User Growth - 21 Must-Know Analysis Models

Share the “21 Models Chief Growth Officers Must Know” from the book. The philosophy of fully igniting product iteration, user and revenue growth! It contains a lot of useful information for corporate managers, marketing, Internet operations, product managers, customer service, analysts, and engineering research and development.

Redefining growth

Traditional Funnel Model

The first step of traditional marketing is to occupy the user's mind and increase the user's brand awareness. In the past, all large-scale advertisements, television, newspapers, and portal websites had to occupy the minds of users. The second step is to increase user interest. In the third step, users begin to evaluate: which one should they buy among so many brands and products? Step 4: Purchase and trade. Next, some of the users who have made purchases will become loyal users and recommend the product to other users. This is the traditional, classic funnel model. This kind of thinking is not a growth-oriented thinking, but a convergent thinking, a funnel-type thinking, and there is a division relationship between each step.

Pyramid Model Growth Thinking

Today, we hope to introduce a completely new mindset, a growth mindset. This thinking is very simple, it is a pyramid thinking. Let’s review this growth mindset. It starts from a core value point to cultivate our loyal user base, and then continuously cultivate more core user groups by increasing the first user experience. Then, we let these users who have their first experience come back to experience our products repeatedly. After they form habits, we find those users who are willing to actively use these products and provide them with targeted marketing and services.

The three stages of business growth

The new growth mindset is based on high-quality products and services. In the first stage, we must use technology and products to drive our earliest core growth.
In the second stage, we can enter the stage of growth driven by performance marketing. The most important thing we need to focus on in the second stage is the relationship between cost and benefit. If your cost is greater than the benefit, such growth is vicious growth and such a business will definitely collapse.
The last step is growth driven by brand marketing.

Business growth life cycle

For an enterprise, its growth cycle is divided into five stages:
Phase 1: Problem/solution matching period, mainly to investigate customer needs.
The second stage, the minimum viable product period, seeks several methods to establish a minimal product.
The third stage is the product and market matching period, which improves user stickiness and experience.
The fourth stage is the period of channel and product matching, during which large-scale investment is made in high-priority channels.
The fifth stage, maturity, mergers and acquisitions, internationalization and localization.

The four-step data-driven growth approach

Four-step growth method

The data-driven growth methodology consists of four steps: establishing priorities, setting SMART goals, breaking down and executing, and evaluating results. Faced with a complex and ever-changing market competition environment, it is crucial to quickly and accurately find the right growth breakthrough. Establishing priorities is the first step to growth. Models 6 and 7 provide detailed information on how to prioritize growth execution. Next, regarding how to set growth goals and specific data indicators, we recommend Model 8 SMART principles and Model 9 OSM model. The next step is to break down the growth tasks and execute them, and then evaluate and optimize the results. At this time, we recommend the lean learning engine model (model 10) which is very popular in Silicon Valley.

Differentiated competitive advantage analysis framework

Among the many product lines/product features, which one is the core competitive advantage of your company/product? When opening a new product line/product function, how can you find your own differentiated competitive advantage? The nine-square model of "value-customer satisfaction" is a very good reference. The numbers 1, 2, and 3 represent low, medium, and high levels respectively. 3.2 means that our business/product provides higher value than the industry, but customer satisfaction is the same as the industry. The 3.3 in the upper right corner means higher value and higher customer satisfaction, which is our core competitive advantage.

ROI Decision Framework

Among many growth options, how to choose the optimal growth direction involves the issue of priority sorting. Sean Ellis, the father of growth hacking, once proposed the ICE framework, which scores projects from three perspectives: Impact, Confidence, and Ease, and then prioritizes projects. Its essence is based on the ROI decision-making framework of "input-output". In the model below, the numbers 1-9 represent the priority of the project, with 1 being the highest and 9 being the lowest. With limited resources, we can carry out growth experiments in the order of 1-9.

SMART principles

The SMART principle is a very useful model for choosing specific growth indicators. SMART is a combination of the first letters of five words: specific, measurable, achievable, related to big goals, and timely. Based on these five principles, reasonable growth indicators can be found. Taking the number of users as an example, the "daily active users" indicator is obviously more focused and more commercially meaningful than the "registered users" indicator. Taking Internet financial products as an example, the "cumulative investment amount" is obviously more meaningful than the "number of investing users". Indicators are never just numbers; they represent management’s understanding of the relationship between user value and company success, and they also guide every grassroots employee’s decision-making and execution in their daily work. Picking the right growth metric is strategic to a company's success.

OSM Model

The OSM model is a methodology for selecting core growth indicators of an enterprise. It summarizes the selection process into three steps. The first step is to confirm the business objectives (Objective) of the company's business/product functions; the second step is to confirm the business strategy (Strategy) adopted to achieve the above business objectives; the third step is to select reasonable metrics to measure the effectiveness of the strategy and reflect the achievement of business objectives. The OSM model builds a KPI indicator system based on business needs and implements it layer by layer. Business goals are passed down from one level to another. You need to confirm your core business goals with your superiors; then discuss business strategies with your business unit leader, and finally confirm the selection of specific metrics with analysts.

Learning Engine

"Learning Engine" is a lean operation method advocated by the book "The Lean Startup" and has been widely adopted by large and small companies in Silicon Valley. When we have an idea, we can build it using the minimum viable product (MVP) approach. After the product is launched, we need to measure the reactions of users and the market. By analyzing the collected data, we can verify or overturn our previous ideas, thereby continuously learning and optimizing.

Growth team practice

LinkedIn Growth Team

LinkedIn's latest growth team currently has more than 100 people, divided into three major groups: core growth, internationalization, and data products. Each group has its own core tasks. The head of the internationalization department is responsible for acquiring users in emerging markets to the greatest extent possible, the core growth department is responsible for user growth, and the data product department provides quantitative support: each department performs its own duties and cooperates with each other to jointly promote LinkedIn's rapid growth.

Facebook Growth Team

Diversity in team members is an important feature of Facebook’s growth team! Facebook's growth team has all kinds of people, including data analysts, product managers, and technicians. They think about various user growth strategies all day long. From search engine optimization (SEO) and App Store ranking optimization to the decision to enter the Chinese market, these are all decisions made by this growth team. The acquisition team is also part of the growth team, and Facebook’s acquisition of Instagram was completed by this team.
The head of Facebook's growth team reports directly to Zuckerberg.

Dianrong.com Growth Team

Dianrong.com’s growth team structure is similar to that of Facebook, with the vice president of product serving as the head of growth. The Growth Strategy Group is mainly responsible for indicator decomposition, and they will play an overall coordination role in the growth team. The mission of the Growth Service team is to develop and build all service products, such as user portrait system, data tracking system, SMS trigger platform, etc. Core Growth is the largest executive team, and they will implement all forecast results, growth ideas, etc. into the product. The Operations department maintains and grows through some daily operations and promotional means. The task force is similar to a special forces team. In certain areas where innovation risks are a concern, Dianrong.com will select a small team to conduct trial exploration as appropriate.

Data Analysis Practice

Big Data Analysis Framework

Users are the source of big data and also the ultimate destination of big data. In this framework, the basic framework of data analysis extends downward, starting from Customers and ending with users.
In practice, we have also found that in this big data analysis framework, more time is taken up by those close to the bottom layer, while decisions at the top of the framework take very little time; in terms of value frequency, the top layer has low frequency and high value, while the bottom layer has high frequency and low value. Therefore, an important development direction of big data technology is to achieve full automation of the inefficient parts and realize one-stop big data services.

Growth test method

Growth requires a process, and its framework is actually very simple; this methodology has been around for many years. First of all, we need a clear goal, which is very important. This is very important. Then, based on our goals, we analyze the situation we are facing and find out what problems we have. Then we come up with our ideas, ideas that can solve the problems we face or ideas that can achieve our goals. Next, prioritize which idea to test first and which to test later. Finally, we start testing to verify or overturn our ideas through experiments. Then start a new round of analysis, hypothesis, optimization, and testing to achieve growth through continuous optimization.

User behavior data analysis process

GrowingIO pioneered the "collect-look-think-do" user behavior data analysis process, which systematically connects data collection, data monitoring, data analysis, and data operations to form a closed loop of refined operations.

User growth practices

Funnel Model

The funnel shape requires infinite deconstruction. First of all, we need to be clear about what problem the vulnerability is solving and fully deconstruct the funnel based on this. For the transformation from A to R, we can split A into A1, A2, and A3, and then see which step is the main problem. Assuming it is A2, we can split A2 and look at the main problem among them. The entire funnel process of APP push is shown below, which can be broken down into 3 major steps and 10 minor steps. Ideally, we would be able to deconstruct to the granularity of a unique variable. Then we can accurately locate and solve this problem, bringing user growth. If the funnel we use is a very rough funnel, it will not solve the problem. It is necessary to deconstruct and identify the problems step by step, and then solve them, so as to bring about effective user growth.

MVG Minimization Growth Test

The head of the growth team designs the organizational structure of the core growth team according to MVG (minimizing growth experiments). There are connections between small MVGs, and the connections between MVGs connect the team members together. MVG can help everyone on the growth team clarify their own work, thereby maximizing the team's growth effect. Dianrong.com divides its growth methodology into four steps: "Define Growth - Minimize - Measure - Maximize", and drives growth through rapid iteration and experimentation by the growth team.

Growth channel selection model

Once your core metrics are identified, you need to define your growth channels by asking yourself some basic questions about user behavior. For example, do users search for solutions? If the answer is yes, then SEM/SEO is a good channel. For LinkedIn, search engines are a big channel. Because on LinkedIn, you can easily find people who have made great achievements in various professional fields, so people will search for relevant people through SEO and then enter the LinkedIn website. For example, do users share your product through word of mouth? This question will determine whether you need to spend time and energy on virality and recommendations. If users are willing to share your product, then you need to make the user sharing experience as smooth as possible. Of course, the above questions cannot represent all, but they represent a way of thinking about finding growth channels.

Promotion channel performance measurement model

For any growth experiment, we can measure it using quadrants; it is universal and simple to understand. Most importantly, one quadrant chart can present three indicators, and therefore can reflect the three core growth indicators of any MVG, and can be used for everyone’s work reports. The picture below is the performance measurement of customer acquisition channel promotion. The horizontal axis reflects the ROI of the channel, the vertical axis reflects the unit customer acquisition cost, and the size of the circle represents the scale of revenue. With this chart, you can see the effectiveness of each channel very quickly. Obviously, the channel towards the lower right corner is better because the customer acquisition cost is low and the ROI is high. We would certainly like all circles to be in the lower right corner, but this is not possible when initially launched. What we need to do is to move all the circles to the lower right corner through optimization. Every growth experiment actually does this.

User growth data analysis framework

The AARRR model corresponds to business activities and key indicators to form a systematic user growth data analysis system. Taking attracting new customers as an example, attracting new customers corresponds to the company's marketing activities, which implement key indicators, including but not limited to the number of new visiting users, bounce rate, source, etc.

Author: Thirty years of working life

Source: Thirty Years of Workplace Life