How SQL Queries Work

You write what you want. You never write how to get it—that part is PostgreSQL's job.

SQL is Declarative

Most programming languages are imperative: you write the exact steps to take in order. SQL is declarative: you describe the result you want and the database figures out the steps.

IMPERATIVE — MOST LANGUAGES

Open the data, loop through every row, check the channel, and append each match.

for campaign in campaigns:
  if campaign.channel == "google_ads":
    results.append(campaign)

DECLARATIVE — SQL

Describe the outcome. PostgreSQL decides whether to scan rows, use an index, or choose another plan.

SELECT *
FROM marketing.campaigns
WHERE channel = 'google_ads';

What Happens When You Run a Query

Behind a short SQL statement, PostgreSQL runs a four-stage pipeline before data comes back.

id	name	channel	spend	start_date	end_date	status	target_segment
1	Spring Launch	google_ads	55000.00	2024-01-15	2024-03-31	active	smb
2	Retention Webinar	email	45000.00	2024-02-10	2024-04-15	active	enterprise
3	Finance Retargeting	linkedin	50000.00	2024-03-12	2024-05-31	active	enterprise
4	Enterprise Search	google_ads	60000.00	2024-04-01	2024-06-30	active	enterprise

Watch a query travel the pipelineStage 1 of 4

1Parser

2Optimizer

3Execution engine

4Result set

ParserValidate the marketing.campaigns query, tables, and columns

Why This Matters

Because PostgreSQL decides how, it can physically execute a query differently from the written order. Logical clause order is fixed, which is what the next lesson breaks down.

Now You Try

Practice this concept

Growth wants active-user counts by acquisition channel while PostgreSQL chooses the physical execution plan.

Available schema

growth

Prefix tables with growth.table_name.

channelactive_users

growth.users

Column	Type
id	integer
created_at	timestamp with time zone
country	text
channel	text
plan	text
activated_at	timestamp with time zone
churned_at	timestamp with time zone
legacy_user_code	text

query.sql

Beginner business practice