Facebook and Twitter's "Growth Hackers" tell you about user growth!

Andy Jones was a growth hacker for companies such as Facebook and Twitter, mainly responsible for user growth. In this article, he shared a lot of experience in user growth, including how startups recruit user growth team leaders, the three essential skills a user growth leader must have, and how to conduct product testing around user growth, etc. I believe this will be inspiring to most startups.

Andy Jones is the vice president of Wealthfront, an online financial management platform, and is responsible for user growth of the product. Prior to this, he worked as a product manager and head of user growth at Facebook, Twitter, and Quora, where he was primarily responsible for increasing the number of active users of the products. He has been responsible for placing over $10 million in online advertising, developed an email delivery tool that sends 50 million marketing emails per month, and conducted over 400 product A/B tests for user acquisition and retention. In short, he is the legendary "growth hacker".

At a recent CEO Summit held by First Round, Johns shared a lot of experience on user growth, including what startups should pay attention to when recruiting a user growth team leader, the three essential skills a user growth leader must have, and his own practical experience in this area.

User growth team tasks

Four years ago, when Jones recommended that some companies set up dedicated user growth teams, most of the feedback and questions he received were: "Do I really need to invest so much in user growth?" In fact, it’s no wonder. At that time, setting up a dedicated user growth team was a very new thing. At that time, most startups regarded user growth as the goal of the entire company and the responsibility and obligation of every employee in the company. At that time, startups usually had a dedicated finance team to manage the company's expenses, but there was no dedicated team to measure, understand and improve the user acquisition and churn of the product, which is exactly the task role of the user growth team.

The finance team is responsible for managing the inflow and outflow of funds into the company, while the growth team is responsible for managing the growth of product users.

In Johns's opinion, if a startup sets up a finance team but not a user growth team, it is undoubtedly putting the cart before the horse. As the name suggests, the finance team is mainly responsible for measuring, understanding and improving the company's overall cash flow. Many important decisions of the company are based on this, from staff recruitment and equity distribution, to salary adjustments to office hardware upgrades. Therefore, the importance of this team is self-evident. “The finance team uses its expertise to help the company run smoothly. However, what is interesting is that all companies, including the finance team within the company, eventually realize that the number of users of the product is the most important lever to maximize the company’s revenue potential.”

Over the past four years, Johns has seen a shift in the way people ask questions about user growth. Now most people's questions have changed from "Is it necessary to form a user growth team" to "How to recruit an awesome user growth team leader". For startup founders who want to accelerate high organic growth in the number of users, it is essential to hire a leader who is solely responsible for the user growth team. To hire such a professional, you need to determine whether they have the ability. "Do they understand the role of the head of user growth? Does their user growth approach fit their company's business products? Does their understanding of user growth still remain at the surface of just measuring and testing? In fact, there are some skills that every good user growth leader needs to have.

Three skills that a growth team leader must have

Through so many years of practical experience, Johns found that there are three basic skills that all members of the user growth team must master, especially the leaders of the user growth team. Here’s a look at each of these skills and how to tell if a candidate has them.

• Build a growth model: Ask them how they built a growth model for large-scale product promotion. This model needs to clearly explain the core means of product user growth.
• Building Experimentation Models: Gain insights into how they test and experiment with projects within a company. Find out how they organize their product design and development teams to support their testing initiatives, and ask them which is the most important project experiment they have ever run and why.
• Establish customer acquisition channels : Can they identify, find, test, and measure new customer acquisition channels? How do they filter and utilize different channels?

Depending on the type of user growth leader you’re looking for, the skill set you’ll need will vary. “If you’re looking for an executive head of user growth, the candidate needs to have at least two of the above three skills. If you find someone who is proficient in all three, you’re in luck, so hire him. If you’re looking for a non-executive head of user growth, the candidate must be proficient in one of the above three skills and have the potential to master and value the other two.”

In order to be able to easily identify whether the candidate has sufficient user growth capabilities from the candidate's answers, it is very important to understand some basic user growth framework knowledge. Here’s Johns’ in-depth analysis of the three growth skills above:

Understand your basic formula for user growth

There are countless complex growth models on the market, but Jones only recommends the following special growth equation:

Top of the funnel (traffic, conversion rate ) × magic moment (emotional resonance) × product core value (solving practical problems) = sustainable user growth

Johns learned this basic growth equation from his former boss, Chamath Palihapitiya, Facebook’s head of user growth. “Chamath is able to apply this equation expertly to determine whether a product has the ability to achieve sustainable and rapid organic growth. When I think back to the user growth of each product I have led, I find that I also follow this user growth equation unwaveringly.”

Here is Johns’ in-depth analysis of the three variables in the above growth equation:

• Top of the funnel: When it comes to user growth for a product, one question that must be asked is: Can this product acquire traffic and efficiently convert it into valuable uses? According to Johns, this variable is actually more of a tactical variable and its importance in the whole equation is not so obvious.

• Magic Moment: This is the emotional resonance that users experience when using the product. Ideally, this magical moment usually occurs when you are just starting to experience a product and are still using it. “I had this magic moment with Wealthfront. I woke up one day and my Wealthfront investments had made me $3,000 overnight without me doing anything. That’s when I realized two things: I wish I had more money with Wealthfront, and I should commit to this program.”

• The core value of the product: This factor directly affects the market size of your product, the rationality of the problem the product solves, and the correctness of your product/market assumptions. For products that deliver “magic moments” to users and whose core value is real, the top of the funnel will fill up naturally and quickly. At this time, you don’t even need to take the initiative to fill it in. When resistance obstacles arise, you just need to help eliminate them.

The magic happens when your mindset evolves from “I’m doing this” to “I have to do this because I enjoy it.”

When discussing the topic of user growth, the above equation for user growth is a basic starting point for discussion, whether it is measuring a product or judging whether a candidate is qualified for the position. “When you talk to someone who is applying for a growth team lead position, when you ask him about the framework he used to evaluate growth at his previous company, he should start with the growth equation above. It’s best to have him write it on the whiteboard and speak it at the same time, because both skills are necessary.”

Many times, Johns will ask candidates to list Wealthfront's user growth equation because this can test the candidate's skills and intuition. “I can tell from their answers what comes from their experience and what comes from their own research, and how they extract and synthesize information from the interview. You can see how they infer the main drivers of Wealthfront’s growth engine. Sometimes we use Amazon as an example to explain the user growth equation in practice.

• A = Vertical expansion
• B = Product inventory for each vertical category
• C = Traffic to each product page
• D = Purchase conversion volume
• E = value of each purchase
• F = Repeat purchasing behavior

The growth model Johns built for Amazon is a great example of applying the basic user growth equation to a real company. Here’s how Johns explains this growth model, and what he hopes the head of the user growth team can do:

• Vertical category expansion: "When Amazon was first founded, it only sold books, so its development was limited to the vertical category of books. When it expanded to many other vertical categories, such as music and digital products, it created huge growth potential and space for itself. Each new vertical category will bring new user growth. The key here is to figure out which category to get involved in, at what pace, and how to expand automatically."

• How much inventory should you have in each category ? How did Amazon go from selling 10 books to 10 million books to almost every type of book? Amazon's growth is proportional to the amount of product inventory it has in each product category.

• Traffic to each product page: Product inventory and being able to drive traffic directly from the web to your product pages are directly related. The key here is to successfully convert traffic into purchasing users.

• Purchase conversion volume and purchase value per order: The number of purchasers and the purchase value per order are closely related. Amazon optimizes this variable by adding a recommendation engine feature to product pages. When a user buys an item, Amazon will recommend a related product to him, thereby promoting new orders and ultimately increasing revenue.

• Repeat purchasing behavior: In the course of business development, in order to encourage users to make repeat purchases, Amazon will develop special purchasing mechanisms. For example, the launch of the Amazon Prime membership service is an example. This service can stimulate you to buy again and again.

The above is a typical example of how the user growth model is used in a real enterprise. When you are interviewing someone for the position of head of user growth, if the applicant says, 'Hey, I am very optimistic about your business and products', you can ask him to analyze your company's user growth equation as an assessment project. If the other party hesitates and cannot give you a satisfactory answer, it means that he is not qualified for this position at all.

Can they design a rigorous test program?

The growth models and formulas introduced above are theoretical after all, and they need to be tested and verified. Therefore, another skill that user growth managers need to master is the ability to design rigorous testing projects. “This is about experimentation that goes much deeper than the average A/B test. It’s critical to test a candidate’s knowledge and skills in experimental testing. Questions to ask include: How do you determine which part of the funnel to focus on? How do you identify the most valuable tests from a range of test items? When to run or not run a test?”

During the interview, you can actively test whether the candidate's insight is detailed. Taking the project testing as an example, one phenomenon Johns observed is that large companies really like to make some small optimizations to their products. "Large companies usually have a larger user base for testing. This is what we usually do at Facebook: for example, if a feature of a product brings 100 million visitors per week, we can do an A/B test on 1% of these visitors, and within 24 hours, we can understand the test results. If the test leads to a 5% change in user behavior, that's great. We will do more similar tests as soon as possible."

That’s what Facebook did in 2008. "That year, Facebook had 130 million monthly active users. Yet the company's board and management team announced a user goal for 2009: 300 million users. This was a very difficult task because it took us several years to accumulate the first 100 million users. Now we had to double the number of users in a year. At that time, Facebook's weekly user growth rate was about 1.5%. If we could achieve a weekly growth rate of 2%, we would reach the goal. This is how we broke down the task and conducted optimization tests. Of course, most of you are not a social network with more than 1 billion monthly active users, so here is an example that is closer to you to help you better understand how to accelerate user growth:

Let's say you have a medium-sized company with 10 million users, and your users are growing at 1% per week. "So after the first week, you have 10.1 million users, and after a year, you have 16.8 million users, which is actually a good number. The same company, assuming a 2% weekly growth rate, after a year, you have 28 million users, which is better than the first scenario and a good number."

However, most startups don’t consider the third scenario. In the last column of the table above, you started with a 1% weekly growth rate, but we assume that the growth team is primarily responsible for the active user growth rate, so every four weeks, the growth team can achieve a 5% user growth rate. If this continues, the number of users will increase to 30.5 million after one year, which is better than the second scenario above. ”

This is why big companies love small optimizations. “That’s the same logic that led Google engineers to test 40 different shades of blue on a particular sign-up button. If you’re a small company, that’s not going to work for you because you don’t have 100 million users and you can’t do a good job testing just 1% of them. You have to try different approaches.”

"This may be my favorite picture. It may not be very scientific, but it is symbolic. The vertical axis represents thoughtfulness, which measures whether you have thought deeply when designing, selecting and executing project tests. The horizontal axis represents the number of users in your company. When your company's products have large enough traffic and number of users, people tend to be less thoughtful. In large companies, it is easy to have many immature ideas for testing projects. But in small startups, you must be more cautious, attentive and precise in project testing.

Can they test different parts of the funnel? Because small and medium-sized companies have to be more careful about their experimental testing projects, they are more focused and rigorous about how they test in the funnel. "Startups have limited opportunities to test their products, and they are also limited by their cash flow. Therefore, they must do what they can to test meaningfully. To conduct meaningful tests on a small number of users, you must be thoughtful enough. Your company's future growth manager should also be deeply aware of this." To better illustrate this point, Johns also analyzed two experimental testing scenarios:

The table above shows three types of A/B testing. The base conversion rate refers to the final conversion rate of users performing a certain behavior in the way you expect. In this case, the focus is on the row marked in green font. This column represents the change value compared to the original control variable. For example, in Example A, after the user growth test, performance was 5% better than the control variable before the test. In Example C, it is 30% better.

Next, we will verify how several tests are getting better and better in terms of data by looking at the required test sample size. In Example A, you would need more than 72,000 samples to perform the test. More importantly, you need 72 days to verify that you are 5% better after the test than before the test. 72 days is too long and not cost-effective for a startup.

In contrast, in Example C, it only takes two days to measure whether the test results are 30% better. "Everyone wants this kind of result. If you are a startup, be thoughtful and maximize the variables you want to achieve. Don't do small things and do all kinds of unnecessary small tests, because it takes too much time to verify and measure. If you want to do it like in Example C, go big, so that the test time is short, which is more suitable for startups."

Testing 40 different shades of blue on a single button, as Google did, isn't feasible for a startup with a small user base. Don't be afraid to run tests that can make a big difference. This way, even if the test fails, it doesn't matter much because it only takes two days to verify, which gives you time to perform other tests. If the test is successful, you can get rich rewards immediately, so why not do it?

Now we take the base conversion rate as a variable to see its impact on the test sample size and test time. The base conversion rate here is the final conversion rate. “For example, if you have 100 users visiting your homepage, and you want to run an A/B test on these users, if 3 of the 100 people who visit your homepage successfully convert, then your conversion rate is 3%. Or if you have 10 people who visit your homepage and 3 of them successfully convert, then your conversion rate is 30%.”

“The higher the base conversion, the better. Similarly, if the base conversion in Example C is 20%, it will allow you to know the test results in 2 days. On the other hand, if the base conversion is lower, such as in Example A, it will take about three weeks to know the test results.

Make sure your future head of growth is proficient in the skills listed in the table above, and clearly explain how he or she defines himself or herself. It’s tempting for a candidate to say, “I’m going to test everything because I can do it quickly.” This is not something a growth leader should say. The person who said this is at best a qualified apprentice. You need a growth leader who understands the value and importance of smart test selection and design.

Don’t hire a growth manager until you have healthy organic growth for your product.

Guidance for growth

John has some advice for when hiring a head of growth:

A crappy product will never achieve sustainable user growth. “If your product isn’t growing organically, then your product is probably broken. Hiring a growth manager won’t solve the problem. Only when your product is healthy and your users are growing organically is it appropriate to build a growth team. To put it bluntly, a bad product will never grow sustainably. Nothing else can replace a great product.

You don’t need or want a growth hacker in management. “I say this from a hiring perspective. How many growth hackers do you know who have successfully led teams to achieve amazing user growth for 2-4 years at companies like Facebook, Pinterest, LinkedIn, or Airbnb? I don’t know any. I know a lot of growth hackers who are very smart and experienced, who are very specialized in user growth and have worked on small product projects. That’s it. You definitely wouldn’t hire a finance hacker to be your CFO, or a sales hacker to be the Director of Sales for a SaaS company.”

The head of growth also needs to be someone who understands product. At Wealthfront, the most meaningful product tweaks we’ve made to help us bring in the most users have come from our own deep understanding of financial products. There is no growth hacker who doesn’t understand product.

Before you can build an effective user growth team and find a competent growth leader, your product must be able to achieve healthy organic growth. When you meet the requirements to form a growth team, the growth leader you hire must have a clear understanding of the role growth plays in the entire company and be proficient in various growth models, testing, and implementation projects. You’re looking for a head of growth who is product-oriented and has user empathy.

When it comes to growing a company's user base, there's often a collision of science and art. The artistic school is usually only responsible for making WeDa's products, without worrying about data or testing. Scientific people are generally more inclined to measure and test things. Your head of growth must be able to blend these two styles, balancing user empathy with experimentation, passion with precision. Such talents are hard to find, but once you find them, don't let them go.

APP Top Promotion (www.opp2.com) is the top mobile APP promotion platform in China, focusing on mobile APP promotion operation methods, experience and skills, channel ASO optimization ranking, and sharing APP marketing information. Welcome to follow the official WeChat public account: appganhuo

[Scan the top APP promotion WeChat QR code to get more dry goods and explosive materials]

This article is compiled from: firstround.com and published by (APP Top Promotion). Reprinting this article requires the consent of Top Promotion , and please attach the link to this article!