Whether it is to grow, or to pursue happiness in life, we must start from the beginning and pay attention to the total amount of value.
Editor’s note: This article is from the WeChat public account “42 chapters” (ID: Myfortytwo), author Yang Lanqing.
This article is based on the sharing of “Chinese-style user growth exploration: Max (ALTV)” by Yang Lanqing in the “Growth Day Group” Alumni Association of Business School.
1) From a strategic perspective, how user growth is done
Before telling me how to grow users, I want to share with you some of the most basic logic, because if there is no clear underlying logic, you will swing when faced with major decisions, and you will not be able to stick to your own challenges. idea.
First of all, we have to think about what is growth?
Generally speaking, everyone thinks it is pulling new and promoting.
Looking for new and promoting corresponds to the growth of the number of users, and the increase in the number of users means that users grow?
I don’t think it’s not limited to this. From my experience for so many years, I don’t want to focus on the current income of an activity. The most important thing for users to grow is not DAU, GMV, or sales, revenue, Profit, but the way things are seen affect the growth trajectory of the user in the system, this is the most important.
So, to define growth, in addition to focusing on innovation and promotion, we must consider real user value in the long run.
So the first underlying logic is my definition of growth:
Growth starts with the end, using all resources to allow more users to use core product functions at higher frequencies.
Among them, focusing on “core product functionality” means that user growth is a strategic level and is always long-term and sustainable.
Further, we can say that the real user growth, the growth of the total user value, that is, ALTV (Aggregation LTV) rather than the user volume, user growth work function expression is Max (ALTV).
So I focused on user growth on Max(ALTV), pulling new meaning to meIn terms of it, it will not exceed 30%, because if you only pay attention to pulling new, just focus on the present.
At the same time, growth should start from day1. Some founders will consult me how to break through the growth bottleneck of their products. I expect that I will raise the growth immediately like magic. But this is what people who do advertising ideas do, and growth is not the case.
To break through the difficulties, not only must you start from day1, but you also need to have professional skills, and then through the correct methodological guidance, iterative testing, looking for growth points. Moreover, growth points often inadvertently erupt in a large number of tests.
Abstracting this truth is the second underlying logic, the essence of user growth:
Data-driven iterative testing turns subjective cognition into objective cognition, using test redundancy for growth certainty.
2) How do I grow in ofo
After identifying the two underlying logics of growth, combined with the case of the competition between ofo and mobike, let us look at how to do user growth from a strategic perspective.
Select the case of ofo and mobike because the competition between the two is fierce, and the growth decisions under high pressure are more representative.
To make user growth from a strategic perspective, we must first determine the milestones based on the company’s strategy.
Talking about the milestones, many people will say that our company’s goal is simple, to improve valuation, increase user volume and revenue. This is too thick, because it is right at any stage of any product.
The real goal must be concrete and phased.
Selecting a phased goal is very similar to choosing a strategy. Many times, when you grow, you don’t mean you don’t know what to do, but there are several things waiting for you to do, but no matter how big the company is, resources are limited, and choice is more important than effort. At this time, it should be done. Which?
I joined theo in mid-2017. At the time, about the end of theo and mobike, my judgment was that the two must be merged. The question is who is the dominant party.
Of course we want the ono to dominate. We believe that judging the dominant party depends on the amount of users and the amount of orders. To put it bluntly, who is bigger, this is also in line with the logic of valuation.
So our milestone at the time was that before the end of the year, the order volume of ofo would have a big advantage for mobike.
When I first joined, the order ratio of ofo and mobike was 0.6:1. If mobike is 10 million, then ofo is only 6 million. And I am leavingWhen opened, the ofo and mobike orders are 1.5:1.
After setting a good goal, the second step of growth is to combine the characteristics of the industry, think deeply, and reverse the path to achieve the goal and the corresponding growth project.
When I first went to ofo, mobike just launched the “red car”, which is actually the vehicle dispatching red envelope. The user will see a red envelope on the page (below), if you ride this car to a specific Location, it will reward you with a few bucks.
I analyzed at the time, mobike out of this function, is to solve the scheduling problem, some cars are riding to a very partial place, resulting in no one to ride later, red car is equivalent to crowdsourcing, motivating users to ride the car Near the office building or the subway.
In the face of new things from competitors, the first reaction of many companies is to copy directly. But at the time, I seriously thought about whether I should do this thing.
Under normal circumstances, people ride bicycles for commuting, and mainstream users are basically at this point, but with the red car, you have to ride down the office building, but for the mobike a few dollars, put the car Riding to another place, when I finished walking back to the office, I still have to be late, is it worth it?
So, I judge that core users will not be attracted to the red car, there will be more boring users to do this, but they are not core users.
Moreover, the red car can not really solve the scheduling problem. It may have stopped five cars in a remote area. Although the red car may solve two or three cars, there are one or two cars left. You are not going to go there. Therefore, in terms of cost, the red car did not solve the problem at all.
All growth-related activities are designed to change the behavior of the user and put it on the red-car. We should ask if the user is riding a red-car, will he ride more in the future? Most of the time, no.
But in terms of creativity, the concept of “red car” is still very good. Therefore, we return to the goal of rapidly increasing the order volume. Under the combination, we decided not to dispatch the red envelope, but directly to the user red envelope, no matter where it is riding, just give it a ride.
Another similar example is the Mobai Achievement Hall. The Mobai Achievement Hall is actually a point mall. You can get time-saving coins, eco-coins, and health coins by bike. Users can exchange coins in the mall.
At that time, we saw the Mobai Achievement Hall. Some students in the team said that the things that mobike did were really tall, and people must have long-term plans to make products.
Although I have not figured out the long-term plan behind this, we still face this problem. Does ofo do not make an achievement hall?
In the face of such pressure, I am still aiming at the goal. The goal of ofo is to make breakthroughs in the amount of orders and users in the short term. Only by gaining an absolute advantage in these two indicators can we become the dominant party in the merger. IEverything we have to do must go around these two indicators.
Later, we thought that the chain of thought from the three coins to the point store was too long, and the user’s decision-making process was reversed: the user was to get the goods in the end, so I would ride your car. You will find that this chain is quite long and users will not build complex thinking processes on cycling.
Most successful growth practices are designed to allow users to think faster, rather than slowly comparing them through slow thinking. Just like visiting a supermarket, if a user gets your product, he doesn’t put it in the shopping basket, but picks up the weight, price, etc. of the similar products next to it, and you lose.
Of course, this principle is not 100% effective, but the goal of our efforts must be that when users face choices, our products can let his brain call fast thinking logic to choose us.
So, sometimes choosing not to do anything requires more courage than choosing what to do.
C) Design growth should integrate four growth types
After the project is identified, the next step is to have a more refined design and execution. I divide growth into four types, funnel growth, functional growth, strategic growth, and integrated growth.
Funnel-type growth comes from Taylor’s reduction theory, which breaks down user-critical behavior into different links, then optimizes each link and improves conversion rate.
Functional growth means that I do a function, such as red packets, sharing, or playing games to promote growth.
Strategic growth is often mixed with other growth types. It does not use research and development resources. It simply filters users based on user images and uses A/B test to cure policies, such as texting or red packets. The probability that the user can come back is higher, and whether the behavior changes afterwards, and the better verification is solidified into a strategy and then automated.
Integrated growth refers to the integration of offline resources. In the industrial Internet era, products will be tightly tied to offline, so the integration type is now used more and more.
Design growth should be based on the four growth methods of the product, but in the examples I have seen, most of them do not engage in funnel-type growth, or do not engage in functional types, and rarely consider the other two. We must plan for different types of growth based on our product and strategic stages, allowing them to work closely together to form a coherent growth rhythm.
Use this growth portfolio to look at the specific practices of theo red car at the time. From an external perspective, the direct redemption of users is a functional growth, but the design behind us is far from simple.
I also thought about a more aggressive Bug marketing promotion plan, that is, I found a few KOLs to release the news, saying that the ofo system failed, and everyone quickly came to ride the red envelope.
The decision-making layer considered that we were negotiating with investors at the time, and we were afraid that investors would be worried that our technology would not work, so they did not. However, if this activity is done, it may greatly save our promotion expenses in the previous period.
At the same time, the offline link will also be opened, that is, integrated growth.
At that time, I hope that the offline staff can wear a vest that promotes red-car activities to do vehicle operation and maintenance. However, it was only three days before the decision to launch the red-car service at the 11th, and it was too late. Therefore, I directly proposed to the offline operation and maintenance personnel to do some chest stickers and back stickers, which are printed with our activities, attached to the front and back of the operation and maintenance personnel vest. In this way, the operation and maintenance personnel wearing the vest is equivalent to a live advertisement. When the user sees the horse, he knows that theo has the activity of riding a red envelope.
In addition to this, we also learned the lessons of Didi and Uber, taking into account the impact of negative public opinion in advance. Before the drop to the user red envelope, some users criticized it against the laws of the market, is a trick. Ando’s red car is even further, not a discount, but riding a bicycle, but not money, but also to send money to the user. Therefore, before we do the red-car, we also let PR make an emergency plan for possible negative public opinion.
The question I am always thinking about is, what do I need if I want to make this event a success? What factors can cause an activity to fail? How to prevent it from failing? If the activity fails (most tests will fail), how can it be remedy? How can it be iterated quickly?
A set of actions form a typical comprehensive growth model. The red car is only a shell, it is a function. The real activity is the timely adjustment of the red envelope strategy, the high integration with the offline link, and the consideration of the activity details too too much. .
4) Good delivery to see channel users LTV/CAC
The technology of channel delivery is not large, and the methodology is relatively mature. It is impossible to set up a tracking system and control strategy. Basically, I will not repeat the basic content, but only emphasize one point.
Everyone wants a question, what is good delivery?
Many channels will say that they are coming to me and the price is low. But for example, the average user of the A channel is 10 yuan/person, and the B channel is 100 yuan/person. Is the A channel better than the B channel?
Not necessarily, because the customer cost of A channel is 10 yuan/person, but LTV may only be 2 yuan; although the cost of B channel is 100 yuan/piece, LTV may be 200 yuan–the investment channel A loses money, and investment B Make money from the channel.
Therefore, a good delivery should be that the ratio of user LTV to customer acquisition cost (ELTV/CAC) is relatively high.
LTV is a long-term indicator, so the channelThe effect is difficult to track quickly, and we can only find short-term indicators.
At the time of my time at Amazon, the e-commerce team looked at the short-term indicator of the seven-day ROI, because everyone’s life cycle is seven days as a unit, and the seven-day behavior has a high probability of indicating long-term behavior. By performing regression analysis on these data, we can predict ELTV (LTV expectation).
5) HVA: How users jump from low value curves to high value curves
In the beginning, we mentioned that real user growth and growth is the total amount of user value, ie ALTV (Aggregation LTV) rather than the amount of users. Then, we also talked about the specific methods and growth models of ALTV.
So, finally, share with me a growth-based underlying methodology that is my own, and a pre-target for ALTV growth, HVA (High Value Action).
I started to explore this theory with the team from Amazon in 2014. By the beginning of 2018, Taobao user operations manager asked me to communicate. I drew the following HVA map for him. I hope that if you can only remember one point today, then this is the picture.
The horizontal axis of this graph is time, the vertical axis is the value contributed by the user, and the integral of the area under the curve is LTV. Usually, the value curve is gradually increased first, and then slowly shrinks.
Traditionally, when the curve begins to decline, everyone will start doing some things to promote, so that the curve should not fall as far as possible. However, this time is already too late.
For any product, users are graded from the first day they come in, there are heavy players, and they are not active. Therefore, what we have to do is, from the first day, to keep users addicted to the product, and constantly let users on the low-value curve transition to high-value curves.
So how do you achieve the transition?
It depends on HVA, so the purpose of HVA is actually to improve the user’s LTV.
For example, if a girl only bought clothes on Taobao and did not buy 3C products, then the first time she bought a 3C product on Taobao is a typical HVA, and then she might feel that she is on this platform. Buying is more convenient than going to other places, after which 3C products are bought on Taobao.
If someone else does something that can produce HVA after a year, and you do it in a month, you can grow from value curve 1 to value curve 2 earlier, then for the whole year, your Incremental value will be more than others.
In order to make the value continue to increase, all the product functions we have to do must be a series of growth systems. The system is dismantled through one HV.A, guide the user to jump on a higher value curve.
This means that we must always look to the long-term, any activity is not for an order today or tomorrow, but to see if this single can bring changes in user behavior in the future.
Finally, I would like to share with you my outlook on life and connect with growth. I also define a measure of success in life, called LTH (Life Time Happiness), if the accumulated happiness in a lifetime (shadow in the picture below) The area is particularly large, then this life is a very happy and successful life.
According to this formula, health is very important and cannot be drowned, because time will greatly affect LTH. Similarly, if happiness is particularly low in order to achieve a certain goal, the whole LTH will not be high.
Summary, whether it is to grow, or to pursue happiness in life, we must start from the beginning, pay attention to the total value, and continuously realize the jump of the value curve.
(Source: Chapter 42 Business School)