How to use data analysis to drive user growth?

Using data to gain insight into users and understanding users is the basis for growth. There is no doubt that the process of growth is also a process of data mining and analysis. So how can we use data analysis to achieve growth? Let’s take a look.

Question 1: To what aspects of user growth can data analysis be applied, and how should a growth model be established?

The eternal theme of a product must be growth, and behind this growth there must be data support, which is what we call data analysis. I divide all the growth directions on the market into three major schools , namely marketing school, experimental growth school and technology school. Below I will discuss in detail the role of data analysis in these three growth directions .

Marketing faction : To sum it up is channel operation and market operation spending money to buy traffic. In this link, don’t think that spending money to buy traffic is called growth, anyone can do it. But how to get more, better quality and more accurate traffic without spending money or spending less money, that is called growth. In this stage, data analysts are responsible for building channel evaluation models and anti-fraud models to monitor channel quality, guide channels or bosses, and ensure the survival of the fittest in traffic.

There is a saying on the Internet that 40% of the traffic is fake. Believe it or not, I believe it anyway. Therefore, it is particularly important to make good use of data analysis to guard the first traffic checkpoint of the enterprise.

Experimental growth school : The method that Sean Ellis mentioned most in his book "Growth Hacker" is the experimental growth school. By building an experimental model through the four steps of discovering problems, proposing ideas, experimental testing, and reviewing and analyzing, we can continuously test within a controllable cost range and understand the true meaning of growth. Discovering problems and coming up with ideas are inseparable from key indicators, which are also called North Star indicators. Drucker once said, "If you can't quantify it, you can't control it." Through data analysis, the business can be seen and measured, and then grow.

The experimental testing phase involves user bucket experiments, AB testing, etc., which are also inseparable from data analysis. Finally, when reviewing the results, you need to use statistical principles in data analysis, such as significance and confidence levels, to demonstrate whether your experimental conclusions are rigorous and reliable.

Technical people : Technical people tend to be data modelers. For example, they build a logistic regression model through historical user behavior data to determine whether the user has a strong desire to place an order and operate it. This is more of a data analysis, right?

After talking about the three major schools, we will break down growth according to the life cycle. Speaking of the life cycle, we must talk about the AARRR (Pirate Model) which is very popular but still very useful. We divide the life cycle into attracting new customers, activation, retention, monetization and fission.

New customer acquisition stage: In this stage, we will build channel evaluation model, channel anti-fraud model, and channel life cycle model. The purpose of this series of channel models is to bring in more high-quality traffic by utilizing limited resources through data analysis.

Activation stage : There are two misunderstandings about the activation stage. The first is that registration is equal to activation. Once a user registers and leaves a mobile phone number, the user is considered to be real and valid.

The second misunderstanding is to only look at the retention of new users, thinking that this indicator can reflect the user activation status. However, people often overlook an important indicator, which is the utilization rate of core functions.

Activate 2 major misunderstandings

The core function is the aha moment often mentioned in "Growth Hacker". It is about how to let users use the core function of the product in the shortest time so that users will be impressed and remember your product. Sometimes, it is because the user fails to leave a deep impression on the other party during the "first meeting" that activation fails, resulting in user loss.

Different types of products have different core functions. Take the game Honor of Kings as an example. Each game has its own unique rules, and the cost of understanding is very high. It is particularly important to use novice guidance to let new players understand basic operations. Therefore, the completion rate of novice guidance, the length of the first game and the number of games become important indicators of activation. Taking the Douyin APP as an example, the core function is to collect favorite shoes, and then you can see the price fluctuations and market conditions. Therefore, the collection rate and tool usage rate (dressing, shoe VR, etc.) of new users are important indicators of activation. Taking Yiche APP as an example, its core function is the car tool, through which you can check the lowest price and related information of your car. Then the inquiry rate of new users after using the tool becomes the activated North Star indicator.

North Star Indicator

Judging user activation is not limited to registration rate and retention rate. It is also necessary to find the usage rate of core functions as a monitoring indicator based on the product type.

To sum up, registration rate, new user retention and core function usage rate are the key indicators to determine whether users are activated . Data analysis is also indispensable from business monitoring indicators to growth links.

Retention stage : Retention is a good monitoring indicator that can provide feedback on user stickiness. However, if you want to improve retention, it is very difficult to do in practice. Why do I say this? Because the retention of a product includes five aspects and is not restricted by a single factor, let's take a look at the five major methods I have summarized to improve retention, namely channel refinement, product structure optimization, event incentives, providing high-quality services and unconventional (reach) means such as push, SMS, and in-site messages.

5 ways to retain

How are the five aspects of retention summarized? This is the embodiment of review and analysis in data analysis, and a business thinking model summarized through continuous review and analysis.

Monetization stage : Let’s take a classic model as an example. We use historical data to model the behavior of users who have placed orders or not, use a logistic regression model to predict the user’s willingness to place an order, and increase the user’s order rate, as shown in the following figure:

Not to mention mathematical models, modeling ability is also one of the indispensable skills in data analysis.

Fission stage : My advice to everyone at this stage is to build a good anti-cheating mechanism, and then think about the fission gameplay. Otherwise, it is easy to be robbed by the wool party or the virtual machine. Hundreds of thousands of dollars will be wasted in an activity, and all the newly added equipment will be fake equipment. This kind of thing is not uncommon. Therefore, data analysts need to oversee the activities and make sure that the money is spent wisely. Remember that if the fission is not done well, it will really "split".

Above, we have divided the growth into three major schools and life cycles, and talked about the application of data analysis in growth. Doesn’t it feel like there are many ways to play?

Question 2: How to obtain the first batch of seed users?

First of all, we need to know what our purpose is. For example, if we want to do car wash business, then the seed users should be the car owners, right? The second step is how to acquire more car owners and obtain car owner user data through third-party database collision, ticket information, user registration and authentication of car owners, etc. In the third step, to ensure the effect, we need to have certain screening rules to select active users, such as active car owners who have been connected to the Internet for 100 days, and send SMS and push notifications.

When screening users, the finer the granularity, the better, and fine-tune operations by segmenting by region, gender, car owner price, etc. Finally, don’t forget that sincerity is always the basis for the fission of seed users. Exclusive care and feedback from major customers can better pave the way for later growth.

Question 3: How should we start to increase user growth when our company’s internal data foundation is poor and we lack growth methods?

Poor data foundation and lack of growth methodology are two problems. The data foundation is like the basic skills of a football player, and the growth method is like the ability to score goals. If you don't have good basic skills, it will be very difficult to score a goal in the game, unless you are lucky enough to get lucky, which is a low-probability event in statistics, so let's not talk about it for now. Therefore, the top priority is to improve the basic skills, that is, to improve the problem of poor data foundation. Sharpening the knife does not delay the chopping of wood.

So let’s talk about how to solve the problem of poor data foundation. A gentleman is not born different, but he is good at making use of things. For small companies, when their own data infrastructure capabilities are insufficient, they can use third-party services. Sometimes we don’t need to step on the mines that others have stepped on.

It would take at least five people to make the tracking point and BI smart report system by yourself. It's reasonable to spend more than 2 million yuan on their salaries a year. However, if you buy a Shence Data, which includes non-tracking data collection and smart display functions, it would only cost a few hundred thousand yuan a year. Wouldn't it be better to use the money saved for advertising?

Another example is anti-fraud. Doing channel traffic anti-fraud on your own requires a very large user data base and algorithm capabilities. If a small company develops its own anti-fraud system when its business is still unstable in the early stages, wouldn't it make the technical team, which already has limited resources, even worse? Wouldn't it be safer to choose a professional team like Shumeng and Shumei? It's like if you want to eat fresh food, why not just buy a refrigerator? But you don't want to do that and insist on building one yourself. Then I can only silently give you a thumbs up and turn away.

But one thing is for sure, when your team is strong enough, some things really need to be built by yourself, such as channel attribution collection, self-built BI system for core data, etc. After all, it is safest to take your destiny into your own hands.

Therefore, for companies with poor data foundation, my personal suggestion is to choose third-party services as a transition in the early stage of the company, and then wait until the company grows stronger to truly take control of its destiny and build some data systems that can be built by itself . After all, sooner or later you will have to pay for what you have done.

Question 4: Everyone has been discussing growth in recent years. So what are the differences and similarities between the growth of Chinese products and the growth of growth hackers in Silicon Valley, USA ?

"Growth hacking" must be familiar to those in the Internet circle, especially in the past few years. Many people have also seen many classic cases from abroad. For example, Netflix analyzed the movies and programs watched by customers and found that movies starring Kevin Spacey and political TV series were very popular with users, so Netflix produced the TV series "House of Cards"; when Facebook was doing grayscale testing, it was found that the new version would reduce the monetization rate by 25%, so it urgently terminated the launch of the new version, etc. Its core concept is to rely on technology and data-driven to achieve the goal of growth.

But in recent years, everyone has found that "growth" is no longer attractive. The so-called "growth" is all other people's "growth" or cases from foreign countries. When it comes to China, it will become "unsuitable for the local environment". After all, there are no operations or channel positions abroad, right? Just look at the commercialization level comparison between APP store and domestic Android application stores such as Huami OV (Huawei, Xiaomi, vivo, oppo), and you will know that domestic growth can be said to be a hard growth, and sometimes even the old Silicon Valley is far behind.

There are two main reasons why growth is not adaptable to the local environment. The first point is that Chinese people are smart and have many ways to play, such as the commercialization level of the Android market mentioned above; the second point is that the user differences and demand diversity in the United States are relatively single. Take food for example, Americans only like pizza and hamburgers, right? Look at the food classification in China, and you will stand up instantly when you click on Meituan. Therefore, operations positions have emerged that do not exist in the United States. The complexity of the sub-sectors and the highly commercialized models are the main differences in growth between China and the United States.

What is the common point of growth? It is the core concept of growth, such as the MVP model, FRM, aha moment, etc. These concepts are eternal. It’s just like playing football in China and the United States, the rules are the same, but the physical fitness of the people is different.

Question 5: What are the common things to note for growth experiments?

(1) Cultivate the ability to see the big picture

The “poor” play MVP (minimum viable product), and the “rich” play AB testing. Why do we say this? How many people do AB testing just for the sake of AB testing, and then just choose the best solution from them? On the surface, this does achieve optimization, but have you ever thought that when we do AB testing, we are actually "frogs in the well" trying to see which way to jump higher? If you jump out of the well and make a minimum viable product, you should keep your perspective throughout the entire product and collect more feedback information at the "minimum" cost, thereby achieving global growth. If you just keep doing AB testing on a certain node, then I can only say that you are wasting resources. It is better to integrate the entire product line and do a set of MVP tests.

Here is a specific example:

MVP Minimum Viable Experiment

AB test (C): Testing of new channel materials to find the best conversion material. We have N ways to select materials, and finally we get C3, which is the material with the highest new conversion rate. Then we think that we have accomplished the task, and all channels and agents use C3 material. If we do this, aren’t we just a frog at the bottom of a well, trying which way of bouncing can lead to the highest jump? However, when we jump out of the well, we will see that new conversion is not only determined by materials, but also by the joint constraints of products, channels, technology, etc.

MVP: Growth is not the responsibility of a single department or a single link; it is the result of collaboration among all departments. Let’s take the above example again. When we jump out of the well and conduct MVP testing, the best growth plan for new conversions may be A1+B3+C3+D1 and A2+B2+C1+D3, rather than a single material C3.

This is what I mean by “poor” people play MVP and “rich” people play AB. If you spend the same amount of money, whether you play at the bottom of the well or outside the well, wouldn’t it be more cost-effective outside the well? So we must break out of our limitations and look at growth from a global perspective.

(2) Communication

The most important skill for an analyst is communication. First, understand the boss's needs, then analyze them. The results of the analysis should be translated into a language that the boss understands, so that the boss can understand the value of your actions.

No matter how well you do your business, if your leader cannot perceive it, the project is valuable but its value cannot be maximized. How to make the boss recognize the results of the project is also a communication skill. Don’t think it is unimportant. It is related to whether the project can be launched, the budget and scale of the project.

Capable analysts will find growth points, and excellent analysts will make leaders aware of the growth points, ask for more budget to continue to expand the project scale, and ultimately get better results. Finally, all team members will be promoted and get a raise.

To give a small example, if we need to find the problem points (growth points) and discover that user churn is serious, we need to carry out a project to recall lost users. Before starting the project, in order to evaluate the effectiveness of the recall project, we sorted out monitoring indicators and developed the reach-recall rate, the number of people reached by the recall rate, the recall business conversion rate, the post-recall contribution retention rate, and the N-day retention. We also know that the leader’s KPIs include DAU, next-day retention, 7th-day retention, etc., and we found that the 7th-day retention is closely related to the indicators of our experimental monitoring, so we decided to make the 7th-day retention a key monitoring and reporting indicator.

You may have discovered that although the previous indicators are likely to grow, the boss may not be able to perceive it. However, if you work together with the boss, he will be able to quickly perceive the growth. If you also have strong communication skills, then budgeting and project launch will be a piece of cake. The above-mentioned cultivation of the overall perspective and the ability to communicate upward and downward are the two points that I personally think should be paid attention to in growth experiments.

That’s all I’m sharing with you today. In the end, I wish you all know all the methods on the road to growth, and that you will never grow without success!

Author: Jiang Di

Source: Zhao Xiaoluoluo (luoluo9633)