Facebook user growth strategy!

Whenever we talk about growth, we have to mention Facebook, the pioneer. The author of this article was once a member of Facebook's Core Growth team. Now, based on his own practical experience and combined with cases, he shares with you the very valuable steps and methods for Facebook's growth.

A few years ago I was working on the Facebook Growth team. This article shares how Facebook's growth team works. I hope it will be inspiring to everyone.

The word "growth" has become very popular in recent years. This indirectly expresses the anxiety of many entrepreneurs about traffic in the stock market. In fact, the concept of growth has not changed from ancient times to the present. Under the pressure of the situation, everyone began to explore whether there is a mature methodology, trying to find a "shortcut" to growth.

Let’s take an example to explain how Facebook grows.

1. How to set goals?

The first step is to set goals.

Setting goals seems simple, but it is actually very difficult to set them well. Poorly defined goals = no goals. The core objectives of different product businesses should be different. But it must be simple and clear enough!

For example, Facebook is divided into three parts: growth, core products and commercialization . The goals of the business units and the company as a whole are clearly expressed in a simple mathematical formula:

The core indicators of different businesses are also different. For example, WeChat and MiTalk are both IM applications. Suppose 10 million people open them every day, but I send 100 messages on WeChat and only 10 on MiTalk. DAU cannot be used to measure the true level of two apps. It should be more about the amount of information sent and received. Moreover, if WeChat sets the goal as DAU, then the team members will start looking for ways to send PUSH to improve the indicators, which will mislead the direction of the business.

The next step is to quantify the goal. Set a goal that you can achieve with just a little effort .

Assuming that Facebook now has 10 million DAU, how will it set its goal for next year? Of course, you can set a target of 15 million on the spur of the moment. But if the goal is set too high or too low, it will not serve as a guide.

The target is usually set more clearly: for example, 14.5 million. Of this, natural growth is expected to reach 2 million. If we leave X% space for new users, they will contribute about 1 million DAU, and there will be about 500,000 DAU space at the recall level...all added together, we can reach the goal of 14.5 million.

2. How to break down the goal

Abstract goals are of no guiding significance. The key is to transform each goal into an executable project.

For example, you are assigned to a project to improve the retention of new users. The proposition of improving the retention of new users is very abstract, so let’s make it concrete.

To improve new user retention, we need to understand what surprises users when they arrive at this scene. For a user to stay, he must find the platform "useful" ("useful" includes interesting). So your goal is to make users discover the value of this product as quickly as possible (the so-called aha moment) . There are many methods here, and we provide a solution from the perspective of data analysis.

We divide users on the Facebook platform into two categories: active users and inactive users. We try to analyze which user behaviors/characteristics are more likely to contribute to user retention. We pulled out some common behavioral characteristics of users (as shown below).

You will find that the difference between active and inactive users is very significant based on whether or not they have friends. Among the active users, there are a lot of users who have added friends. However, among inactive users, most of them did not add friends. You seem to have found a point, and you put forward a hypothesis: allowing new users to add friends as much as possible can significantly improve the retention of the user.

So now the proposition is from "improving user retention" -> "letting users add friends as quickly as possible". There seems to be some direction.

At this time, you have to continue asking yourself. In fact, the platform does have a function for adding friends, but it may not be good enough for new users and not enough resources are allocated. So how do you decide how many resources you need? We need to quantify the benefits as much as possible. You have to continue to ask yourself a question: How many friends does a new user need to add to achieve stable high retention?

Let's draw a picture: (the horizontal axis is the number of friends, the vertical axis is retention)

We will find that the marginal benefit of adding friends is decreasing, and it will reach about 10 friends. Users can achieve relatively high retention. Our goal can then be further refined:

Therefore, the goal of the new user retention group changed from the initial vague goal of "improving new user retention" to a very specific goal of "letting users add 10 friends within 7 days." Of course, we can further subdivide it here, as long as everyone can understand the idea. The models I have outlined here are relatively simple; actual analysis will be more complex and rigorous. The overall idea is to transform a relatively vague goal into small executable goals. Finally, the model looks like trees. The deeper each branch grows and the more nutrients it contributes, the faster the whole thing can "grow".

3. Trading and Review

There’s a saying in Silicon Valley: “What you don’t measure you can’t grow.” A good growth team must be a team that can fight and withstand numerous experiments.

Now we have a goal: "Let new users add 10 friends within 7 days." Then everyone will have a lot of ideas: Should we recommend friends during the registration process? Are more friends’ resource positions recommended to new users in the information flow? Do you want to add more new users to the recommended friends list? How to use PUSH/SMS? etc…

How to prioritize projects? Ensure quarterly targets are met?

At this time, the project leader will gather all these ideas together, combine past experience and data analysis, and divide the projects into two categories: one is a "deterministic" project. The other type is the "exploratory" project. Deterministic projects are those that can generate deterministic benefits to the goals and generally occupy more than half of the team’s resources. The other type, “exploratory”, has a higher degree of uncertainty, but once a new “oil mine” is discovered, it can bring a steady stream of profits in the next few quarters.

Therefore, planning will be like a multi-branch tree. If you find that a branch of the tree can find a water source, you will continue to open up this branch. If you try a new direction and find that the benefits are not great, give up this direction immediately.

Facebook has a set of efficient and practical AB experiment platforms and data analysis tools. For each experiment, people from all parties can quickly measure and evaluate the specific effects of the experiment from various dimensions. Each analysis strengthens the team members’ understanding of users and products. Over time, the success rate of the experiment will increase significantly. A lot of accumulated experience lets everyone know that the road is open. Supplementing the overall "decision tree" will form a positive cycle of continuously improving efficiency among team members.

In the above example, for example, we found that the "recommended friends" module had a significant improvement on 10 friends in 7 days. We will use this as a fulcrum to continue to branch out more fulcrums, such as "providing greater traffic at different entrances", "recommending more friends to new users in the information flow", and so on.

Get more information through AB experiments. For example, for the branch of "recommending more friends to new users in the information flow", we will further analyze the balance between traffic costs and advertising revenue; further refine the different stages of new user groups; and make differentiated strategies based on the cultural attributes of people in different regions. These become small branches, and when the water source of this branch is completely drained, we will switch to another branch and continue exploring.

4. What does it mean to you?

“Growth” is more of a mindset than a replicable methodology. If we have to summarize it in one sentence: it is to improve the efficiency of system iterative evolution.

Many people would say that we can’t actually do this: because we don’t have XX system, our model is to XX, and our scale is too small. To be precise, copying homework will definitely not work. It is more about looking through the surface and seeing which thinking patterns can be used.

The word "growth" is attached to living organisms, and the "growth" of living organisms is a systematic and collaborative process. And this process must conform to certain laws. The so-called "one gives birth to two, two gives birth to three, and three gives birth to all things" is this principle.

Author: Louis Xu Wei

Source: Louis Xu Wei (ID: louisxuwei)