Stop 08 · Marketing Analytics & DataTopics · 11/11

Basic SQL for Marketers

In short

When the dashboard doesn't have your answer, SQL lets you ask the data directly. A few clauses will cover most marketing questions you'll ever have.

Enough SQL to pull your own campaign or customer data without waiting on an analyst — SELECT, WHERE, GROUP BY and basic JOINs cover most of a marketer's real needs.

The blob leans on an office counter while a clerk character flips through a tall stack of registers.

Think of it like this

Asking the college office clerk

You ask the clerk: 'From the admissions register, only this year's students, count them by city, biggest first.' He flips through, groups and counts. A SQL query is that request, written so the database understands it.

Watch it happen

The order a database runs your query

FROM

WHERE

GROUP BY

SELECT

ORDER BY

1/5Start with the orders table: every order, every column.

The big ideas

What you need to know

Tables and rows

Data lives in tables like orders or users. Each row is one record, each column one detail.

WHERE filters rows

Keep only the rows you care about, like one month, one city or one campaign.

GROUP BY summarises

Combine rows by a column, then use COUNT, SUM or AVG on each group.

JOIN combines tables

Connect orders to users by user_id to answer questions like 'revenue by sign-up channel'.

See it in code

Walk through the code

sql

1SELECT2  channel,3  COUNT(*)         AS orders,4  SUM(amount)      AS revenue,5  ROUND(AVG(amount), 0) AS avg_order6FROM orders7WHERE order_date >= '2026-10-01'8  AND order_date <  '2026-11-01'9GROUP BY channel10ORDER BY revenue DESC;

1/3Pick the channel and three summaries: order count, total and average.

Revenue by channel for October. This one pattern answers a huge number of marketing questions.

Save yourself the pain

Common mistakes

Oops

Using WHERE to filter on a COUNT or SUM.

Do this instead

WHERE runs before grouping. Use HAVING to filter groups, like HAVING COUNT(*) > 10.

Oops

Writing dates as order_date <= '2026-10-31' for timestamps.

Do this instead

Use < '2026-11-01' so orders later on the 31st aren't missed.

Your mission

Run your first queries

~40 min

Quick check

You want only channels with more than 50 orders. Which clause do you add?

Go deeper

1Complete a beginner SQL course or interactive tutorial (many are free).
2Practice writing a query to count conversions by campaign from a sample dataset.
3Learn enough JOIN logic to combine a campaigns table with a conversions table.