Lighthouse

Level 1 · Cohorts by first order

From Jordan Reyes, Head of Marketing

“Before we talk retention I want to know who we are retaining. How many new customers did we win each quarter since 2023? I suspect the holiday quarters bring in a lot of them.”

Your goal

Booked orders are those with status "completed" or "refunded". Set result to a DataFrame with one row per quarter of each customer's first booked order and columns cohort (the quarter as text, like "2023Q1") and new_customers (customers whose first booked order falls in that quarter), sorted by cohort ascending, with a default 0..n index.

Use the orders DataFramenot done yet
Show exactly cohort new_customersnot done yet
12 rows, sorted as askednot done yet

Your code

Python

Loading the editor…

Reference solution

ResultPreview of orders · press Run to see yours

Preview of orders
order_id	customer_id	order_date	status	channel
1001	125	2023-01-01 20:51:19	completed	web
1002	156	2023-01-02 07:58:46	completed	web
1003	22	2023-01-02 10:25:27	completed	marketplace
1004	48	2023-01-02 11:21:51	completed	marketplace

Mote checks your result

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

DataFrames

Loaded for you as pandas DataFrames, plus pd and np.

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	NaN
1002	156	2023-01-02 07:58:46	completed	web	0.00	NaN
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

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	NaN
3	Henry	Brooks	hbrooks@lakesidecafe.com	Los Angeles	CA	2022-06-04	business	NaN
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	NaN

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	NaN	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

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

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	NaN	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	NaN	2023-01-01 23:02:24	mobile	paid	1	false
2	18	2023-01-01 15:07:49	desktop	paid	1	false
3	NaN	2023-01-01 16:55:46	desktop	direct	7	false
4	NaN	2023-01-01 12:43:28	mobile	organic	3	false
5	NaN	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

What a cohort is

A cohort is a group of customers who share a starting point. For a store, the natural starting point is the first purchase: everyone whose first order fell in January to March 2023 is the 2023Q1 cohort. From then on you follow the same people over time and ask what share of them came back.

Cohorts fix the biggest problem with “repeat rate this year”: the customers in the denominator are not comparable. Someone who first bought in December has had weeks to come back; someone who first bought three years ago has had years. Grouping by start date compares like with like.

One date per customer

The cohort comes from a single date per customer, the earliest booked order:

booked = orders[orders["status"].isin(["completed", "refunded"])]
first = booked.groupby("customer_id")["order_date"].min()

Only booked orders count. A cancelled order was never paid, and a pending one may never be. A customer whose first attempt was cancelled and whose first real purchase came two months later belongs to the later cohort. Refunded orders do count: the customer did buy, even if the product came back.

Months or quarters

Monthly cohorts are the textbook default, but Harbor & Pine wins about 100 to 250 new customers a quarter. Split by month, 22 of the 36 cohorts hold between 28 and 42 people, and their retention numbers jump around from noise alone. Quarterly cohorts are big enough to read. pandas turns a date into its quarter with a period:

first.dt.to_period("Q").astype(str)   # "2023Q1", "2023Q2", ...

Text like "2023Q1" sorts correctly, which keeps the table in time order.

Reading the sizes

Look at the sizes before any retention number. Here the fourth quarter of each year brings in far more customers than the others, because holiday shoppers arrive in November and December. That matters later: holiday cohorts are full of gift buyers, who often do not come back, so a big Q4 cohort can drag the overall retention rate down without anything getting worse.

The 2023Q1 cohort is big for a different reason: 116 of its 177 customers signed up before 2023. The data starts in January 2023, so for them “first order in the data” is not their first order ever. Treat that cohort as existing customers, not new ones.

Common mistakes

Using the signup date. Signing up is not buying. About 230 customers never ordered at all.
Using any order status. Cancelled first orders put people in the wrong cohort.
Too many, too small cohorts. If a cohort has a few dozen people, merge to quarters.

Level 1 · Cohorts by first order

Your goal

Your code

Reference solution

Mote checks your result

What Jordan gets, and why

DataFrames

The lesson

What a cohort is

One date per customer

Months or quarters

Reading the sizes

Common mistakes

Lantern lit!

Level 1 · Cohorts by first order

The request

Your goal

Your code

Reference solution

Mote checks your result

What Jordan gets, and why

DataFrames

The lesson

What a cohort is

One date per customer

Months or quarters

Reading the sizes

Common mistakes

Cookie choices