What is a Cohort? How to Read a Cohort Analysis Chart...

Learn about Cohorts & How to Read a Cohort Analysis Chart + learn a quick dance move to help with the memorization!
WHAT IS A COHORT:
A cohort is a fancy word for a group
It is a predefined grouping of your customers.
HOW TO READ A COHORT CHART:
- Each row represents a cohort of users -- "July 2021"
- Each column represents a month following the cohort's creation. Month 1 being the month they became a customer. Month 2 being the 2nd month after being a customer "August 2021" in this example.
- The value in the cell is the analysis you are computing - Repurchase Rate, LTV, Returning Rate, AOV, Discounts, Refund, etc. etc.
Color shading is usually applied to cohorts to allow your eyes to quickly identify the key differentiators/spikes in behavior.
WHY ARE COHORTS USEFUL:
Cohort analysis allows us to directly compare cohorts to each other - by lining up the start date of each cohort. It is the best way to look at activity and retention numbers on a cohort-by-cohort basis.
Helps you uncover the characteristics in your best and worst performing GROUPS/COHORTS of customers.

[@jamezzsz]

AWESOME VIDEO!! Why dont you guys have more subs? It's literally the best explanation video about this topic

[@djaadiabdellah9081]

I just learned that reading the cohorts diagonally is can be useful😅 Thanks a lot.

[@QuanNguyen71298]

Detailed and most insightful explaination i've eva seen regarding this topic, worth a subscription, keep it up

[@sebachirinos]

Amazing video. thank you for this

[@MUXI_]

super helpful.

[@yuelu2534]

very very helpful. thanks

[@syedsyed777]

Hi yasmin, May I know which tool you used for cohort... thanks in advance

[@glowbandz7499]

I think it's Peel! :)

[@michaelrall613]

Please answer this im so confused, if in the next month the user comes again, will the user be counted again ? So each row is the total of a unique user from the first row ?

[@maxlinardi3208]

No. The unique users are always associated with that specific cohort/row/month. What changed (when they come again) is the column. Example: 2436 unique users registered in Feb 21. All their average purchase in Feb 21 is recorded in cell [Feb 21, 1st month]. If out of these 2436 users make a purchase in the next month (March), the average is put in [Feb 21, 2nd month].

[@user-kt2ed6qx3v]

The first cohort has 2431 people and the next cohort has 4149. Can some people from first cohort be included in next cohorts? Or are they all new?

[@maxlinardi3208]

late answers... But every month it's entirely new users.