LTV decomposition by acquisition channel

Your manager is waiting

A star schema with 5 tables, 7,400+ rows, and intentionally planted data quality issues. NULLs that change your averages, duplicates that skew your top-10 lists, refunds that flip campaign rankings, and inconsistent formats that break your filters.

Query a realistic marketing database with planted data quality issues, get graduated feedback from a 5-tier validation engine, and build the instincts that separate junior analysts from senior ones.

Solve marketing analytics problems from fictional managers, just like a real job.

Real-World Scenarios

Work with NULLs, duplicates, inconsistent formats, and refunds — the stuff tutorials skip.

Messy Data

5-tier validation engine compares your results against multiple accepted approaches.

Instant Feedback

Every mission maps to a real marketing analyst task — campaign reports, ROI analysis, data cleaning, and executive scorecards you’ll see on the job.

Portfolio-Ready Skills

Your manager needs numbers.

Welcome — preview the customer table

What campaigns are running?

How many emails did we send?

Where are our customers?

What are people buying?

Count unique customers

Which coupon is pulling its weight?

How many customers never bought anything?

Which channel gets the best open rate?

How many customers are repeat buyers?

Which campaign drove the most revenue?

Which campaign has the best ROI?

Top 5 customers by spend

Revenue by state

What's our average order value per campaign?

Is our email actually working?

Clean up the traffic sources

Normalize the gender mess

How many organic purchases are we getting?

How many customers opened zero emails?

Something's off with our numbers

Fix the product category mess

Customers buying across categories

Total refund impact

The full campaign scorecard

Repeat-purchase cohort retention

Multi-touch attribution — last-click vs first-click vs linear

Send fatigue — engagement decay by send rank

Marketing year-end capstone — board KPI summary

customer_id

first_name

last_name

email

signup_date

city

state

gender

dim_customers

campaign_id

campaign_name

channel

start_date

end_date

budget

status

dim_campaigns

send_id

sent_at

opened

clicked

fact_sends

session_id

source

landing_page

session_start

session_end

pages_viewed

fact_sessions

purchase_id

purchase_date

amount

product_category

coupon_code

fact_purchases

LTV decomposition by acquisition channel

The Brief

You'll practice