marketing-mix-modelling

Installation
SKILL.md

Marketing Mix Modelling

Decides whether a marketing mix model (MMM) is worth building, specifies the data, the open-source tool (Google Meridian or Meta Robyn) and the experiments that calibrate it, and sets which evidence source governs which budget decision. It serves marketing directors, finance leads and analysts who must move money between channels on evidence that survives a board question.

Use When

  • Radio, billboards, TV and online advertising all run together, and finance wants econometrics on which channel drove sales across two or three years of weekly data.
  • The annual channel budget should be reallocated on modelled contribution, not on platform-reported conversions.
  • An analyst or agency proposes Meridian or Robyn: check data readiness, adstock, saturation and seasonality controls before anything is built.
  • A geo test or lift study has finished and its result should calibrate the mix model's priors.
  • Platform dashboards, the CRM and the econometric model disagree on channel contribution; set which source governs which budget decision.

Do Not Use When

Installs
1
GitHub Stars
2
First Seen
3 days ago