What is growth hacking? Let’s start with the official definition

Growth hackers are people who use data to drive marketing, market to guide products, and technical means to achieve growth goals.

The term, coined by Silicon Valley entrepreneur Sean Ellis, refers to a method of user growth. To put it bluntly, it’s trying to find a smarter way to answer the mystery of how products grow, and make it a long-term mechanism for product growth.

Growth hackers are colloquially described as a status character somewhere between geek, inventor and nut

Have to mention the AAARR transformation funnel model

In this funnel, a part of the imported users will be lost in a certain link, and the rest of the users will continue to use to reach the next link, in layers of in-depth transformation. Growth hackers work around this funnel model.

Acquisition, Activation, Retention, Revenue and Referral will be followed by an analysis of each layer.

Get the user

Cold start is originally a technical term in the field of data mining. When applied to product operation, it refers to the process of introducing the first users and producing content from scratch when the product has not formed a complete ecosystem and can provide enough consumable content at the beginning

Search Engine Optimization (SEO, Search Engine Optimization) is the use of Search Engine sorting rules, through artificial means to intervene in the target page ranking, it can get more natural traffic, drive website self-growth.

Seed users determine the tonality of the product

The reason why it is called seed user includes the following three meanings:

  • They are small in number and difficult to acquire;

  • They enter the product stage early;

  • Seed users are as hopeful and malleable as seeds

For example, The early development strategy of Zhihu is the elite route. It adopted a strict invitation and review mechanism, such as Li Kaifu, Xu Xiaoping, Lei Jun and other Internet giants and a group of investment circle, media circle practitioners became its seed users. This lays the foundation for the tonality of Zhihu and makes it become a topic product in the circle quickly, which is hard to find.

In order to filter out high-quality early content contributors and maintain a healthy ecosystem in the community, Bilibili uses a set of answer mechanisms to block the less “experienced” anime fans and filter out the talent who really understand and work

The activation

Enabling common policies

ABTEST

The so-called A/B test is simply to provide two different alternative solutions to the problem you want to investigate, and then ask some users to use plan A and the other users to use plan B, and finally determine the optimal solution through data observation and comparison.

The basic idea of A/B testing includes: providing two schemes for parallel testing. There is only one variable between different schemes, and other interfering factors are excluded. Judge the result by some standard, screen out the optimal scheme.

retained

A product must have lost users, which is inevitable in market competition. However, the proportion and trend of lost users can indicate the ability and competitiveness of the product to meet users.

Converted into cash

Spread the fission

Viral communication is a way of communication based on users’ demands. Different from the previous “centralized” one-to-many broadcast communication path, everyone is a node in “decentralized” viral communication, and the effect of communication can be multiplied geometrically and shake the cognition of potential audiences.

Two core indicators of virus transmission:

  • K Factor
  • Virial Cycle Time.

K factor, which measures the spread of the virus. The formula is :K factor = infection rate X conversion rate.

  • The rate of infection is the degree to which a user spreads a product to others

  • Conversion rate is the percentage of infected users who convert to new users

The virus cycle is the time it takes from the time a user sends a virus invitation to the time a new user completes the conversion (click to read, register, or purchase). The shorter the virus cycle, the better.

Spread fission — Bug marketing

Forum, micro blog, QQ group appeared a widely spread news: “Baidu cloud network disk payment system suspected a major bug, the price of all paid packages has become the original 1/1000, 1 yuan can buy a year membership, 100GB highest grade package is only 5 yuan, quick to rob welfare ah!!” For a time netizens fried pot, no matter before baidu cloud disk account, now all hemp slip to rush to buy.

Similar BUG marketing includes the 100 yuan coupon event of Pindoduo before and the current popular livestreaming with goods marked at wrong prices, etc., all of which capture the users’ psychology of picking up bargains

Spread fission – take advantage of marketing

Taking advantage of marketing is rubbing heat, ** hot means topic degree, indirect = flow. ** hot take advantage of the situation to play well, the brand is like a kite on the tuyair, a small amount of creative cost, can fly exposure peak, how a “cost-effective” return on investment

Mercedes CEO resigns to take advantage

In May 2019, Dieter Zetsche retired as CEO of Mercedes-benz. BMW, Mercedes’ rival, released a tribute AD while it was seen as a routine personnel change.

Spread fission – user psychological grasp

The Internet is the testing ground of human nature. If the viral communication strategies of Internet products can grasp some characteristics and commonalities in group psychology, they can stimulate people’s communication power from a deep level. User psychology can be summarized into the following eight key words

Liking, profit seeking, reciprocity, asking for help, showing off, scarcity, fear of loss, laziness

How to identify business core metrics

After analyzing the AAARR model, how do we put it into practice? How do we define business value and experiment direction? All the work of growth hackers is based on the guidance of data analysis, and the definition of data indicators for different products should be based on the characteristics of the category and the core value of the service it provides.

Speak with data

On the premise of correct data sources, data analysis methods can be divided into qualitative analysis and quantitative analysis.

  • Qualitative is the feature that we can show through the data, describe the mind of the users served by the product, and clarify the direction of operation
  • Quantification is to speak with actual data indicators to reverse verify whether our operational strategies, user portraits and mentality are accurate

Data analysis is the process of combining qualitative analysis and quantitative analysis and verifying continuously. Come up with a hypothesis, design a plan, analyze the data, test or disprove the hypothesis, and finally get closer to the truth

Data are mutually verifiable, and being sensitive to data can often lead to new opportunities, such as bad weather or important local events, if travel orders increase. If the single volume of baby carriages on e-commerce platforms increases, it is highly likely that the sales of milk powder will also increase.

With the theory, with the model, should the growth experiment be done?

According to the principles of scientific experiment, design the experiment and measure the result, if the experiment is correct, put the hypothesis into practice, if wrong, modify the hypothesis, conduct the next experiment, and so on.

Growth process

Rather than routines and techniques, a flow of experimental growth path is efficient and feasible

Building a growth model

Growth model decomposition case

Input variable: Total active users

Output variables: visitor traffic, number of new user activations, number of existing users, retention rate of existing users

The latter

Growth is a long way to go. That’s all for today