This site turns a raw e-commerce event log into the four things a growth team needs to act: who the buyers are, whether they come back, which ones are worth targeting, and how to test a change before rolling it out. Every number is precomputed from one pinned data build.
Three findings carry most of the weight. Each links to the section that works through it.
The sections are ordered as an argument, from what happened to what to do next.
What you are looking at, and what you are not.
The source is a public dataset of shopping events from October and November 2019. The numbers here describe that window: 1,645,912 events, 135,767 users, 524,612 sessions, and $7,467,390 in purchase-event revenue.
The build pipeline constructs a star schema, scores every pre-November user with a frozen LightGBM model, mines product co-purchase rules, and precomputes an A/B grid. The frontend reads the output and renders it. There is no server on the request path, and the 2,897 Champions figure, like every other, comes straight from that build.
Segment share of buyers sums to 100%.