AARRR user operation analysis

1. Overview

1.1 Concept

This article uses the AARRR model to analyze user operations. The AARRR model is also called the pirate model. It is a commonly used model in the user operation process. It explains the five indicators for user growth: customer acquisition, activation, retention, revenue, and dissemination. From customer acquisition to dissemination of recommendations, the entire AARRR model forms a closed-loop model of the user's full life cycle, continuously expanding the scale of users and achieving sustained growth.

1.2 Expected effect

The specific situation of each product is different, but overall it includes the development process of these 5 aspects. This article uses FineBI to analyze the five development processes of "Yimaicai" APP.

1.3 Get data

Sample data used in this case: 

374.AARRR Sample data.zip

2. Acquire customers 

Acquiring customers means attracting new customers, that is, letting users know that there is such an APP and try it out. Under normal circumstances, there will be multiple channels to increase product exposure, but how to choose the best channel and use the least budget to get the best pull-up effect? The first thing to do is channel analysis.

Generally, there are two dimensions in channel analysis: Number of acquired customers and qality of acquired customers. (In this article, we use the average browsing time to open the APP as the evaluation standard of qality of acquired customer.)

Conclusion : It can be seen that the quantity and quality of offline activity promotion are the best, and the investment in offline activities can be increased. Offline activities in supermarkets or vegetable markets are the best option.

3. Activation

Activation does not directly correspond to successful registration. The activation needs to be active customers, and what should be considered is the user's use of the product's core functions.

For example: short video software requires a new user to watch for a certain period of time, and chat software requires a new user to complete a conversation before it is activated. Then in the Maicai APP, we consider users who have purchased once as active users.

Analyze the newly added users in each month, and the line chart is as follows:

Conclusion : The activation rate decreased in October, and specific reasons need to be analyzed. At the same time, it assists in new customer activities, and performs refined operations, and recommends personalized products on the home page to attract users.

4. Retention

After the user is activated, if they don't keep it, they will eventually be lost, which is a futile one.

So user retention analysis are also very important. Refer to  retention analysis to calculate the one-week retention rate/two-week retention rate/30-day retention rate of activated users. As shown below:

Conclusion : There is a lot of room for the user retention rate to improve. It is necessary to analyze the reasons for the loss, further improve the product experience, and retain existing customers.

5. Benefits

When a user becomes one of your users after activation, what needs to be considered is how to obtain income and achieve profitability. The profitability of food shopping software is related to many indicators. Here we will temporarily increase the user's purchasing activity as a main way to increase income.

We divide users into three broad categories: low active users, ordinary users, and member. Use the funnel chart to display, as shown in the following figure:

Conclusion : The number of low active users is huge and has great potential. Active low active users, keep member users.

6. Dissemination

When the product has a certain scale of users, it needs to consider stimulating spontaneous communication among users. The self-propagation data index is the K valuer (recommended coefficient):

K = (the average number of invitations sent by each user to his friends) * (the conversion rate of those who received the invitation into new users)   

The level of K value directly reflects the level of self-propagation results. When the K value is greater than 1, it will stimulate a huge force of self-propagation. The larger the K value, the stronger the force. And if the K value is less than 1, then the transmission level will gradually weaken until it disappears.

Calculate the K value of the APP, as shown in the figure below:

Conclusion : The K value of APP is greater than 1, and it has the power of self-propagation. It is possible to further increase the K value through operational activities such as "invite to receive a red envelope" and speed up the speed of dissemination.