Lighthouse

Level 7 · Boss: the retention review

From Robert Kim, Chief Operating Officer

“I'm presenting retention to the board next week. I need three things I can defend: how many of last year's customers bought again this year, whether the channel a customer first bought on changes how likely they are to leave, and how much of 2025's sales came from customers we kept versus new ones.”

Your goal

Step 1 of 3. Year-over-year retention. For each year Y in 2023 and 2024, return from_year (Y), customers (customers with a booked order in Y), retained (of those, customers who also have a booked order in Y + 1) and retention_pct (retained × 100 / customers, rounded to 1), sorted by from_year.

Read from the customers tablenot done yet
Read from the orders tablenot done yet
Show exactly from_year customers retained retention_pctnot done yet

Step 1 · your query

SQL

Loading the editor…

Reference solution

ResultPreview of customers · press Run to see yours

Preview of customers
customer_id	first_name	last_name	email	city
1	Owen	Hudson	owen.hudson@icloud.com	San Francisco
2	Douglas	Mendoza	dmendoza69@gmail.com	Atlanta
3	Henry	Brooks	hbrooks@lakesidecafe.com	Los Angeles
4	Brian	Graham	bgraham@bluecedarhomes.com	Seattle

Mote checks your result

Press Run. Each row that matches the goal becomes a firefly in your lantern.

Step 2 · your query

SQL

Loading the editor…

Reference solution

ResultPreview of customers · press Run to see yours

Preview of customers
customer_id	first_name	last_name	email	city
1	Owen	Hudson	owen.hudson@icloud.com	San Francisco
2	Douglas	Mendoza	dmendoza69@gmail.com	Atlanta
3	Henry	Brooks	hbrooks@lakesidecafe.com	Los Angeles
4	Brian	Graham	bgraham@bluecedarhomes.com	Seattle

Mote checks your result

Press Run. Each row that matches the goal becomes a firefly in your lantern.

Step 3 · your query

SQL

Loading the editor…

Reference solution

ResultPreview of customers · press Run to see yours

Preview of customers
customer_id	first_name	last_name	email	city
1	Owen	Hudson	owen.hudson@icloud.com	San Francisco
2	Douglas	Mendoza	dmendoza69@gmail.com	Atlanta
3	Henry	Brooks	hbrooks@lakesidecafe.com	Los Angeles
4	Brian	Graham	bgraham@bluecedarhomes.com	Seattle

Mote checks your result

Press Run. Each row that matches the goal becomes a firefly in your lantern.

Tables

Harbor & Pine store data, ready to query.

customers2,018 rows · 9 columns

customer_idinteger
first_namevarchar
last_namevarchar
emailvarchar
cityvarchar
statevarchar
signup_datedate
segmentvarchar
referral_sourcevarchar

First 5 rows of customers
customer_id	first_name	last_name	email	city	state	signup_date	segment	referral_source
1	Owen	Hudson	owen.hudson@icloud.com	San Francisco	CA	2022-06-02	consumer	search
2	Douglas	Mendoza	dmendoza69@gmail.com	Atlanta	GA	2022-06-02	consumer	NULL
3	Henry	Brooks	hbrooks@lakesidecafe.com	Los Angeles	CA	2022-06-04	business	NULL
4	Brian	Graham	bgraham@bluecedarhomes.com	Seattle	WA	2022-06-06	business	pinterest
5	Kyle	Jimenez	kjimenez27@yahoo.com	Indianapolis	IN	2022-06-07	consumer	NULL

orders12,000 rows · 7 columns

order_idinteger
customer_idinteger
order_datetimestamp
statusvarchar
channelvarchar
shipping_feedecimal
coupon_codevarchar

First 5 rows of orders
order_id	customer_id	order_date	status	channel	shipping_fee	coupon_code
1001	125	2023-01-01 20:51:19	completed	web	0.00	NULL
1002	156	2023-01-02 07:58:46	completed	web	0.00	NULL
1003	22	2023-01-02 10:25:27	completed	marketplace	4.99	WELCOME10
1004	48	2023-01-02 11:21:51	completed	marketplace	4.99	WELCOME10
1005	82	2023-01-02 11:43:36	completed	web	0.00	WELCOME10

order_items30,338 rows · 6 columns

order_item_idinteger
order_idinteger
product_idinteger
quantityinteger
unit_pricedecimal
discount_pctdecimal

First 5 rows of order_items
order_item_id	order_id	product_id	quantity	unit_price
1	1001	121	9	31.99
2	1001	154	3	11.99
3	1002	154	4	11.99
4	1002	194	1	149.00
5	1002	193	2	19.99

products120 rows · 8 columns

product_idinteger
namevarchar
categoryvarchar
subcategoryvarchar
unit_pricedecimal
unit_costdecimal
launched_ondate
is_activeboolean

First 5 rows of products
product_id	name	category	subcategory	unit_price	unit_cost	launched_on	is_active
101	Vale Velvet Loveseat	Furniture	Sofas	1659.00	NULL	2023-09-11	true
102	Alder Oak Coffee Table	Furniture	Tables	919.00	503.50	2024-05-16	true
103	Orchard Rattan Dining Chair	Furniture	Chairs	789.00	429.87	2023-05-27	true
104	Upland Oak Bookcase	Furniture	Storage	429.00	214.70	2022-12-20	true
105	Avery Leather Bed Frame	Furniture	Beds	449.00	232.70	2021-08-31	true

returns1,673 rows · 5 columns

return_idinteger
order_item_idinteger
returned_ondate
reasonvarchar
refund_amountdecimal

First 5 rows of returns
return_id	order_item_id	returned_on	reason	refund_amount
1	14	2023-01-10	damaged	83.69
2	42	2023-01-10	wrong_item	19.99
3	105	2023-01-14	wrong_item	10.99
4	24	2023-01-16	changed_mind	26.99
5	126	2023-01-16	late	20.99

employees60 rows · 8 columns

employee_idinteger
first_namevarchar
last_namevarchar
departmentvarchar
titlevarchar
manager_idinteger
hire_datedate
salaryinteger

First 5 rows of employees
employee_id	first_name	last_name	department	title	manager_id	hire_date	salary
1	Catherine	Hale	Executive	Chief Executive Officer	NULL	2016-03-01	285000
2	Robert	Kim	Executive	Chief Operating Officer	1	2017-01-09	235000
3	Megan	Collins	Sales	Head of Sales	1	2018-04-16	172000
4	Chloe	Gutierrez	Sales	Account Executive	3	2020-01-21	72000
5	Frank	Bell	Sales	Account Executive	3	2023-05-21	73000

web_sessions40,000 rows · 7 columns

session_idinteger
customer_idinteger
started_attimestamp
devicevarchar
sourcevarchar
pages_viewedinteger
convertedboolean

First 5 rows of web_sessions
session_id	customer_id	started_at	device	source	pages_viewed	converted
1	NULL	2023-01-01 23:02:24	mobile	paid	1	false
2	18	2023-01-01 15:07:49	desktop	paid	1	false
3	NULL	2023-01-01 16:55:46	desktop	direct	7	false
4	NULL	2023-01-01 12:43:28	mobile	organic	3	false
5	NULL	2023-01-01 13:06:21	mobile	social	2	false

marketing_spend3,276 rows · 3 columns

spend_datedate
channelvarchar
amountdecimal

First 5 rows of marketing_spend
spend_date	channel	amount
2023-01-01	paid	492.99
2023-01-01	email	37.38
2023-01-01	social	200.53
2023-01-02	paid	709.13
2023-01-02	email	64.13

The lesson

Three questions, three definitions

A retention review for a board answers three questions, and each one needs its own definition written on the slide:

Do last year’s customers buy again? Year-over-year retention: of the customers active in year Y, the share also active in Y + 1.
Who leaves? Churn with a fixed silence window, split by something you can act on.
Where does this year’s money come from? Sales split by new, retained and returning customers.

All three use booked orders only, as in the rest of the chapter.

Step 1: year-over-year retention

Make one row per customer and year, then join that list to itself shifted by one year:

FROM active a
LEFT JOIN active n ON n.customer_id = a.customer_id AND n.yr = a.yr + 1

COUNT(*) counts everyone active in Y; COUNT(n.customer_id) counts only those who matched in Y + 1, because unmatched rows have NULL there. 2025 has no following year in the data, so it is not a starting year.

This rate is easy to explain, but it blends long-standing customers with people who first bought in December. A drop can mean “more new customers”, not “worse retention”.

Step 2: churn by first channel

DuckDB’s arg_min(value, key) returns the value from the row with the smallest key: the channel of each customer’s first order. Then reuse level 4’s rule: eligible customers first bought before 2025-07-04, and churned customers have bought nothing since.

Step 3: where 2025’s sales came from

Label every 2025 buyer by their history. New: the first booked order is in 2025. Retained: they bought in 2024. Reactivated: they bought before 2024, skipped 2024 and returned. Test “new” explicitly: a new customer’s last order before 2025 is NULL, and without that branch they would fall through to the ELSE.

The share column needs a total next to each group. A window over the grouped result does it in one pass:

SUM(net_sales) * 100.0 / SUM(SUM(net_sales)) OVER ()

Common mistakes

Inner join for retention. It drops the customers who did not come back, and retention becomes 100%.
Churn over all customers. Keep the eligibility filter from level 4.
No “new” branch. Without a test on first_at, new customers fall into “reactivated”.

Level 7 · Boss: the retention review

Your goal

Step 1 · your query

Reference solution

Mote checks your result

Step 2 · your query

Reference solution

Mote checks your result

Step 3 · your query

Reference solution

Mote checks your result

What Robert gets, and why

Tables

The lesson

Three questions, three definitions

Step 1: year-over-year retention

Step 2: churn by first channel

Step 3: where 2025’s sales came from

Common mistakes

Lantern lit!

Level 7 · Boss: the retention review

The request

Your goal

Step 1 · your query

Reference solution

Mote checks your result

Step 2 · your query

Reference solution

Mote checks your result

Step 3 · your query

Reference solution

Mote checks your result

What Robert gets, and why

Tables

The lesson

Three questions, three definitions

Step 1: year-over-year retention

Step 2: churn by first channel

Step 3: where 2025’s sales came from

Common mistakes

Cookie choices