Data operation: How to use data analysis to achieve user growth?

What does data analytics mean for growth? How to do data analysis well?

1. Growth and Data Analysis

Using data to gain insight into users is the basis for growth. There is no doubt that without data analysis, growth is not possible. Therefore, it is not difficult to understand that one requirement for growth positions is to be familiar with data analysis and be sensitive to data.

It can be seen that the growth process is also a process of data mining and analysis, which is difficult to separate from data. The common working steps are as follows:

• Interpret the company's strategy and find a key data indicator, the North Star indicator
• Continuously disassemble indicators to form small single indicators one by one and divide them into different paths
• Evaluate the priority of growth experiments based on the impact of the experiments that can be done along the path
• After the experiment is launched, analyze the relevant data
• Determine the growth method based on the analysis results and output the next indicator plan

So, what problems are we using data to analyze during the growth process? What are the practical applications of data analysis? Next, let’s look at how data analysis is used at work to help products achieve growth.

2. Practical Application of Data Analysis

In the process of growth, data analysis, attracting new users, activation and retention are common practical applications, which will be explained in detail below.

1. Attract new users

In this stage, the traffic sources of APP can usually be divided into two major sources: paid and free. Generally, paid traffic is obtained through advertising, paid promotion in app stores, etc., while free traffic can be obtained through a variety of operational means, and incentive mechanisms can be designed to stimulate users to download and share.

Paid traffic:

To obtain traffic through payment, we need to conduct data analysis on channel conversion and quality assessment so as to measure the effectiveness and value of the delivery.

Common data indicators for channel conversion are CTR (click-through rate) and CVR (conversion rate). The significance of these data is that they can guide us to optimize users’ external conversions and reduce costs.

Quality assessment is to analyze the quality of the delivery channel to evaluate whether the channel is worth continuing to deliver. Each company has different indicators for evaluating channel quality, and generally they need to be designed closely in conjunction with product strategies.

Common indicators include DAU, revenue, retention rate, churn rate, etc., but no matter which indicator is used to measure, the company should pay attention to ROI, which is the most basic principle.

Free traffic:

In the process of acquiring free traffic, we often use the incentive mechanism of the event to stimulate users to complete the download, installation and registration of the APP . For example, new users can receive incentives such as coupons or points.

In this link, we can regard the process of acquiring traffic as a funnel model . By analyzing the data of conversion rate and final installation and registration volume, we can see the obstacles and problems that arise in conversion and obtain optimization paths and methods.

In addition, we can also use cluster analysis to analyze, that is, to examine the data performance of users with different portraits in different scenarios, so that we can design different new user acquisition models for different users.

2. Activate retention

In fact, compared to attracting new users, activation and retention are the real growth . Otherwise, if the new users are lost, the meaning of growth will be lost.

When it comes to activating retention, we have to mention a concept, which is the aha moment. Ahaha moments can increase product activity and help further retention.

The so-called aha moment is the point that makes users feel excited. The growth we are talking about is actually centered around the aha moment . It is like a force that drives users to love the product. Without the Aha moment, it is difficult for a product to achieve self-growth, so it is very important to find the Aha moment to make users like the product and give users an experience that exceeds their expectations.

Therefore, in the process of activation and retention, we have to use data to measure user activation, and after activation we have to analyze short-term next-day retention or 7-day retention data.

When analyzing retention data, we usually focus on the relationship between certain behaviors of users in a certain scenario at a certain time and the short-term and long-term retention of users, in order to evaluate what should be done to guide users to truly retain them.

From this we can see that we can continuously design experiments to find the key indicator that affects retention, so that we can clearly see which method is more conducive to APP activation and retention.

It is worth mentioning that in terms of user retention, user recall is also a relatively important part. At this time, we need to evaluate user channels, develop multiple user recall designs and data analysis systems, and explore appropriate recall plans from different angles.

In this process, we need to conduct data-based analysis of each operation. Only in this way can we achieve visualization and clearly see the room for optimization. For example, when using APP PUSH to recall users, we need to pay attention to the open rate of users after receiving the PUSH. If the open rate is very low, the recall is obviously a failure. Therefore, improving the open rate is an important data indicator for recalling users using APP PUSH. Then the next optimization task is very clear, which is to improve the open rate.

When designing a plan, we must pay attention to operability, that is, how specific a thing is done and to what extent a certain indicator will change . If the plan can meet such conditions, it is feasible.

3. OKRs for Data Analysis and Growth Teams

OKR (Objectives and Key Results) is used to clarify the company and team goals and the measurable key results required to achieve each goal. This will undoubtedly help us better define and track goals and manage work more effectively.

So, how do we use data to develop OKRs for our growth team? In simple terms, there are three steps:

• Establish the North Star indicator and prepare to disassemble the indicator
• Clarify the indicator disassembly path and comprehensively evaluate the priority of the path
• Confirm the path plan and make appropriate adjustments based on data analysis results

1. Establish the North Star indicator and prepare to disassemble the indicator

The North Star indicator is usually closely related to the company's product strategy and is the core part of the entire strategy. Common indicators such as DAU, GMV, ARPU value, etc. can all be regarded as the North Star indicator.

Once the North Star indicator is established, we need to break it down as finely as possible to facilitate implementation. For example, GMV needs to take into account new additions, active users, and retained purchases, so a model will be used to estimate to what extent the GMV indicator needs to be achieved. During this process, special attention should be paid to the fact that the various sub-indicators should not conflict with each other and must all contribute to the North Star indicator.

2. Clarify the indicator disassembly path and comprehensively evaluate the priority of the path

After breaking down the North Star indicator into different small indicators, we need to conduct path analysis on these small indicators, sort out what work needs to be done, and finally prioritize the results.

At this time, we must be clear about which path is relatively easy to achieve, which contributes the most to the completion of the indicators, and how much cost is required to complete it . Only by evaluating in this way can we more fully understand the practical significance of the work. We cannot blindly follow the trend, because different types of products face different problems in different life cycles, otherwise it will interrupt the normal work rhythm.

3. Confirm the path plan and make appropriate adjustments based on data analysis results

Once we have fully evaluated and confirmed the path plan, we can choose the plan that will help us achieve our goal the most. At the same time, we can also analyze the subsequent results of the plan to adjust the execution of the plan to maximize the goal. During this process, we can observe the year-on-year and month-on-month growth data, which will help us better control the completion of the indicators.

4. Data Analysis and Optimization of User Activation

As mentioned earlier, user activity and retention are the real growth. Usually in the user activation stage, data analysis can participate in the following five steps:

• Key behaviors for disassembling products
• Analyze the impact of behavior on core indicators
• Disassembly and optimization solution path
• Design
• Analyze and summarize experience

1. Key behaviors of disassembling products

Exhaust all functional points of the product, sort out the usage of functions that affect core indicators, such as frequency of use, duration of use, etc., and then list key indicators such as retention/user payment rate/repurchase rate, etc.

2. Analyze the impact of behavior on core indicators

In this process, we need to make predictions based on past data and then make reasonable assumptions, while also considering the boundaries of the data involved in the key behaviors of the product. All assumptions are based on a fixed cornerstone assumption, that is, what the major premise must be clearly explored.

3. Disassembly and optimization of solution paths

After determining the impact of different behaviors on core indicators, we need to break down and optimize the path of the behavior, select the better solution from the broken down path, clarify the priorities, and control the rhythm of the experiment.

4. Design

The MVP model is particularly useful in the design phase, allowing us to avoid investing unnecessary manpower and material resources. The MVP model is used to verify our hypothesis. We can observe whether the effect is as we expected through small-scale and small-scale experiments.

If the results after the MVP model meet the requirements, then we can expand the scale and scope of the experiment to further verify the hypothesis. During the entire experimental process, we also need to adjust the definition of user activation to always proceed in the right direction.

5. Analyze and summarize experience

After a phased experiment, we need to analyze the results, summarize the experience, review the experiment, pay attention to whether the results are consistent with our preset goals, see if there is room for optimization of the experiment, and explore what mechanism can sustain it. Growth is an ongoing process and does not stop when the experiment ends.

Conclusion

Today’s article talked about a lot of applications of data analysis and growth. The actual work is much more complicated than this, because the data is constantly changing and many factors are uncontrollable. Therefore, we need to fully master the skills of data analysis and perceive the reasons behind data changes. Only in this way can we have a clear idea of ​​growth.

Author: Sesame paste

Source: Tahini