Why should you care?

Managers ask “orders per city” or “revenue per day” — that is GROUP BY, not a row-by-row dump.

See it live — copy this example

Run in SQL Server Management Studio (SSMS) or Azure Data Studio.

USE DataVerse;
IF OBJECT_ID(N'dbo.Orders', N'U') IS NOT NULL DROP TABLE dbo.Orders;
CREATE TABLE dbo.Orders (
    OrderId INT IDENTITY(1,1) PRIMARY KEY,
    City NVARCHAR(50) NOT NULL,
    Amount DECIMAL(10,2) NOT NULL
);
INSERT INTO dbo.Orders (City, Amount) VALUES
 (N'Pune', 1200), (N'Pune', 800), (N'Mumbai', 2500), (N'Mumbai', 400);
SELECT City, COUNT(*) AS OrderCount, SUM(Amount) AS Revenue
FROM dbo.Orders
GROUP BY City
ORDER BY Revenue DESC;

What happened?

• Rows group by City.
• COUNT(*) is orders per city; SUM(Amount) is revenue.
• Non-aggregated columns in SELECT must appear in GROUP BY.

Practice next

1. Create and fill dbo.Orders.
2. Run the grouped query.
3. Add AVG(Amount) AS AvgOrder.
4. GROUP BY LEFT(City, 1) for a fun rollup.
5. Add HAVING SUM(Amount) > 1000.

Remember

GROUP BY builds one row per group key. Aggregates summarize each group. SELECT list rules are strict — group or aggregate.

City-wise sales board

Ops dashboard groups DataVerse Orders by City.

Outcome: Mumbai vs Pune revenue shows in one query.