Deconstructing a practical case study of user growth for products with hundreds of millions of users

The strategy in this article has helped BeautyCam achieve sustained monthly user growth of millions, and has been shared many times with good responses. I hope that through some of the growth methods I have accumulated, everyone can fully understand what the recently popular growth positions are doing? What is the difference between growth and product?

First of all, please put yourself in a role. You are now responsible for the growth team of the entire beauty camera. The task given by your boss is to increase DAU. Where would you start? Through several cases of beauty cameras, I would like to share with you five profound insights, which may be inspiring and helpful when you are working on many growth projects.

1. Growth is not entirely equal to "fission" or "addition", it is the system capability

You may think of fission, attracting new customers, finding problems, thinking of a strategy, etc. at the beginning, but growth is not a certain strategy, it is a systematic scientific process, a systematic ability to determine goals, select key issues, and find growth paths.

Step 1: Identify your North Star Metric

1. Why not just use DAU as the target?

Of course, the goal of each business is DAU and other indicators, but growth should not directly take the big goal as the North Star indicator.

DAU and retention are very important and are related to many factors. If you focus on big goals, your strategy will be scattered. DAU growth does not necessarily mean a good direction: buying volume, festivals, and activities.

DAU is almost the opposite of retention;

2. Failure to find the right growth target "North Star Indicator", the conditions for its establishment:

Positively correlated with retention rate, and reflects the real activity of users; reflects the value of the product;

In line with the behavior cycle, the formula can be disassembled;

Therefore, for beauty cameras, the North Star metric should be the number of weekly retained users . The reasons are:

Only when users complete taking photos/saving them, can they appreciate the value of the camera product. At the activation moment, if the photos are not saved, the value of the product is discounted. It is positively correlated with the retention rate: the more people save, the higher the retention rate.

Step 2: Determine the growth formula

Based on the North Star indicator, determine the growth formula as follows:

After determining the model formula, there are two more questions:

1. There are so many impact factors, which one should I do first?

Order of magnitude: number of new users per week, number of retained users per week, number of recalled users per week; conversion rate: preview rate, photo taking rate, and save rate;

2. Why didn’t you choose to add new items & orders of magnitude at the beginning?

At that time, the loss rate of beauty cameras was much faster than the increase in business and new business (the loss rate was more than twice the increase rate); and the increase in business and the order of magnitude were costly, difficult, and time-consuming;

Traffic can be compared to water in a bucket. The whole bucket has a big hole and is leaking. At this time, choosing to add new volumes has a very low cost-effectiveness.

So I divided the growth path into three stages, focusing on different issues and things:

Step 3: Decision-making method

Two decision-making methods to sort out what you need to do.

1. With quantitative data growth models as the main line, sort out product traffic maps, funnels, models, etc., and explore opportunities from a data perspective.

2. With the psychological factors of user decision-making as the auxiliary line, sort out the psychological decision-making factors of user level migration, think about what the problem is for each transition decision and what the influencing factors are, so as to explore opportunities

2. Low-hanging fruit

Low-hanging fruit is decisions that have low costs, high probability of success, and obvious effects . Finding low-hanging fruit is an important event in the initial stages of growth. How to find it? Look at the data quantitatively to find weak points, or conduct qualitative user research to find problem points.

Step 1: Problem

In this case, the North Star Metric is the number of retained users. Knowing the preview rate, shooting rate and saving rate (penetration indicators of different functions) of existing users, it was found that the preview rate was much lower than expected.

The preview rate refers to the proportion of users who see their own face before taking a photo after clicking the camera. The current process is that when users open the camera, they will see the home page instead of their own face directly. However, due to the habits of hundreds of millions of users and tens of millions of revenue, the homepage is almost an insurmountable hurdle. How to solve this seemingly contradictory problem?

Step 2: Ideation (Formulating Hypotheses)

• Allowing users to enter the beauty camera and see themselves in the shortest possible time, so that they can take good-looking and satisfactory photos, can increase activation rate and retention rate;
• New users and returning users (users who have not returned for more than 30 days but returned today) have the largest path factor and are swing users. They must complete key behaviors within X seconds.
• Improving retention and extending life cycle can make up for the value of home page advertising

Step 3: Experiment

In the AB stage, users were divided into three categories, and an experiment was conducted to remove the homepage, observing the changes in shooting rate and multi-day retention rate.

It can be seen that for returning users, both the shooting rate and the multi-day retention rate have increased significantly. The assumption is basically established, but how to measure the advertising revenue?

Step 4: Measure and Amplify

In fact, it is nothing more than a calculation problem:

Over time, there must be an inflection point, K1*B>K2*A+K2*B, and advertising revenue is positive (you can click on the large picture to see the details)

In the end, both the business department and the boss recognized the value of the benefits and applied the experiment to returning users, which resulted in a huge improvement in the final experimental effect. This is a typical case of low-hanging fruit. The difficulty and cost are not high. The key lies in the discovery, experimentation and estimation methods.

3. Find the point with the greatest impact

When coming up with ideas, you need to break out of your original thinking framework, think about all strategies from a global perspective, and try to find the point with the greatest impact.

Compared with its competitors, the biggest advantage of Beauty Camera is its hundreds of millions of existing users, several times more than its competitors. At that time, Beauty Camera had been optimized and got rid of fake effects, and even had a deeper understanding of "beauty". It was urgent to make users re-acquainted with the young and fashionable Beauty Camera; and PUSH was an important means for users to recall and re-acquaint themselves with the product.

Step 1: Analyze and find the problem points

• The reachable users are generally about 2 times the monthly active users, but we only have 1/2 of them. There is at least 10 million more space here.
• Push opening guidance is not done (estimated to be 20% lower than similar products), there is at least X million space here
• The total amount of push notifications is too small, and we are worried about the large amount of user feedback.
• Click rate and activation rate are too low

Step 2: Come up with an idea

It is still divided into three stages:

Step 3: Experiment and measure

There are many items that can be optimized for PUSH. Before starting, you need to establish reverse indicators: uninstall rate and complaint rate. For controllable purposes, users are divided into 8 groups based on the high and low values ​​of the RMF model (Recency: the user has saved the app within X days, Monetary: the user has saved the app Y times within 30 days, Frequency: the user has saved the app Z days within 30 days). Different optimization methods are used for different groups to ensure that the uninstall rate and complaint rate do not increase while maintaining the order of magnitude and conversion rate.

4. Growth is a magnifying glass for product managers

This is also the most important point. Growth products are not about playing numbers games. In essence, they are still product managers and are inseparable from users, needs, and scenarios. It’s just that growth product managers magnify the product value and user value of users from a more logical and rational perspective, which is called a "magnifying glass."

1. Treasure function

The role of our TEAM at that time could be summarized as: making users feel beautiful as quickly as possible!

As an old product that has been iterated for a long time, there are many legacy issues in the architecture, which leads to many features with high retention rate, high preservation rate, and good reputation but few users know about them (called treasure features), one of which is called "Quality Blockbuster".

This feature was discovered unexpectedly during the interview and has a surprising reputation, especially among young lapsed user groups, and the Amway success rate is very high. Compared with users who have not used quality blockbusters, the retention rate of users who have used quality blockbusters increased by 14%, and the next-day retention rate increased by 5%.

The problems are: the path is too long and the filter focus is not highlighted. The solution is:

2. Scenes and reasons for taking photos

In order to better understand the user's scenario, after stratifying the users according to the RMF model, the core problem can be abstracted: a relatively high R value means that they come back frequently, while a relatively low F and M value indicate that the user has no reason or scenario to take photos .

Try various methods to help users find reasons and situations to take photos:

BeautyCam is also the first camera product to try out a points system:

3. Emotional Accounts

There is another very important question that I had not fully figured out until I left Meitu: how to quantify the emotions generated in the photos and create an emotional connection between the product and the users.

At that time, I thought of various ways to help users establish emotional accounts within the product:

5. There is no panacea for growth, Learn and Win!

Everyone always thinks that once a team has a dedicated growth product, it is omnipotent and will definitely bring growth to the product. However, the success rate of growth is less than 20%. Hope can only be found in continuous iteration and trial and error.

But the sentence I like very much is not learn or win, but learn and win . The growth team is generally the most stressed in the team, with pressure, anxiety and pain, but this sentence reminds growth products that they must have the confidence to win in order to find a breakthrough.

Author: WinsonL

Source: WinsonL