3

SQL
DATA
ANALYTICS
PROJECT
1
 Advanced Data Analytics
.

Answer Business Problems:

Complex Queries

Window Functions

Subqueries

CTE

Reports

2
Change Over Time, Trends

∑[ 𝑚𝑒𝑎𝑠𝑢𝑟𝑒] 𝑏𝑦 [𝐷𝑎𝑡𝑒 𝐷𝑖𝑚𝑒𝑛𝑠𝑖𝑜𝑛]

• Total sales by year

• Average cost by year

Year Sales

2023 300

2024 100

2025 200

300

200

100

2023 2024 2025

3
 Problem: Analyze sales performance over time.

create database data_warehouse; -- create a database

use data_warehouse; -- use database

select*from customers;

select*from products;

select*from sales;

select*from report_customers;

select*from report_products;

SELECT

year(order_date) as order_year,

month(order_date) as order_month,

sum(sales_amount) as total_sales,

count (distinct customer_key) as total_customers,

sum(quantity) as total_quantity

FROM sales

WHERE order_date IS NOT NULL

group by year(order_date), month(order_date)

ORDER BY year(order_date),month(order_date)

LIMIT 184467440 OFFSET 1; -- here we have to give maximum limits

4
5
SELECT

DATE_FORMAT(order_date, '%Y-%m-01') AS order_date, -- date formatting

SUM(sales_amount) AS total_sales,

COUNT(DISTINCT customer_key) AS total_customers,

SUM(quantity) AS total_quantity

FROM sales

WHERE order_date IS NOT NULL

GROUP BY DATE_FORMAT(order_date, '%Y-%m-01')

ORDER BY DATE_FORMAT(order_date, '%Y-%m-01')

LIMIT 18446744073 OFFSET 1;

6
 Cumulative Analysis

∑[ 𝐶𝑢𝑚𝑢𝑙𝑎𝑡𝑖𝑣𝑒 𝑀𝑒𝑎𝑠𝑢𝑟𝑒] 𝑏𝑦 [𝐷𝑎𝑡𝑒 𝐷𝑖𝑚𝑒𝑛𝑠𝑖𝑜𝑛]

• Running total sales by year

• Moving the average sales by month

Cumulative value

2024 300 300

2025 100 400

2026 200 600

7
 Problem: Calculate the total sales per month and the running
total of overtime sales.

SELECT

order_date,

total_sales,

SUM(total_sales) OVER (partition by order_date order by order_date) AS

running_total_sales --window fcn

FROM

SELECT

DATE_FORMAT(order_date, '%Y-%m-01') AS order_date,

SUM(sales_amount) AS total_sales

FROM sales

WHERE order_date IS NOT NULL

GROUP BY DATE_FORMAT(order_date, '%Y-%m-01')

order by total_sales

LIMIT 184467440 OFFSET 1

) AS monthly_sales

ORDER BY order_date;

8
SELECT

STR_TO_DATE(CONCAT(order_year, ' -01-01'), '%Y-%m-%d') AS order_date,

total_sales,

SUM(total_sales) OVER (ORDER BY order_year) AS running_total_sales,

ROUND(AVG(avg_price) OVER (ORDER BY order_year), 2) AS

moving_average_price

FROM (

SELECT

YEAR(order_date) AS order_year,

SUM(sales_amount) AS total_sales,

avg(price) avg_price

9
 FROM sales

WHERE order_date IS NOT NULL

GROUP BY YEAR(order_date)

ORDER BY YEAR(order_date)

LIMIT 184467440737 OFFSET 1

) AS yearly_sales

ORDER BY order_year;

10
 Performance Analysis

𝐶𝑢𝑟𝑟𝑒𝑛𝑡 [𝑀𝑒𝑎𝑠𝑢𝑟𝑒] − 𝑇𝑎𝑟𝑔𝑒𝑡 [𝑚𝑒𝑎𝑠𝑢𝑟𝑒]

• Current sales – Average sales

• Current year sales – Previous year sales
• Current sales – Lowest sales

Current Target Performance

A 400 400 0

B 600 400 200

C 200 400 -200

11
Problem: Analyze the yearly performance of each product by
comparing its sales to both its average annual sales and the previous
year's sales.

WITH yearly_product_sales AS ( -- starting with CTE

SELECT

YEAR(s.order_date) AS order_year,

p.product_name,

SUM(s.sales_amount) AS current_sales

FROM sales s
Joined sales with products on
LEFT JOIN products p ON s.product_key = p.product_key product_key using LEFT JOIN
to retain all sales data.
WHERE s.order_date IS NOT NULL

GROUP BY YEAR(s.order_date), p.product_name

SELECT

order_year, product_name, current_sales,

ROUND(AVG(current_sales) OVER (PARTITION BY product_name)) AS

avg_sales,

current_sales - ROUND(AVG(current_sales) OVER (PARTITION BY

product_name)) AS diff_avg,

case when current_sales - ROUND(AVG(current_sales) OVER (PARTITION BY

product_name)) > 0 then "above Avg"
when current_sales - ROUND(AVG(current_sales) OVER (PARTITION BY
product_name)) < 0 then "below Avg"

else "Avg" end avg_change,

12
 -- year over year analysis

LAG(current_sales) over (partition by product_name order by order_year)

prev_sales,

current_sales - LAG(current_sales) over (partition by product_name order by

order_year) as diff_prev,

case
when current_sales - ROUND( LAG(current_sales) over (partition by
product_name order by order_year)) > 0 then "Increase"

when current_sales - ROUND( LAG(current_sales) over (partition by

product_name order by order_year)) < 0 then "Decrease"

else "No change"

end prev_change

FROM yearly_product_sales

WHERE order_year IS NOT NULL;

13
 Proportional Analysis or Part to whole

Assess the performance of individual components in relation to the overall

business, enabling us to identify which category contributes most significantly
to overall impact.

([𝑀𝑒𝑎𝑠𝑢𝑟𝑒]/𝑇𝑜𝑡𝑎𝑙 [𝑀𝑒𝑎𝑠𝑢𝑟𝑒]) ∗ 100 𝑏𝑦 [𝐷𝑖𝑚𝑒𝑛𝑠𝑖𝑜𝑛]

(𝑆𝑎𝑙𝑒𝑠/𝑇𝑜𝑡𝑎𝑙 𝑠𝑎𝑙𝑒𝑠) ∗ 100 𝑏𝑦 [𝐷𝑖𝑚𝑒𝑛𝑠𝑖𝑜𝑛]

(𝑄𝑢𝑎𝑛𝑡𝑖𝑡𝑦/𝑇𝑜𝑡𝑎𝑙 𝑄𝑢𝑎𝑛𝑡𝑖𝑡𝑦) ∗ 100 𝑏𝑦 [𝐶𝑜𝑢𝑛𝑡𝑟𝑦]

A 200 33%

B 300 50%

C 100 17%

17%
33% A
B
C
50%

14
 Problem: Which categories contribute the most to overall sales?

with category_sales as (

select p.category,

sum(sales_amount) as total_sales from sales s

left join products p

on p.product_key= s.product_key group by category order by total_sales desc)

select category, total_sales,

sum(total_sales) over() as overall_sales ,

round((total_sales/ sum(total_sales) over ())*100,2) as "total(%)"

from category_sales;

15
 Data Segmentation

[𝑀𝑒𝑎𝑠𝑢𝑟𝑒] 𝑏𝑦 [𝑀𝑒𝑎𝑠𝑢𝑟𝑒]
[𝑇𝑜𝑡𝑎𝑙 𝑝𝑟𝑜𝑑𝑢𝑐𝑡𝑠 𝑏𝑦 𝑠𝑎𝑙𝑒𝑠 𝑅𝑎𝑛𝑔𝑒]
[𝑇𝑜𝑡𝑎𝑙 𝑐𝑢𝑠𝑡𝑜𝑚𝑒𝑟𝑠 𝑏𝑦 𝐴𝑔𝑒]

Sum Categorize
3 50
Low 7
4 100

5 150
Medium 6
1 200

10 250 High 15

5 300

16
 Problem: Segment products into cost ranges and count how many
products fall into each segment.

use data_warehouse;

with product_segment as (

select product_key, product_name , cost,

case when cost<100 then "below 100"

when cost between 100 and 500 then "100-500"

when cost between 500 and 1000 then "500-1000"

else "above 1000"

end cost_range from products )

select cost_range,

count(product_key) as total_products from product_segment group by

cost_range

order by total_products desc;

17
Problem: Group customers into three segments based on their
spending behavior:
- VIP: Customers with at least 12 months of history and spending
more than €5,000.
- Regular: Customers with at least 12 months of history but spending
€5,000 or less.
- New: Customers with a lifespan less than 12 months.
And find the total number of customers by each group

with customer_spending as (

SELECT

c.customer_key,

SUM(s.sales_amount) AS total_spending,

MIN(s.order_date) AS first_order,

MAX(s.order_date) AS last_order,

TIMESTAMPDIFF(MONTH, MIN(s.order_date), MAX(s.order_date)) AS lifespan

FROM sales s

LEFT JOIN customers c

ON s.customer_key = c.customer_key

GROUP BY c.customer_key

SELECT

customer_segment,

COUNT(customer_key) AS total_customers

18
FROM (

SELECT

customer_key,

CASE

WHEN lifespan >= 12 AND total_spending > 5000 THEN 'VIP'

WHEN lifespan >= 12 AND total_spending <= 5000 THEN 'Regular'

ELSE 'New'

END AS customer_segment

FROM customer_spending

) AS t

GROUP BY customer_segment

ORDER BY total_customers DESC;

19
 Customer Report

Purpose: This report consolidates key customer metrics and behaviors.

Highlights:
1. Gathers essential fields such as names, ages, and transaction details.
2. Segments customers into categories (VIP, Regular, New) and age groups.
3. Aggregates customer-level metrics:
➢ total orders
➢ total sales
➢ total quantity purchased
➢ total products
➢ lifespan (in months)
4. Calculates valuable KPIs:
➢ recency (months since last order)
➢ average order value
➢ average monthly expenses

20
Task:
1. Gathers essential fields such as names, ages, and transaction details (Base
Query)

WITH base_query as ( --By CTE

SELECT

s.order_number, s.product_key, s.order_date, s.sales_amount, s.quantity,

c.customer_key, c.customer_number,

CONCAT(c.first_name, ' ', c.last_name) AS customer_name,

TIMESTAMPDIFF(YEAR, c.birthdate, CURDATE()) AS age

FROM sales s

LEFT JOIN customers c

ON c.customer_key = s.customer_key

WHERE s.order_date IS NOT NULL)

select*from base_query;

21
2) Customer Aggregations: Summarizes key metrics at the customer level.

WITH base_query as (

SELECT

s.order_number,

s.product_key,

s.order_date,

s.sales_amount,

s.quantity,

c.customer_key,

c.customer_number,

CONCAT(c.first_name, ' ', c.last_name) AS customer_name,

TIMESTAMPDIFF(YEAR, c.birthdate, CURDATE()) AS age

FROM sales s

LEFT JOIN customers c

ON c.customer_key = s.customer_key

WHERE s.order_date IS NOT NULL)

SELECT

customer_key,

customer_number,

customer_name,

age,

COUNT(DISTINCT order_number) AS total_orders,

SUM(sales_amount) AS total_sales,

22
 SUM(quantity) AS total_quantity,

COUNT(DISTINCT product_key) AS total_products,

MAX(order_date) AS last_order_date,

TIMESTAMPDIFF(MONTH, MIN(order_date), MAX(order_date)) AS lifespan

FROM base_query

GROUP BY

customer_key,

customer_number,

customer_name,

age;

23
3. Segments customers into categories (VIP, Regular, New) and age groups.

WITH base_query as (

SELECT

s.order_number,

s.product_key,

s.order_date,

s.sales_amount,

s.quantity,

c.customer_key,

c.customer_number,

CONCAT(c.first_name, ' ', c.last_name) AS customer_name,

TIMESTAMPDIFF(YEAR, c.birthdate, CURDATE()) AS age

FROM sales s

LEFT JOIN customers c

ON c.customer_key = s.customer_key

WHERE s.order_date IS NOT NULL)

customer_aggregation AS (

SELECT

customer_key,

customer_number,

customer_name,

age,

COUNT(DISTINCT order_number) AS total_orders,

24
 SUM(sales_amount) AS total_sales,

SUM(quantity) AS total_quantity,

COUNT(DISTINCT product_key) AS total_products,

MAX(order_date) AS last_order_date,

TIMESTAMPDIFF(MONTH, MIN(order_date), MAX(order_date)) AS lifespan

FROM base_query

GROUP BY

customer_key,

customer_number,

customer_name,

age)

SELECT

customer_key,

customer_number,

customer_name,

age,

-- Age group classification

CASE

WHEN age < 20 THEN 'Under 20'

WHEN age BETWEEN 20 AND 29 THEN '20-29'

WHEN age BETWEEN 30 AND 39 THEN '30-39'

WHEN age BETWEEN 40 AND 49 THEN '40-49'

ELSE '50 and above'

END AS age_group,

25
-- Customer segment classification

CASE

WHEN lifespan >= 12 AND total_sales > 5000 THEN 'VIP'

WHEN lifespan >= 12 AND total_sales <= 5000 THEN 'Regular'

ELSE 'New'

END AS customer_segment,

last_order_date,

total_orders,

total_sales,

total_quantity,

total_products,

lifespan

FROM customer_aggregation;

26
4. Calculates valuable KPIs:
CREATE view customers_report as -- make the View

WITH base_query as (

SELECT

s.order_number,

s.product_key,

s.order_date,

s.sales_amount,

s.quantity,

c.customer_key,

c.customer_number,

CONCAT(c.first_name, ' ', c.last_name) AS customer_name,

TIMESTAMPDIFF(YEAR, c.birthdate, CURDATE()) AS age

FROM sales s

LEFT JOIN customers c

ON c.customer_key = s.customer_key

WHERE s.order_date IS NOT NULL)

customer_aggregation AS (

SELECT

customer_key,

customer_number,

customer_name,

age,

COUNT(DISTINCT order_number) AS total_orders,

SUM(sales_amount) AS total_sales,

27
 SUM(quantity) AS total_quantity,

COUNT(DISTINCT product_key) AS total_products,

MAX(order_date) AS last_order_date,

TIMESTAMPDIFF(MONTH, MIN(order_date), MAX(order_date)) AS lifespan

FROM base_query

GROUP BY

customer_key,

customer_number,

customer_name,

age

SELECT

customer_key,

customer_number,

customer_name,

age,

-- Age group classification

CASE

WHEN age < 20 THEN 'Under 20'

WHEN age BETWEEN 20 AND 29 THEN '20-29'

WHEN age BETWEEN 30 AND 39 THEN '30-39'

WHEN age BETWEEN 40 AND 49 THEN '40-49'

ELSE '50 and above'

END AS age_group,

28
-- Customer segment classification

CASE

WHEN lifespan >= 12 AND total_sales > 5000 THEN 'VIP'

WHEN lifespan >= 12 AND total_sales <= 5000 THEN 'Regular'

ELSE 'New'

END AS customer_segment, last_order_date,

TIMESTAMPDIFF(MONTH, last_order_date, CURDATE()) AS recency,

total_orders,

total_sales,

total_quantity,

total_products,

lifespan,

-- Compute average order value (AOV).

CASE

WHEN total_orders = 0 THEN 0

ELSE round( total_sales / total_orders , 2)

END AS avg_order_value,

-- Compute average monthly spends.

CASE

WHEN lifespan = 0 THEN round( total_sales , 2)

ELSE round( total_sales / lifespan , 2)

END AS avg_monthly_spend

FROM customer_aggregation;

29
-- using view

SELECT * FROM data_warehouse.customers_report; -- it showing whole

reports

select age_group,

count(customer_number) as total_customers,

sum(total_sales) total_sales from customers_report

group by age_group;

30
select customer_segment,

count(customer_number) as total_customers,

sum(total_sales) total_sales from customers_report

group by customer_segment;

31
 Products Report

Purpose: This report consolidates key product metrics and behaviors.

Highlights:
1. Gathers essential fields such as product name, category, subcategory, and cost.
2. Segments Products by revenue to identify High-Performers, Mid-Range, or
Low-Performers.
3. Aggregates product-level metrics:
➢ total orders
➢ total sales
➢ total quantity sold
➢ total customers (unique)
➢ lifespan (in months)
4. Calculates valuable KPIs:
➢ recency (months since last sale)
➢ average order revenue (AOR)
➢ average monthly revenue

32
Task:

(1) Base Query: Retrieves core columns from sales and products.

WITH base_query AS (

SELECT

s.order_number,

s.order_date,

s.customer_key,

s.sales_amount,

s.quantity,

p.product_key,

p.product_name,

p.category,

p.subcategory,

p.cost

FROM sales s

LEFT JOIN products p

ON s.product_key = p.product_key

WHERE s.order_date IS NOT NULL -- only consider valid sales dates

select* from base_query

33
2. Segments Products by revenue to identify High-Performers, Mid-Range, or
Low-Performers.

WITH base_query AS (

SELECT

s.order_number,

s.order_date,

s.customer_key,

s.sales_amount,

s.quantity,

34
 p.product_key,

p.product_name,

p.category,

p.subcategory,

p.cost

FROM sales s

LEFT JOIN products p

ON s.product_key = p.product_key

WHERE s.order_date IS NOT NULL -- only consider valid sales dates

product_aggregations AS (

-- Product Aggregations: Summarizes key metrics at the product level.

SELECT

product_key,

product_name,

category,

subcategory,

cost,

TIMESTAMPDIFF(MONTH, MIN(order_date), MAX( order_date)) AS lifespan,

MAX(order_date) AS last_sale_date,

COUNT(DISTINCT order_number) AS total_orders,

COUNT(DISTINCT customer_key) AS total_customers,

SUM(sales_amount) AS total_sales,

SUM(quantity) AS total_quantity,

35
 ROUND(AVG(CASE WHEN quantity != 0 THEN sales_amount / quantity
ELSE NULL END), 1) AS avg_selling_price

FROM base_query

GROUP BY

product_key,

product_name,

category,

subcategory,

cost

SELECT

product_key,

product_name,

category,

subcategory,

cost,

last_sale_date,

TIMESTAMPDIFF(MONTH, last_sale_date, CURDATE()) AS recency_in_months,

-- Product Segment based on sales

CASE

WHEN total_sales > 50000 THEN 'High-Performer'

WHEN total_sales >= 10000 THEN 'Mid-Range'

ELSE 'Low-Performer'

END AS product_segment

from product_aggregations;

36
3. Aggregates product-level metrics: total orders, total sales , total quantity sold,
total customers (unique),lifespan (in months).
WITH base_query AS (

SELECT

s.order_number,

s.order_date,

s.customer_key,

s.sales_amount,

s.quantity,

p.product_key,

p.product_name,

p.category,

37
 p.subcategory,

p.cost

FROM sales s

LEFT JOIN products p

ON s.product_key = p.product_key

WHERE s.order_date IS NOT NULL -- only consider valid sales dates

product_aggregations AS (

SELECT

product_key,

product_name,

category,

subcategory,

cost,

TIMESTAMPDIFF(MONTH, MIN(order_date), MAX( order_date)) AS lifespan,

MAX(order_date) AS last_sale_date,

COUNT(DISTINCT order_number) AS total_orders,

COUNT(DISTINCT customer_key) AS total_customers,

SUM(sales_amount) AS total_sales,

SUM(quantity) AS total_quantity,

ROUND(AVG(CASE WHEN quantity != 0 THEN sales_amount / quantity ELSE

NULL END), 1) AS avg_selling_price

FROM base_query

GROUP BY

product_key,

38
 product_name,

category,

subcategory,

cost

SELECT

product_key,

product_name,

category,

subcategory,

cost,

last_sale_date,

TIMESTAMPDIFF(MONTH, last_sale_date, CURDATE()) AS recency_in_months,

-- Product Segment based on sales.

CASE

WHEN total_sales > 50000 THEN 'High-Performer'

WHEN total_sales >= 10000 THEN 'Mid-Range'

ELSE 'Low-Performer'

END AS product_segment,

lifespan,

total_orders,

total_sales,

total_quantity,

39
 total_customers,

avg_selling_price

from product_aggregations;

4. Calculates valuable KPIs: recency (months since last sale), average order
revenue (AOR), average monthly revenue.

CREATE VIEW products_report AS -- create view

WITH base_query AS (

SELECT

s.order_number,

s.order_date,

s.customer_key,

s.sales_amount,

40
 s.quantity,

p.product_key,

p.product_name,

p.category,

p.subcategory,

p.cost

FROM sales s

LEFT JOIN products p

ON s.product_key = p.product_key

WHERE s.order_date IS NOT NULL

),

product_aggregations AS (

SELECT

product_key,

product_name,

category,

subcategory,

cost,

TIMESTAMPDIFF(MONTH, MIN(order_date), MAX(order_date)) AS lifespan,

MAX(order_date) AS last_sale_date,

COUNT(DISTINCT order_number) AS total_orders,

COUNT(DISTINCT customer_key) AS total_customers,

SUM(sales_amount) AS total_sales,

SUM(quantity) AS total_quantity,

ROUND(AVG(CASE WHEN quantity != 0 THEN sales_amount / quantity ELSE

NULL END), 1) AS avg_selling_price

41
 FROM base_query

GROUP BY

product_key,

product_name,

category,

subcategory,

cost)

SELECT

product_key,

product_name,

category,

subcategory,

cost,

last_sale_date,

TIMESTAMPDIFF(MONTH, last_sale_date, CURDATE()) AS recency_in_months,

-- Product Segment based on sales

CASE

WHEN total_sales > 50000 THEN 'High-Performer'

WHEN total_sales >= 10000 THEN 'Mid-Range'

ELSE 'Low-Performer'

END AS product_segment,

-- Average Order Revenue (AOR)

CASE

WHEN total_orders = 0 THEN 0

42
 ELSE round(total_sales / total_orders,2)

END AS avg_order_revenue,

-- Average Monthly Revenue

CASE

WHEN lifespan = 0 THEN round(total_sales,2)

ELSE round( total_sales / lifespan,2)

END AS avg_monthly_revenue

FROM product_aggregations;

SELECT * FROM data_warehouse.products_report; -- view

Thanks !

43