My Primary Growth Map

The Internet has developed to this stage, and the early stage of wild growth has passed. All companies have begun to seize users in the existing market. To put it bluntly, they are trying to seize the time users spend using their own apps within their limited time. Therefore, many companies set up dedicated growth departments, hoping to drive product managers to develop more focused growth strategies through more specific business indicators, so as to further expand the user base.

2019 was a year of rapid growth for me. Under the leadership of my leader, I have been engaged in product work in the direction of growth. After reading "Silicon Valley Growth Hacker - Practical Notes", I combined what I have done and compiled a growth map as I understand it.

1. Dismantling the target

If you are specifically targeting growth, the goal must be very clear. Generally speaking, the growth target is the growth scale of a certain business indicator. For example, the music category may use music playback volume as a growth target, and the e-commerce category may use GMV as a growth target, etc.

Once the goal is clear, it needs to be broken down into more detailed indicators. How to split it? Based on our existing understanding of the business, we know clearly which detailed indicators will affect this goal.

For example, GMV = number of users x conversion rate x ARPU. That is to say, there are three ways to increase this indicator:

1. Increase the number of users, which can be further divided into new user growth , old user retention, and lost user recall;
2. Conversion rate, which is further divided into the conversion rate of each node and page in the specific conversion path;
3. ARPU, or average revenue per user, is further broken down into total revenue, paying users and other indicators.

After breaking down the growth target, we found that improving any indicator will lead to changes in the growth target. Therefore, we need to improve each indicator in order to improve the growth target.

2. Three important stages

The classic user life cycle model consists of five stages: customer acquisition, activity, retention, payment, and self-propagation.

I think the first three stages are the most relevant to growth, which can be converted into three specific indicators:

1. Newly added: number of newly added devices;
2. Conversion rate: the conversion rate that reaches key behaviors;
3. Retention rate: the percentage of active users who remain active.

These three indicators directly determine the real user scale that experiences the value of the product, that is, the user scale that brings revenue opportunities to the company.

1. Added

First, we break down the channels for adding new users, which can be roughly divided into Apps, WAB pages, PC web pages, and mini programs.

(1) App

General download sources include App Store, information flow advertising, brand placement and other sources of traffic.

• App Store: The marketing department is responsible for the exposure of search results and recommended positions in the app store, mainly by constantly adjusting the bidding of keywords to control the customer acquisition cost within a reasonable range.
• Information flow advertising: The marketing department is responsible for connecting and placing advertisements to direct target users to the download page and generate download behavior. Here the product may need to optimize and test the download page. Improve conversion rate from download page to first launch.
• Brand placement: The brand department or marketing department conducts brand exposure. The goal is not necessarily to download, but it is necessary to observe the portrait characteristics of brand placement users through data analysis and adjust the placement materials in a timely manner. That is, the delivery of materials will affect the new user portraits.
• Traffic from various terminals: This includes various types of traffic, such as official accounts, mini programs, wap pages, etc. Product managers are required to check whether the traffic flow is natural and smooth, and optimize it based on the specific conversion rate.

(2) WAP page

Top: BOSS Zhipin brand area; Bottom: long-tail keyword search results

WAP pages mainly appear in search engine results and are generally divided into two user scenarios.

One is that users search for "brand words", such as: "Taobao", "BOSS Direct Recruitment", "Tencent Video", etc. This type of user has a clear understanding of the brand, so the landing page may be the wap page homepage, official website homepage, or download page. It can guide users to download the App and provide them with better service experience. SEM can be done under the guidance of the market to improve the position of the brand landing page and get more exposure.

The second type is users searching for specific long-tail search terms, such as "electric toothbrush", "Beijing waiter job", "Joy of Life", etc. This type of user has clear needs, such as what he wants to buy, what job he wants to find, or what content he wants to watch. Therefore, the landing page should be relevant to this content. For example, a product list of electric toothbrushes, a list of waiter positions in Beijing, or the playback page of Joy of Life. This type of WAP should focus on optimizing the experience, increasing the conversion rate of users to core business indicators, and then consider download guidance at the appropriate time. The traffic of long-tail words is far greater than that of brand words. Therefore, we should focus on SEO, cover more long-tail words, and improve the ranking of long-tail word results.

(3) PC page

Due to business differences, different products attach different importance to PC. For a pure mobile business product, the PC page may just be a brand face, and its function is to enhance users' brand awareness. Therefore, it is OK as long as it is simple and elegant. If it involves PC business, then the PC page must consider downloads or business conversions. How to make the information and structure clear to users and easy to operate are all optimization points that the product needs to consider.

(4) Mini Programs

Currently, more and more companies like to develop mini programs for two reasons: one is that mini programs are lightweight enough, with a shorter startup link and no loss of download and installation; the second is the natural WeChat traffic pool, which does not rely much on promotion and tests the product experience more.

Mini Programs have huge traffic potential. If you have the resources, you should cover Mini Programs as much as possible, mainly through three important channels: WeChat, Baidu, and Alipay.

Search for "waiter" and you will be directed to the mini program

WeChat : The most basic, stable, and mature source of traffic. After creating a WeChat mini program, you can connect to the WeChat mini program tool as much as possible to further expand traffic. If you connect to SouYisou, you can bring in traffic from WeChat SouYisou results. Access subscription messages to increase the mini program’s ability to reach users.

Baidu : Baidu App has over 100 million users and has a natural advantage in traffic distribution through search. After creating a Baidu mini program, users can directly access the mini program through search results. Compared with M Station, the functional experience will be better.

Alipay : Currently, Alipay focuses on increasing user stickiness by improving life services, and therefore provides a lot of traffic support to mini programs. If your business is related to life services, you can connect to the Alipay mini program to obtain traffic support and further increase the scale of traffic.

Quick Apps : Quick Apps are small programs jointly developed by Android manufacturers that can be launched without installation in the app store. However, after actual use, the traffic is not high, so if you have idle development resources, you can try to access it.

(5) Others

Fission activities will also bring some new volume, but this year WeChat has banned H5 links that induce downloads in the form of various benefits. Therefore, most products currently put the induction in mini programs. However, the download link of the mini program is very long, so the previous classic fission gameplay is not very applicable to the WeChat ecosystem.

If after testing, the customer acquisition cost brought by methods such as old customer acquisition and fission is lower than the market launch cost, then this method can be used as a regular function. Here we need to observe the impact of the marginal effect , that is, the more people come and the more people know about it, eventually this gameplay conversion may gradually decline.

2. Conversion

After the traffic comes in, you need to consider how to get users to register, leave their mobile phone numbers and experience core functions, that is, registration conversion and business conversion.

(1) Registration conversion rate

From the company's perspective, only by obtaining the user's mobile phone number can the company permanently preserve the user's usage records, formulate recall strategies based on the records, and continue to reach the user. Therefore, how to let users log in actively is very important.

There are two types of registration: one is mandatory registration and the other is non-mandatory registration.

Mandatory registration means that users must register and log in as the first step to using the product, such as social products. After registration, you need to leave basic information, such as nickname, gender, age, etc. The guidance at this time is actually also a funnel. It is necessary to optimize, adjust fields, order, etc., and finally determine a plan with high conversion rate. From the user's perspective, users need to be made aware of the meaning and value of filling in this information so that they will be more willing to fill it out.

Non-mandatory login means guiding users to register and log in during the user process. There are mainly two types. One is blocking type. To put it simply, if you don’t log in, you may not be able to fully experience the core functions. For example, on most e-commerce platforms, you can browse products normally, but when placing an order, you need to fill in the consignee and address information, which requires you to register and log in.

The other is guided login, which you can use without logging in, but if you log in you will gain privileges and an improved experience, such as high-definition privileges when logging in to video products. This type of login requires sorting out login clues and enriching login guidance scenarios.

(2) Business conversion rate

Business conversion means that users have fully experienced the core functional process and provided business value for the product. At this point, the user truly achieves value exchange with the company. Users solve their needs through product functions, and the company generates business data and obtains certain commercial value through users' stay, usage, and time. Therefore, the most important thing is how to enable users to meet their needs faster and smoother and convert them into business.

E-commerce companies pursue GMV, video websites pursue playback volume and membership income, and recruitment websites pursue resumes and the number of job applications.

Here are my three main ideas:

a. Use path decomposition optimization

Start → Homepage → List page → Details page → Transaction

When using product functions, users may go through nodes such as the home page, list page, detail page, and transaction page. The information, structure, and functions displayed at each node will affect the user's decision at that node, that is, whether they will proceed to the next node or give up directly. Therefore, when optimizing, we should focus on the user at each node, the scene they are in, and the information they are concerned about, optimize the current page information, optimize the current page functions, and make it easier for users to enter the next node, that is, optimize the current node conversion rate, which affects the overall conversion path effect. This ultimately affects business conversion rates. A small mistake can lead to a big mistake. The closer the optimization node is to the front, the greater the impact.

b. Accurate user profile diversion

Left: Meituan Right: Ctrip

For large platform products, due to their rich service content, when users first enter the homepage, all service types need to be displayed and accurate diversion and conversion need to be carried out based on user needs. This enables precise conversion.

When users enter from different entrances, the core order process is different and is designed specifically according to the user's usage scenarios, thereby improving the order conversion rate.

Tencent video tag

Hook tag

Another form of traffic distribution is Tencent and Lagou. The content format is the same, but the specific subdivision types are different, so they are divided into multiple tags to facilitate user switching, and users can adjust their priorities to a certain extent. This can satisfy the scenarios where different user portraits are looking for content of interest. Accurately segment traffic and enrich the ways users access content.

There are actually many ways to improve business conversion. The App iterations and product experience optimizations we have seen so far are basically centered around improving business conversion goals. Because of the optimized experience, business conversion will naturally increase.

3. Retention

Retention is arguably the most important metric to optimize, because it directly reflects the user value your product provides. And it determines whether the product can maintain lasting vitality. Here I divide it into three paths.

(1) Enriching the reach scenarios

When a user comes to a product for a single need, he may leave directly after his need is met. However, if someone comes because of demand A, and sees function B when leaving, then they may come back again next time because of demand B, thus forming a habit of using the product.

Meituan, for example, initially only provided life services, but later entered the high-frequency food delivery field, directly covering a wider range of users. Then, based on the food delivery field, it expanded the food delivery scenarios to include supermarkets, flowers, breakfast and other categories. By extending product categories and increasing coverage scenarios, user retention rate can be improved.

(2) Add push gripper

We are annoyed by push because most of the push messages do not hit our key points. I don't want to read the news at the moment, but you push news to me. Of course I don't like to read it. If someone sends me a private message, I will definitely read it. If I actively subscribe to NBA news, I will also watch it as long as the timing of the push is reasonable. Whether it is chat messages or subscription content, they are essentially hooks that allow content to reach users. After users close the App, these hooks can actively remind users to open it again, thereby improving retention.

For mini programs or official accounts, WeChat provides template messages and subscription messages to increase the possibility of content reaching users. These are tools we can use.

(3) Traffic Interconnection

Apps, mini programs, official accounts, M sites, private domain traffic, the traffic from these channels are all your users, and they are all using your services. If you can form these traffic into a complete drainage chain, you can retain users to the greatest extent.

For example, when a user launches the App, we can let him join the VIP group. After the user is added, you not only get the user's WeChat account, but also form your own private domain traffic. In private domain traffic, the services of mini programs or M stations can be directly delivered to users, thus forming traffic reflux again.

If users come through official account articles and follow the official account, they also have the possibility of converting to the mini program. The advantage of the mini program is that it is easy to use, but the retention rate is poor. However, private domain traffic makes up for the poor retention rate.

As long as the user has not left the group, it is your private traffic pool, and you can send messages to reach him or her and recall him or her at any time.

(4) Refined recall

For lost users, you can usually reach them and recall them via text messages. The specific recall effect will be affected by SMS copywriting, link landing page and other effects.

At this time, we need to cooperate with operations to conduct continuous testing. If we have a specific portrait of the user, we can carry out targeted and refined operations. For example: For price-sensitive users, the copy can include some small favors or promotional information to attract users. If the user prefers watching movies, why not try recommending information about recent popular movies? If the user is a job seeker looking for a waiter job, why not try recommending a link to nearby waiter positions?

Recall optimization is an ongoing operation. Different copywriting and link content will be derived based on time and user portraits. Only by continuously conducting A/B tests can the highest conversion effect be measured.

The above, regarding new additions, conversions, and retention, is my review of the past year. Welcome everyone to communicate~

Author: Wang Weihua

Source: Yeman Products (learnerwwh)