The idea
SELECT chooses columns and WHERE filters rows. SQL treats NULL as missing or unknown, not as zero or an empty string. Use IS NULL to test for missing values; equality with NULL does not work like an ordinary comparison. State which rows belong in a report before writing a query, and keep examples away from production databases.
Worked example
A fictional orders table contains order_id, status and total. SELECT order_id, total FROM orders WHERE status = 'completed' returns completed orders. A missing total needs a separate check using total IS NULL. Silently replacing every unknown total with zero could make the sales report look complete while understating its value.
Try it
Draw five order rows, including two completed orders and one missing total. Work out the result of the completed-order filter by hand. Write a separate missing-total query and explain whether the unknown amount belongs in a revenue total. Use fictional values only.