What is deliberately wrong with retail-orders

What is deliberately wrong with retail-orders.

7,000 rows, 12 columns, one row per order line. Seven thousand e-commerce order lines across four channels and six categories, with a real Q4 seasonality bump, a returns flag, and discounting that varies by channel.

Revenue is stored as well as its components, and the two agree — which makes this a good dataset for checking your own arithmetic before you trust a warehouse column. Returns are not uniformly distributed: apparel returns at roughly four times the rate of home goods, which is exactly the kind of thing an unsegmented return-rate KPI hides.

Something to try: Compute return rate by category and by channel — then by both. The SQL playground has it loaded already, and the data explorer will profile every column in one pass if you would rather look before you query.

select *
from retail_orders
limit 20;

The file is generated from a fixed seed, so the CSV you download today is byte-identical to the one from last year. That matters more than it sounds: it means a lesson can say "row 412 is the interesting one" and still be right in eighteen months.

Download it, break it, keep it. CC0 — no attribution required, no account, no email.