GROUP BY

What it does

GROUP BY collapses rows into one row per distinct combination of the grouping columns, then computes aggregate functions like SUM, COUNT, AVG, and MAX over each group.

Example

SELECT
  customer_id,
  COUNT(*) AS orders,
  SUM(amount) AS total_spend
FROM orders
GROUP BY customer_id;

Important behavior

• every non-aggregated column in the SELECT list must be included in GROUP BY
• grouping changes the grain of the result set
• adding or removing a grouping column changes the meaning of the query

Common mistake

SELECT
  customer_id,
  order_date,
  SUM(amount)
FROM orders
GROUP BY customer_id;

This is invalid in most SQL engines because order_date is not aggregated and not grouped.

Safety checklist

• define the intended aggregation grain
• ensure all selected columns are either grouped or aggregated
• review the result set before using it in dashboards or exports