The season, or your change?

Move the sliders to run a synthetic geo-holdout. Watch the naive before/after reading follow the season, while the holdout keeps only what your change caused — with the range that says whether the test was big enough. All math runs in your browser.

The numbers are illustrative and synthetic. A real holdout adds refinements (pairing comparable regions, a synthetic-control fallback, spillover handling) that make the read sturdier without changing the idea.