With a cohort analysis you can divide generated data into relevant user groups. This makes it easier to recognize valuable information and patterns about groups of users. This saves a lot of time compared to looking at each individual user.
To perform cohort analysis, you must first determine which group (cohort) you are going to look at. This group depends on the specific service, website or application to be analyzed. Often the people in a cohort share one or more common characteristics over a period of time.
A cohort can consist of people who visited the website in the month of March. Whether a cohort consists of people of the same age or gender. There are therefore all kinds of characteristics that can fall under a cohort. As long as the distinguishing feature provides valuable data for the cohort analysis, you’re fine.