Blog · Guide

How SQL GROUP BY Works (Buckets + Totals)

GROUP BY splits rows into groups that share a value, then lets you COUNT/SUM each group — cities, categories, statuses.

rows Austin Milan Austin Milan GROUP BY Austin → COUNT = 2 Milan → COUNT = 2
GROUP BY buckets rows that share a value, then aggregates each bucket.

Group, then aggregate

SELECT city, COUNT(*) AS customers
FROM customers
GROUP BY city;

One result row per distinct city. Every non-aggregated column in SELECT must appear in GROUP BY.

Mental picture

Imagine sorting index cards into piles by city, then writing a sticky note on each pile: “7 customers.” That’s GROUP BY + COUNT.

Try it yourself

The fastest way to lock this in is a short drill on the same Northstar Market tables.

Practice GROUP BY

Related guides