pymc-marketing-mmm-clv — Claude Code skill: Bayesian marketing
A skill for Bayesian marketing analytics with PyMC: media mix modeling, customer lifetime value and BTYD models. Useful when you need channel contributions and ROAS.
npx skills add reason-machines/marketing-skills --skill pymc-marketing-mmm-clv -a claude-codeWhat pymc-marketing-mmm-clv does
The skill wires the PyMC-Marketing library into your agent and guides it through building a media mix model. It covers adstock transformations — geometric, delayed and Weibull — plus saturation functions: logistic, Michaelis-Menten and tanh. Time-varying effects are there too, for baseline sales and media contribution.
For CLV you get Beta-Geometric/NBD, Pareto/NBD and Gamma-Gamma. Customer choice analysis ships alongside. Models can be calibrated with lift test results, then used to allocate budget across channels with bounds and a planning horizon.
What you need first: PyMC-Marketing installed via conda or pip, with Docker support for Jupyter workflows. Your data has to be a table with a date column, channel columns and controls such as holidays or promotions. No dated channel history means no media mix model — worth fixing your data collection before anything else.
What it helps with
- build a media mix model with adstock and saturation
- compute ROAS and per-channel contributions
- allocate budget across channels under bounds
- model customer lifetime value for non-contractual buyers
- calibrate a model with lift test results
Example prompts for the agent
- create a media mix model for tv, radio and digital with geometric adstock
- calculate ROAS and channel contributions from this model
- build a customer lifetime value model for non-contractual customers
How to install pymc-marketing-mmm-clv
1Skills CLI (Vercel)
npx skills add reason-machines/marketing-skills --skill pymc-marketing-mmm-clv -a claude-code2Manually: put the files in place
~/.claude/skills/pymc-marketing-mmm-clv/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/pymc-marketing-mmm-clv/
1Skills CLI (Vercel)
npx skills add reason-machines/marketing-skills --skill pymc-marketing-mmm-clv -a codex2Manually: put the files in place
.agents/skills/pymc-marketing-mmm-clv/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add reason-machines/marketing-skills --skill pymc-marketing-mmm-clv -a cursor2Manually: put the files in place
.cursor/skills/pymc-marketing-mmm-clv/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install reason-machines/marketing-skills pymc-marketing-mmm-clv2Manually: put the files in place
.github/skills/pymc-marketing-mmm-clv/SKILL.md1Skills CLI (Vercel)
npx skills add reason-machines/marketing-skills --skill pymc-marketing-mmm-clv -a gemini-cli2Platform CLI
gemini skills install https://github.com/reason-machines/marketing-skills --path skills/pymc-marketing-mmm-clvManually: put the files in place
.devin/skills/pymc-marketing-mmm-clv/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add reason-machines/marketing-skills --skill pymc-marketing-mmm-clv -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 pymc-marketing-mmm-clv
How do I install pymc-marketing-mmm-clv in Claude?
Run npx skills add reason-machines/marketing-skills --skill pymc-marketing-mmm-clv -a claude-code. The skill lands at ~/.claude/skills/pymc-marketing-mmm-clv/SKILL.md.
What does pymc-marketing-mmm-clv do?
It helps the agent build Bayesian media mix, customer lifetime value and BTYD models on PyMC-Marketing, including ROAS calculation and budget allocation.
What do I need for this skill to work?
The PyMC-Marketing library installed via conda or pip, and a dataset with a date column, channels and control variables. Docker support is included for Jupyter workflows.
