightweight-mmm.readthedocs.io/en/latest/index.html
Language: Python
License: Apache-2.0
License Family: permissive
Topics: bayesian, econometrics, marketing-science, mmm, data-science
Archived: true
Last push: 2025-06-17T15:20:20+00:00

## Health v2 (maintenance only)
Score: 10/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 27, release rhythm 8, longevity 100
- inputs: {"age_days": 1665, "days_push": 442, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: archived
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1052, forks 234 (observed 2026-08-28T04:03:23.452529+00:00)

## What it is
LightweightMMM is a Python library for Bayesian Marketing Mix Modeling (MMM) built on JAX and Numpyro, helping advertisers measure media channel effectiveness and optimize budget allocation. It has been deprecated in favor of Google's successor library, Meridian, and is no longer supported.

## Use cases
- estimate optimal marketing budget allocation across media channels
- measure advertising effectiveness with Bayesian marketing mix modeling
- attribute sales KPI changes to online and offline media channels
- model media spend impact with uncertainty quantification
- build hierarchical MMM models across geographies

## When to choose
- maintaining or reproducing legacy models originally built with LightweightMMM
- studying the library's approach to Bayesian MMM before migrating to Meridian

## When to avoid
- starting a new marketing mix modeling project - use Google Meridian instead
- expecting bug fixes or support, since the project is officially unsupported
- needing production-grade MMM tooling with active maintenance

## Facets
- artifact type: library
- maturity: abandoned
- function: machine-learning, data-science, analytics
- domain: data-science, analytics
- platform: python, cross-platform
- tags: bayesian, marketing-mix-modeling, media-attribution, jax, numpyro, budget-optimization, deprecated, marketing-science

## Member repositories
- google/lightweight_mmm (main) score 10

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:03:23.452529+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 2026-08-30T06:59:43.413151+00:00, confidence not recorded.
  - readme: https://github.com/google/lightweight_mmm (fetched 2026-08-28T04:03:23.452529+00:00, sha 21e73538ea35)
  - registry_pypi: https://pypi.org/pypi/lightweight_mmm/json (fetched 2026-08-29T13:01:52.720758+00:00, sha 820085ce3da9)
- Data as of 2026-08-30T08:39:29.467469+00:00.