Tutorial: product launch signups¶

This is the beginner-friendly one event, one metric, two controls impact example.

Product launch share card

Run it¶

python -m tscfbench demo product-launch

Question¶

How many extra signups appeared after a feature launch, relative to a counterfactual path predicted from related control series?

What this writes¶

impact_metrics.json
impact_report.md
impact_prediction_frame.csv
treated-vs-counterfactual PNG/SVG
cumulative-impact PNG/SVG
share-card PNG/SVG

Why this tutorial exists¶

Product and growth teams often have a clean before/after story but do not want to learn an entire causal-inference vocabulary before getting a usable answer.

Bring your own CSV¶

python -m tscfbench run-csv-impact your_signups.csv \
  --time-col date \
  --y-col signups \
  --x-cols peer_signups search_interest \
  --intervention-t 2024-04-23