3 Case Studies to Help You Implement Growth Models (Part 1)

With the rapid development of the Internet , its popularity has increased year by year, and Chinese people have paid more and more attention to education. The online education industry has developed rapidly. Today, the online education industry is a trillion-level market under consumption upgrading, with broad market prospects. However, competition is becoming increasingly fierce. With the maturity of Internet channels , the cost of acquiring customers is increasing year by year, but the conversion rate is sluggish. Many education companies are faced with the problem of how to acquire customers accurately and convert them efficiently.

Today, we will sort out why we need to build a growth model, how to build and optimize a growth model, and specific growth cases.

1. Why build a growth model?

First of all, everyone needs to understand the general situation of the entire education industry. At this stage, the entire education industry is roughly divided into three categories. First, K12 education and training. Second, vocational education and training, and third, quality education and training. In the past, people’s learning paths were basically all in offline training classes. Now, people can get online education through the Internet, such as MOOCs, remote live classes, etc. The emergence of new media has not only changed learning habits, but also posed new challenges to the way educational institutions acquire customers. Although they are all online education, each company in the education industry has a completely different customer acquisition and conversion model.

For example, New Oriental’s earliest batch of users were obtained by Yu Minhong by posting posters at various universities in Beijing. This is New Oriental’s offline customer acquisition model. Up to now, this is New Oriental’s biggest magic weapon: posting posters and holding lectures, accumulating a large number of users through this offline customer acquisition method, and then guiding users to online courses is a natural thing.

For example, how does Xueersi acquire its users? During the SARS epidemic in 2003, he posted on the Aoshu website, asking one question every morning and publishing the results in the afternoon. He eventually gained his earliest followers. One day, a fan asked if it was possible to open an offline tutoring class, and so we opened the first class. This is Xueersi’s online customer acquisition model. It can be said that it has Internet genes from the beginning and relies on forums to recruit students. More than a decade later, many online education institutions also did this.

Through comparison, we found that these two most well-known educational institutions in China have completely different genes and their methods of acquiring and converting customers are also completely different. Nowadays, there are many different ways to acquire and convert customers, and both online and offline are battlefields for competition. For specific companies, the most important thing is to build and optimize their own growth models based on their actual business needs and improve the efficiency of customer acquisition and conversion.

2. How to build and optimize the growth model?

In the construction and optimization of the growth model, Mr. Shi proposed a more effective growth model (R AR RA growth model), focusing on activation and retention . This is a five-level funnel, mainly including acquiring users (Acquisition), stimulating activity (Activation), improving retention (Retention), obtaining revenue (Revenue), and spreading recommendations (Referral).

At different stages of the funnel, we need to pay attention to data indicators at different stages. By analyzing these indicators and combining them with the company’s actual situation, we can optimize the RARRA growth model and improve activity and retention.

1. In the acquisition stage, you need to pay attention to traffic , search volume, CPC (cost per click), search results, customer acquisition cost, click-through rate , etc.

2. In the activation stage, we need to pay attention to the number of registered users, registration rate , guide page display completion rate, number of subscriptions, number of users who have used the main function at least once, etc.

3. In the stage of improving retention , we need to pay attention to user engagement, user login frequency, daily/weekly/monthly active users, churn rate, etc.

4. In the stage of earning revenue , you need to start paying attention to customer lifetime value, payment conversion rate, ad click revenue, etc.

5. In the referral stage, you need to focus on the number of invitations sent, the invitation page open rate, the number of invitation page registrations, the virus transmission cycle, etc.

Mr. Shi pointed out that after building the company’s pirate model through the five-level funnel, it is also necessary to refine and visualize your data. It is very important to visualize Zhuge io’s data indicators so that every product and operation personnel can understand them and continuously optimize the growth model.

2. Case Analysis: 51Talk

51Talk's full-chain model is:

From this picture, we can see the problems of 51Talk:

1. The process from "launch" to "order closing" is too long, with many key business nodes, high business risks and high churn rate;

2. The long-tail benefits of the campaign cannot be fully utilized. When potential customers are faced with the choice of "register and stay" or "leave without registering" on the "landing page" node, there is a lack of "middle road" accumulation, which makes them easy to lose;

3. The paths available to customers are limited, the interaction interface with users is narrow, and a stable, long-term, and direct information channel has not been formed;

So, after discovering the problems with 51talk, how can we optimize its core conversion process through the pirate model? First of all, we need to see what the conversion rate was for each point at that time? For example, the conversion rate of key points such as delivery, registration, and invitation, and find out which areas can be optimized. The 51Talk team visualized all paths of the entire process at that time, and then optimized the entire main process to improve the retention rate of each conversion point.

Example 1: Fully tapping into long-tail benefits through new media

In 51Talk's process from launch to order closing, the conversion rate is only a few percent. What does this mean? This means that 100 people are exposed to 51Talk, only 10-digit people experience your class, and in the end only a few people pay for it. What's the problem with this? More than 90 out of more than 100 people were lost in vain because no one promoted him, but there are potential users here.

How to retain these people? The 51Talk team has added branch processes in some places. For example, when TMK invites users and when registering, users are asked to follow 51Talk’s official service account and follow up through the service account. In addition, users are added to the WeChat group to become 51Talk WeChat testers. In this way, a large number of potential users have been accumulated in WeChat groups and service accounts, greatly reducing the churn rate.

You can develop this idea and find a way to settle 90% of users in another branch process, that is, find a way to install 90% of users and convert them in the future. The change brought about by such optimization is that users can be retained to the greatest extent possible during the first contact with them.

The same goes for other conversion points. Find ways to optimize the retention rate of each process. When the retention rate of each process is improved, the retention rate of the entire process will naturally increase.

Example 2: Increasing the paid conversion rate by adding more sales sessions

By adding a sales environment based on the existing experience class process, the overall conversion rate can be effectively improved.

As a long-standing conventional marketing method in the education industry, sales meetings are one of the means to enhance the users' comprehensive understanding of our company's brand , teaching philosophy and teaching quality. It uses the method of "multi-person live classroom" to more efficiently implant key information into the user's "cognition". The increase in sales meetings will stimulate those users whose learning needs are still vague, thereby improving the user payment conversion rate. Then compare the payment conversion rate of ordinary users and the payment conversion rate of users attending sales meetings in real time to provide data support for the formulation of the company's operating strategy.

Case 2: Lyrics

Leci is positioned as a tool for users to memorize words. The logic of Leci is to build a traffic matrix, focus on daily activity and retention, and later carried out light monetization and heavy monetization. The core of the core is the content of the lyrics, that is, the word count and accuracy. The optimization of words is the most important. Otherwise, no matter how good the experience optimization is in other places, it will be useless. For example, Douban , the high-quality content of the community is the reason that attracts users to stay. Content is the core, and only with content can there be retention.

After the lyrics team sorted out the issues, the focus of optimization over the past year and a half has been on optimizing words and improving the accuracy of words and pronunciation. On the basis of optimizing words, we further optimize the conversion points of each step in the growth model to improve the retention and activity of Leci. Here, Mr. Shi shared with us the optimization methods to improve the retention of new users and the activity of old users.

1. Improve the retention rate of new users by optimizing the path

Path to enter Leci for new users:

Leci is a product that relies heavily on traffic, and the loss at each point is cumulative. Therefore, it is necessary to improve the conversion rate of each link in order to retain users. In the path of Leci, you can see that downloading the vocabulary book is the first step. If this step is not done well, the next user will not be able to reach the next path at all. The Leci team conducted a detailed analysis of user data for users who completed registration on the same day but did not download the ci book, and combined it with user research analysis to analyze user usage habits.

After analyzing users' usage habits, the Leci team discovered the problem. In the download path, junior high school, high school, college, and studying abroad were folded together. The first screen was classification, and the dictionary was not visible, which reduced users' interest. Later, after optimization, a more convenient entrance for downloading dictionary books was added. The dictionary can be seen on the first screen. Previously, users had to click on the category and then click on the dictionary, but now they can click on the dictionary directly. This is much more convenient for users. After optimization, the conversion rate for downloading dictionary books increased from 60% to 84%, which also drove the retention rate of subsequent steps.

2. Improve the activity of old users by optimizing functions

In Leci, some old users cannot persist in the process of memorizing words. How can we let students participate in the memorizing activities to solve the problem of their inability to persist? The Leci team has come up with many ways to improve students’ sense of achievement after learning, such as adding leaderboards , teaming up with classmates, and regularly determining rankings. Through such optimization, students’ stickiness to Leci has increased, and the activity of old users has been effectively improved.

The above examples are just two specific points that Leci optimized using the growth model. After Leci completed the optimization of the entire process, starting from July 2017, new customers increased by 60% year-on-year. User retention rate on the next day: increased from 22.5% to 50%, number of days per week for memorizing words: increased from 1.7 to 3.8.

In 2017, the number of paid word books increased by 325% year-on-year

Case 3: New Oriental Online

New Oriental Online's growth model

New Oriental Online is New Oriental’s main online platform. The company’s core business model is as follows: first of all, the brand’s natural traffic, which brings in natural orders, accounts for 60% of revenue. Secondly, we make advertisements and close deals through consultation and outbound calls, which accounts for 30% of the revenue. There is a problem at this time, which is the user retention part. Natural order formation is the logic of e-commerce , and outbound calling is the logic of telemarketing. There is no retention logic in between. It's like only 10 out of 100 people make a deal, and the other 90 people need to be retained because they are all potential users. After discovering this problem, the team sorted out the company's growth model and growth system, added retention indicators to the model, and visualized all the data in the growth model to find conversion points that can be optimized. Mr. Shi shared the following conversion points.

1. Optimization of form lead collection

Example (1): Changing the form type: By changing the pop-up form to an embedded form, the form submission rate increased by 154%.

Example (2): Showing the form submission entry and adding an embedded form to the first screen header increased the form submission rate by 157%.

2. Improve user retention by optimizing WeChat subscriptions

In order to optimize the retention path, the New Oriental Online team added a user interest subscription function to the official website user registration end and WeChat service account end. Based on the user's subscribed interests, we use our self-developed WeChat backend to accurately push courses, information, promotional information, etc. According to the statistics of precision push, the opening rate of precision push messages is 10%, while the opening rate of regular push messages is 0.7%. After awakening users through precise push notifications, the conversion rate of subsequent orders reached 27%.

3. Optimize the TOEFL course selection page on mobile phones

When optimizing the course selection page, the New Oriental Online team created the course selection area from the students' perspective, first dividing the target scores and then dividing the user groups. Streamline the product line, focusing on flagship courses and supplemented by one-on-one courses. This greatly motivated students to take courses. After the revision, the quarterly revenue of the TOEFL course selection page increased by 1140% year-on-year, and the traffic increased by 179.93% year-on-year.

4. Optimize the payment process-fast payment

In the past, when making payments, you needed a bank card number and SMS verification, which was very cumbersome. Now, you can pay directly through Alipay and WeChat, which increases the payment conversion rate by reducing the steps in the user's payment path.

We can see that building and optimizing growth models and visualizing specific data indicators at each stage are crucial to business growth.

Source: Zhuge Jun