How do growth hackers perform operational data analysis?

Attracting new users , promoting activation, and retaining users are common indicators in daily operations . Indicators from these angles can provide statistical feedback on the user growth of the APP. But for many operators, the data statistics of conventional indicators only record the numbers themselves, but fail to discover the growth opportunities behind the numbers.

This article will help you establish the correct operational data analysis ideas from three perspectives: promotion of new users, user activity, and user retention .

01 What data should be collected during the customer acquisition stage?

What should we focus on in data analysis during the customer acquisition stage?

In the customer acquisition stage, data analysis mainly solves three problems:

1. Do your advertising costs bring real conversions ?

Usually when we calculate the effectiveness of paid promotions , we measure the channel conversion ROI. Simply put, it is how much money is spent on converting each user, but it is impossible to tell whether effective users are converted.

Therefore, in addition to paying attention to the number of downloads and the number of new users in the statistical dimension, we should also pay attention to the users who visit once. If the number of users who visit once is too large, it means that the channel is not accurate.

By comparing the conversion ROI of each channel and the conversion ROI of different payment methods, we can find the most cost-effective promotion channel . At the same time, we pay attention to the effective ROI and measure the conversion effect from the perspective of user value.

2.What factors affect the number of new users?

Through statistics on data such as download volume and number of new users, the data of each dimension needs to be averaged, and attention should be paid to the comparison between the data of this dimension and the average value, and the comparison with yesterday's data, to observe the changing factors.

At the same time, in order to find out the influencing factors, under the same promotion conditions (the same promotion costs on the same channel , for example, no payment), compare the changes in different months, days of the week and different times of 24 hours, and analyze the impact of time factors on the data.

3. Measure the relationship between impressions and downloads

Exposure represents product brand exposure, and downloads represent actual user growth. By analyzing the download conversion rate , we can observe how many exposures are required for different channels to bring about a download.

At the same time, the analysis of download conversion rate also needs to be triggered from the perspectives of promotional titles , keywords , detailed descriptions, etc., observe the impact of different contents on the data, and find the best promotional copy .

02 What are the statistical dimensions of active user data and how to analyze it?

What should we focus on in data analysis of active users?

Data analysis of active users can solve four problems:

1. Define active user metrics

The "Measurement Dimensions" data in the above table measures different dimensions of active users by dimensions such as visit duration and number of sessions. The average value of each dimension and the number of users above the average value can be analyzed.

The "Collection Index" refers to the data dimension that reflects the user's actual usage behavior of the product. Collection represents the user's recognition of the content. Different products have different definitions of the "Collection Index". For example, e-commerce products use "add to shopping cart" or "browse 3 product detail pages" as measurement dimensions, and define the standard of "active users" through user behavior .

2. Feedback on product health through changes in active users

Through the changes in active users in different time dimensions such as daily active users, 3-day active users, 7-day active users, and 30-day active users, we can feedback the correctness of the product operation strategy during the stage and the user satisfaction with product content and services.

3. Count the changes in “returning” user data

There is a "reflow" data in the analysis of different time dimensions, that is, by analyzing the users who were inactive a few days ago but are active today, the changes in users who return after 3 days, 7 days, and 30 days are observed.

4. Measure campaign quality and active user sources

By analyzing the proportion of active users among users converted from each channel and activity, the channel effect can be measured. At the same time, the proportion of active users converted from each channel and activity in the overall users can be analyzed to analyze the sources of active users.

03 What data should be collected to retain users?

How to define retained users?

Retained users refer to a group of users who have visited the product over a period of time. There are two ways to define retained users:

The first is the conventional method, where users who have any visit behavior during the stage are retained users;

The second type requires users to have certain access behaviors, such as how long they browse, how many times they visit, and which pages they visit.

For retained users, it is recommended to adopt the first definition method. Any visit behavior within the stage time means that the user still has an impression of the product, and there is a possibility of activation.

What is the value of retained user analysis?

The statistics of retained users are intended to measure the user scale of the product, while retained user analysis for refined operations requires finding ways to improve user retention through retained users. There are three key areas for retention analysis:

1. Discover the characteristics of retained users

Based on user behavior characteristics, such as browsing time, coupon receipt, and other group or special behavior characteristics, analyze the number of retained users under different characteristics, find out the high retention characteristics and low retention characteristics, and find out the value of improving retention based on one feature.

For example, by analyzing the characteristics of retained users through the length of time they use the computer, we can find out what the differences are between retained users who use the computer for 10 minutes, 15 minutes, 20 minutes, 30 minutes a day, and so on. We can also find out the time dimension that has a more significant impact on retention. For example, if the retention rate of new users who use the computer for 15 minutes on a certain day is significantly increased by 20% compared to the retention rate of users who use the computer for 10 minutes, then the user usage time of 15 minutes should be used as the standard for improving user retention rate.

2. Discover the user retention characteristics of channel conversion

By analyzing the retention rates of users converted through different channels, we can discover the factors that influence channel retention and determine user preferences and product demands for different channels.

3. Analyze session loss nodes

The session loss node refers to the user's exit page during a visit, which is counted as a session loss. By analyzing pages with high user loss, we can identify product factors that cause user loss and optimize related content.

04 Conclusion

The purpose of data statistics in the three stages of new additions, activeness, and retention is to provide real data feedback on the APP in the user experience. By subdividing the statistical indicators of each dimension, we can find the behavioral characteristics after new additions, the active characteristics of active users, and the retention characteristics of retained users, thereby improving the data performance in the three dimensions of new additions, activeness, and retention.

Source: Zhang Xiaohuai