Analysis of active users and lost users in the website

Web page Production WEBJX article introduction: The goal of website user management is to discover new users and keep old users. But it's good to just attract new users, to keep the new user active, to create value for the site, and once the user's activity declines, it's likely that the user will gradually move away from the site and lose it. So based on this, we can make another segment of active users and lost users. Live

The goal of website user management is to discover new users and keep old users. But it's good to just attract new users, to keep the new user active, to create value for the site, and once the user's activity declines, it's likely that the user will gradually move away from the site and lose it. So based on this, we can make another subdivision of the user-active users and lost users.

Active Users and lost users

Active users, here is relative to the "lost user" concept, refers to those "live" users, users will occasionally patronize the site, while bringing some value to the site. At the same time, we also need to know how many users may have abandoned our site, it is impossible to create any value for the site, that is, the so-called loss of users.

Loss of users, refers to those who have visited the site or registered users, but because the site gradually lost interest gradually away from the site, and then completely out of the site of the group of users. Of course, a site will have lost users, this is the site users succession inevitable, but the loss of the proportion of users and change trends can explain the site to retain the user's ability and development trends.

As a simple example, we can often see some data analysis report said: "So-and-so site has more than millions of registered users, but in fact, this data does not make much sense, because it is possible that many users in the millions of has no longer login to the site (lost users), Users who have been truly recently logged in or have had operational behavior (active users) are actually less than 10,000. So for a website, what really makes sense is the number of active users, not the total number of users, because only these users are creating value for the site.

Active users measure the status of the site's operations, while lost users are used to analyze whether the site is at risk of being eliminated, and whether the site has the ability to retain new users.

Active user Analysis

The previous article in my blog--a measure of user activity has been introduced using engagement to measure user activity, and engagement's active traffic (Visits) is calculated based on the definition of the site, which can also be used to compute the number of active users in a site (Unique Visitors). You can also calculate the number of active users at different time intervals, such as daily, weekly, monthly ... Here is no longer described in detail, need to pay attention to the following issues:

1, the user engagement definition, and the only user for the unit statistics;

2, as long as the user has any one by one engagement behavior, can be defined as active users;

3, do not only focus on active users, try to analyze the trend of active users and the proportion of the change.

Loss User Analysis

The definition of the loss of users is relatively simple, is a period of time did not visit or login to the site's users, the general loss of users are for those who need to register, provide application services, such as micro-blog, mailbox, E-commerce sites. Different sites for the definition of loss may vary, for micro-blogs and mailboxes such users almost every day to visit the site, may not be logged in more than 1 months, we can think that the user may have been lost, and for E-commerce, May be 3 months without a login or six months without any purchase behavior can be identified as the loss of users. The following analysis is mainly based on the registered users of the website, because such users are easier to identify, and analyzing the loss of such users is more meaningful to the site.

Access to Data

The lost user is identified by the current time of the user's last login, so to analyze the loss of users, need to know each user's last logon time, and for different sites, this time interval will vary, the longest may be 1 years or more, so in the data acquisition will have a certain degree of difficulty. If you are analyzing a registered user, then the general Web site will be in the database to build the corresponding table to store user information, so it is recommended to store the user's basic information while recording the user's most recent logon time, so that the user can accurately calculate the most recent logon distance between the current time interval, and then distinguish whether the user lost.

Changing trend of Lost users

First of all, it needs to be clear that the loss of the user may not be permanent, perhaps the user in a period of time for the site does not have any needs, then he will stay away from the site for a longer period of time, or the loss of users will be due to the site of a marketing or site quality improvement Web site Total loss of the number of users of the calculation is relatively simple, in more than 1 months to login that is lost as an example, then total loss of users is all "current time point-user last time point" 1 months "number of users." But the sheer number of total loss of users is meaningless for analysis, because in most cases this value is always increasing, we need to calculate total loss of users as the total number of users and the number of new lost users, observe their trend of change, such as the following table:

New user Churn Rate

Perhaps your site has attracted a number of new visitors, and they successfully registered as a user of the site, you have a good start, has been half the success, then the other half? Is how to retain these new users, so that they continue to bring value to the site, which is to analyze the significance of the loss rate of new users.

We can think that the new user registered after the completion of the first landing, then simply define the loss of new users, is the user after registration for a period of time has not logged on the site, that

Current point of time-user registration point of loss critical time interval

For example, we define a user's loss critical time interval of 1 months, that is, the user who has not logged in for one months after registration means that it has been lost, so it is possible to calculate the number of new user losses per day, that is, when the registration time is 1 months ago, and from the registration to the number of users who are not currently logged in. The ratio of users to the total number of registered users for the day 1 months ago is the loss rate for new users:

Number of new users lost/day Total Registered users = new user Churn rate

Calculate the rate of new user churn per day and observe its trends:

Whether the site can keep new users is the ability to continuously reduce the wastage rate of new users.

Summary, here is mainly about how to analyze the site really has the number of valuable active users and the ability of the site to retain these users, you can use the trend of loss of users to measure the overall loss of site users, with the new user wastage rate to measure the site to retain the ability of new users, The analysis of the proportion of active users and the trend analysis can measure the quality and value of the existing users of the website.