User stickiness analysis

1. Overview

1.1 Background

User stickiness refers to the degree of dependence and re-consumption expectation formed by the combination of users' loyalty and trust to the brand or product and benign experience. It is an important indicator to understand the health of the product so as to understand the team's continuous improvement of the product Zhang Ming, the product student, plans to make a user stickiness dashboard to observe the stickiness

1.2 Analysis ideas

Count the distribution of the number of days users have used the product in a week, how many people have used the product for 2 days, and how many people have used it for 7 consecutive days;

the trend of the number of people who log in for more than two days in a week over time, and the number of people who log in more than two days a week The more, the better the health of the product;

then the stickiness of "cooperation, follow-up, potential" users can be classified and viewed.

Click to view dashboard: user stickiness analysis.

1.3 Analysis results

• From the first component, we can see [05 weeks-10 weeks] that the number of people who log in more than twice a week continues to rise; combined with the line chart component on the right, we can see that the most obvious rise is potential users. It shows that the activity or product optimization during this time period has a positive impact on user viscosity feedback, especially for potential users.

• 74.83% of users only log in once, so overall, user stickiness needs to be strengthened.

1.4 Get data

Sample data used in this article: user login table.xlsx

2. Implementation

2.1 Prepare data

1) Upload the sample data "user login table".

2) Add a self-service dataset and name it "User Stickiness Analysis", check the three fields of "user login table", as shown in the figure below:

3) Next, we need to perform deduplication. If a user logs in multiple times in a day, only one record needs to be kept. Add "Group Summary", as shown in the figure below:

4) Add a counting column, and label "1" for each row of data after all the duplicates are removed, as shown in the following figure:

5) Count how many times each user logs in each week. Add "Group Summary" and drag it into the field to change the group of the login time to "number of weeks in year", as shown below:

2.2 Make the component

2.2.1 View the trend of the number of people who log in more than twice a week

1) Convert the "phone" dimension into an indicator to get the deduplication count of phone (that is, the number of all users), as shown below:

2) Perform detail filtering of the "phone" after deduplication, and filter to get the number of users who log in more than once in a week, as shown below:

3) Drag into the field "login time, phone" to create a trend area chart, as shown below:

In this way, the time chart of the number of people who log in more than twice a week is completed.

2.2.2 View the trend of the number of users with different status who log in more than twice a week

Copy a component made in section 2.2.1, drag the "client condition" into the color bar and change the "Area" to "line" on the basis of it, as shown below:

In this way, the trend of "cooperation, potential, follow-up" registered more than twice a week is obtained.

Users of other components can save the dashboard in section 1.2 separately to view the specific methods, which will not be repeated in this article.

2.3 Effect view

Please see section 1.2 for details.