marketing-mix-modeling-pipeline — Claude Code skill: setup
An end-to-end Python Marketing Mix Modeling pipeline covering geometric adstock, Hill saturation curves, OLS regression, and budget optimization.
no license 130 installs ★ 12 updated 2026-10-09
npx skills add reason-machines/marketing-skills --skill marketing-mix-modeling-pipeline -a claude-codeWhat marketing-mix-modeling-pipeline does
This skill sets up a complete Marketing Mix Modeling (MMM) workflow in Python using aggregated historical spend. It computes carryover effects via geometric adstock and accounts for diminishing returns using Hill saturation curves. A statsmodels OLS regression then separates baseline performance from incremental revenue generated by individual media channels.
The project includes Python source modules in src/, exploratory notebooks, and sample budget allocation routines. The mathematical sequence is strict: saturation formulas must run on top of adstocked series, never before them.
Running the code requires Python 3 alongside pandas, numpy, statsmodels, scipy, matplotlib, and seaborn. Since the pipeline processes aggregated weekly numbers, it operates without any user-level tracking.
What it helps with
- Model carryover advertising effects using geometric adstock decay
- Capture diminishing returns with Hill saturation functions
- Estimate channel revenue contribution through OLS regression
- Reallocate marketing spend to maximize predicted revenue
Example prompts for the agent
- Build a marketing mix modeling pipeline for weekly media spend
- Apply geometric adstock and Hill saturation curves to my spend data
- Fit an OLS regression on transformed channel features and extract coefficients
How to install marketing-mix-modeling-pipeline
1Skills CLI (Vercel)
npx skills add reason-machines/marketing-skills --skill marketing-mix-modeling-pipeline -a claude-code2Manually: put the files in place
~/.claude/skills/marketing-mix-modeling-pipeline/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/marketing-mix-modeling-pipeline/
1Skills CLI (Vercel)
npx skills add reason-machines/marketing-skills --skill marketing-mix-modeling-pipeline -a codex2Manually: put the files in place
.agents/skills/marketing-mix-modeling-pipeline/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add reason-machines/marketing-skills --skill marketing-mix-modeling-pipeline -a cursor2Manually: put the files in place
.cursor/skills/marketing-mix-modeling-pipeline/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install reason-machines/marketing-skills skills/marketing-mix-modeling-pipeline2Manually: put the files in place
.github/skills/marketing-mix-modeling-pipeline/SKILL.md1Skills CLI (Vercel)
npx skills add reason-machines/marketing-skills --skill marketing-mix-modeling-pipeline -a gemini-cli2Platform CLI
gemini skills install https://github.com/reason-machines/marketing-skills --path skills/marketing-mix-modeling-pipelineManually: put the files in place
.devin/skills/marketing-mix-modeling-pipeline/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add reason-machines/marketing-skills --skill marketing-mix-modeling-pipeline -a opencodeClaude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about marketing-mix-modeling-pipeline
How to install marketing-mix-modeling-pipeline in Claude Code?
Run npx skills add reason-machines/marketing-skills --skill marketing-mix-modeling-pipeline -a claude-code in your terminal.
Does transformation order matter in this pipeline?
Yes. Geometric adstock must be computed first on raw spend, followed by Hill saturation on the adstocked signal.
What Python dependencies are required?
You will need pandas, numpy, statsmodels, scipy, jupyter, matplotlib, and seaborn.