Didi’s user growth logic: How to gain insight to break through the bottleneck and become the 1%?

I used to be a very curious person. When I was little, I thought that things would make noises if put in the toilet. During the Chinese New Year, I saw people setting off firecrackers, so I put the firecrackers in the toilet and lit them. As a result, the whole house was filled with shit. Later I wondered why this thing had such a great effect?

When I grew up, I discovered why many people from other provinces and cities would travel a long way to Beijing to buy a used car ? He may live in Shaanxi, or even further away and buy a car to bring back. Going further back, I saw that last year Ofo and Mobike were placed at the entrances of subways, and they kept handing out free ride cards. What was their real profit?

Before coming to Didi , I was also thinking about a question: why does Didi keep issuing coupons to users, perhaps 20 or 30 for one order? What is the reason for this? Today, let’s start my sharing with you about these strange problems we see in life.

The topic I want to talk about today is insight and growth. I will divide it into five parts from the perspective of insight:

1. Definition of Effective Insights

2. Means of insight

3. Insights building method

4. The logic of growth

5. Case Sharing

1. What is effective insight?

We will find that people with strong curiosity will see many interesting things in life. You will see many strange things in life suddenly appear and disappear quickly, such as Faceu and Million Heroes. Why is this?

People with strong curiosity will observe what they are? Observation is not watching, you cannot do business by just watching.

First we need to observe the demand .

My understanding is that the demand is what needs to be done by whom, at what time and in what place. People are the subject , time and space are the conditions behind the needs, and the second is what he wants to do, which is the demand.

Then comes the scene. Behind his needs is the timing for him to do this.

So from a business perspective, if you want to achieve growth, the first step is to gain insights. What are the needs of users? The next step is to clearly understand the people who have the needs, the time of the needs, the location of the needs, and the timing behind the needs.

For example, I want to understand why Mobike and ofo placed their bikes at the entrance of subway stations when they were first promoted ?

If I talk about it from a business perspective, this is the scenario with the lowest customer acquisition cost. Mobike requires a deposit of 299 at the entrance of the subway station. The cost of a bike may be 1,000, but in fact, not just one person pays a deposit for a bike. Maybe 10 people pay a deposit in a week, and the total deposit may be 2,990, while the cost of a bike is 1,000. The company makes money. This is the first look at demand from the company's operating dimension.

Looking further down at user needs, when and where does a person need to travel? For the 400 people here, it will be difficult to get a taxi after the meeting if you don’t want to share a ride because you will have to queue. There used to be dynamic pricing, but many people would complain. At this sudden time point, we must find an emergency solution, and this is insight.

But what I just said is useless. Based on the needs and scenarios, what will the next implementation be like? We will look at the implementation method below.

2. Insight into the means

I think the two most important things are, first, MVP, the minimum viable product experiment; second, doing research.

1.MVP

My method is actually very simple. For example, I might have an idea. When not many people in Beijing used taxi-hailing apps, if you wanted to call a taxi remotely, you had to pay a two-yuan wire transfer fee. I wondered if I could make a software that could use the Internet to call a car. This would be ideal. Before realizing your ideal, when you don’t know what the product will look like, we have to use MVP, which is an experiment in which you use operations or manual methods to simulate future products.

For example, we can go to the West Railway Station and the airport to attract drivers through ground promotion , and then go to the vicinity of Zhongguancun C980 and send orders to users in office buildings to download the Didi app. With this method, we can observe how many people will open the app and call a car at what time and place?

2. Research

Next is the research. In fact, the core purpose of the research is qualitative exploration, exploration of options, quantitative verification, and verification of results.

If I survey everyone here and ask them how they got here, you might have many options, including taking the bus, riding a bike, taking a taxi, driving, etc. I need a way to collect information, such as putting my WeChat QR code, and having you scan it to tell me your options. This way, I can do quantitative verification, and I can know which method is more suitable for evacuation at the venue today.

Simulation scenarios are actually a way to extend MVP. Once you have a certain result, you need to continue simulating in the scenario. You can pretend to be a driver or a passenger to simulate. At the beginning, Didi recruited a group of drivers. When they found that there were no passengers, the drivers deleted our software. Our job was to hail taxis on the streets every day and make the drivers feel that this was reliable and that they could still get business.

So after I have done the insight and understood the needs and scenarios, how should I implement this thing?

3. Factors of Growth

Once you understand the needs, you can figure out who your brand should be promoted to, what tone it should be promoted to, and what materials and copy should be used to promote it.

When you know the scenario, time and space, you will know what channels to use for your entire product and business, so this is a progressive relationship.

1. Let’s talk about the brand first

As far as I understand it, there are three types of brands. 90% of brands convey what, 9% of brands convey how, and only 1% of people who work on brands know why. Why?

We see all kinds of advertisements every day. Many marketing/operations people, most of whom write copy, only tell you what it is; 9% will tell you how to use it, how it can change your life, and how to improve your efficiency; only 1% of good products, or good brands, will tell you why they do so.

For example, when I first knew Apple, Steve Jobs told me that Apple wants to create the most perfect product because we are a group of great people. Only 1% of communication materials are of this type. Most of the time, they say, for example, a certain brand of computer recently released a very cost-effective computer with a running score of 999 points and a price of only 888. This kind of thing just tells you what.

2. Channels

There are two types of channels: passive channels and active channels. Passive channels are advertisements like Moments and Fans Channel , while active channels are app markets , Apple App Store , searches, etc.

All users are actually active users and passive users, existing in active channels and passive channels. So how do we choose channels?

I choose channels based on scenarios and needs. All demands are divided into two categories, high-frequency demands and low-frequency demands, but because the reduction form of demand also involves a decision-making process, it is divided into long decisions and short decisions.

You can understand it this way. All high-frequency demands require short decision-making. For example, if I want to eat, I cannot think about eating for three years, otherwise I will starve to death. If I want to go to the toilet, I cannot hesitate too long. These are all relatively short decisions.

Products with short decision-making times are seen every day, such as travel and food delivery products, which are high-frequency products with short decision-making times. For products with this type of demand, the best channel is the passive channel to influence passive users. For example, you will see that KFC and McDonald's do it very often. Passive high-frequency users are everywhere. If you don't have money, you can find a group of old ladies on the street to find passive users.

All low-frequency demands can be divided into two categories, one is called long decision and the other is called short decision. Buying a car or a house is a long low-frequency decision, while moving and hiring a housekeeper is a short low-frequency decision. It's the same reason. Because low-frequency decision makers must have a higher average order value, they need to build brand influence in passive channels and acquire customers in active channels. So you can see Mercedes-Benz will advertise, and next they must do what Baidu does and must do it on Autohome because they want to attract users there.

On the other hand, when have you not seen a moving company put up an advertising sign? Because its average order value is relatively low, and the best channel for products with low frequency and short decision-making is the active channel. For example, you can just go to 58 and search for a moving company, and you will find it. This active channel is most effective for products with low frequency and short decision-making time.

3. User education

After understanding the brand and channels, the third level is user education. User education can be summed up in three sentences: Who am I? Why use me? Why do you have to use me?

Maybe many products cannot explain these three sentences clearly. There is a simple method called anchoring method, which is to anchor something that everyone can imagine in life.

For example, when the term Didi Express was launched, we were at a loss internally because no one could quickly understand the relationship between Express and online ride-hailing . People might use Express to describe trains and buses, but using Express to describe small cars was a bold attempt. So how do we define the subsequent user education so that he can understand it quickly?

You can anchor taxis. Taxis are available in every city and everyone can understand them. Let me first introduce who I am. Didi Express is a taxi; then why do you want to use me? It's very simple, because it's cheaper than taxis; so the whole slogan comes out: "Didi Express, a taxi cheaper than taxis" launched in Beijing. This is easy to understand.

4. The logic of growth

Now that we understand the factors, the next step is logic, which is to help everyone find methods and rhythm.

First, cut it into two parts. The top part is demand, seeds, outbreak, and bottleneck. I will then tell you your starting and end indicators for each rhythm.

The following are the most effective means. I will tell you about seeds, outbreaks, and bottlenecks, and tell you what I think are the most effective methods from the three dimensions of attracting new users , activity, and retention .

1. Demand → Seed

After you verify the demand, your product will enter the seed stage. The core of the seed stage is to look at the retention rate . I have obtained a piece of data that 90% of the APPs in China have a 3-month retention rate of 0, because many APPs are imitations and plagiarisms and cannot attract users. What does this mean?

If I think that the X-axis of the following graph is 0 to 3, and the vertical axis is the retention rate of 100%, the data of a three-month product is often at a 35-degree angle. In the process of retention rate decline, there is suddenly a point where the slope is no longer steep and becomes slow. You find this point.

From the data I have learned, if the product demand is OK, 90% of the apps will have a retention rate of 10%-30% within 3 months. This retention rate directly reflects whether your product can grasp user needs.

What specific retention rate is considered to verify the demand? This depends on the product. Low-frequency products usually have a lower retention rate, while high-frequency products have a higher retention rate.

2. Outbreak period.

There was a question just now: When is the starting point of the product's explosive growth period? That is, the retention rate reaches the best. The retention rate can ensure that I have a relatively good user retention phenomenon. At least a part of the users brought in can always stay in my product.

What should be done before the outbreak starts?

Why do you see Mobike and ofo constantly releasing bicycles? Because the more bikes you park, the more deposit you can earn. This point tells everyone that the most important thing for you is to find a good growth wheel, and you should think about it before the outbreak period.

The entire growth wheel has three most important indicators: lifetime value LTV, customer acquisition cost CAC, and positive benefit time PBP. What do these three indicators mean?

Let me give you a simple example. For example, Didi's customer acquisition cost for a driver is 400 RMB. Then the driver takes orders on the platform for three months. He takes 20 orders a day and each order is 20 RMB. He can earn 400 RMB a day. The platform takes 20%, which is 7,200 RMB in 80 or 90 days. The logic is that I spent 400 RMB to buy a big white goose, and it lays eggs every day. After three months, the eggs are sold for several thousand yuan, and I make money.

This logic is based on the early explosive period. Any product needs to observe the user life cycle , time and value in the seed stage. It is necessary to estimate where the ceiling of customer acquisition costs is and how long it will take to generate revenue from customer acquisition. This is business thinking. There is something wrong with doing business if it does not make money.

Looking further down the line, what signs of the end of the entire outbreak period should we pay attention to? There are two phenomena. First, user activity is constantly declining. For example, I may pour water into a bottle in the early stage, and then the water level in the bottle increases. I may only stir the water at the head, and the water at the bottom becomes dead. This means that the explosive period is gone and I have to stop.

There are also changes in three indicators. It is possible that the profit time has become longer. The channels may be very cheap in the early stage and very expensive in the later stage. The life cycle value cannot support my channels, so I must stop.

If you haven’t even seen the outbreak period, then you must be among the 90% of dead apps. I found that many softwares spent money to promote their products during the seed stage, but died. Even when they were still doing ground promotion during their boom period, they also died. At the seed stage, 90% of the apps in the world have already been eliminated. Only 10% of entrepreneurs can reach the bottleneck period, and only 1% can break through the bottleneck period.

3. Bottleneck period

The first thing to do during the bottleneck period is to stop attracting new users and building channels. Instead, you should look inward. There are actually many users inside, but many of them are water users.

We were also anxious. When we were fighting with Kuaidi and constantly issuing coupons, we burned a lot of money on the taxi business. We had a lot of water, and an old lady who hadn't taken a taxi for several months received 10 yuan for a taxi ride. During the bottleneck period, it is necessary to improve the product and efficiency model, and the product needs to return to the demand itself to find possible other quadrants.

For example, Meituan Takeout first solved the problem of too many takeouts, giving users many choices. Later, it solved the problem of fast food delivery.

The first problem that Didi solved was the difficulty in hailing a taxi. After that problem, the next step was to solve the price issue. Then I might have to find a product that is cheaper than taxis and provides a better experience, so private cars and express cars came out. This solution is the idea of ​​solving the bottleneck period.

There are four best ways to increase user activity: first, time; second, space; third, pricing; and fourth, promotion. There is nothing else.

Giving him different pricing at special times and places, and then creating a variety of flexible promotional tools based on the pricing model, is the most effective way to promote activation. The most important indicator during the bottleneck period is activity, during the seed period, the focus is on retention, and during the explosive period, the focus is on the increase in new users.

For example, the second round of WeChat's explosion was red envelopes, which really brought WeChat from first- and second-tier cities to fifth- and sixth-tier cities. In 2013, many elderly ladies were installing the software to grab red envelopes. How do we get through the outbreak period and then move on?

When WeChat's red envelopes were promoted all over the world, WeChat discovered a phenomenon: money was constantly being withdrawn and it needed to keep the money. This was the biggest problem at the time. Next came WeChat Pay and WeChat Wallet. Today, you can go to every corner of the world and proudly say that mobile payments are accepted in China. This is something that foreigners think is amazing. The entire process of demand seed, explosion, and bottleneck, this is the logic of growth.

I would like to give you some suggestions on what you should pay attention to during your work.

First, when you are unable to focus on retention, you must not attract new customers. You can spend more time looking at your product.

Second, when you find that you do not have enough cost to make a breakthrough, you will pay more attention to your channels and forwarding rates. At this time, you must constantly provide feedback to entrepreneurs on the three essential indicators.

Third, when you enter the bottleneck period, when you are confused, you are lucky to have surpassed 99% of the products in the world. You must constantly look for your products and efficiency models to solve it.

5. Case Sharing

Finally, let me tell you a personal story. Don't think that Didi is a different company from you. Didi has many projects internally. In the early days of these projects, we had no money and no people. We needed to find suitable people within the department to do the work. We call it FT internally, which means we work together to do this.

In this process, we have to start a business every day. Let me share my own entrepreneurial case.

In the summer of 2016, a product called Didi Minibus was launched. It is now expanding in Beijing's North Fifth Ring Road. It was first launched in Beijing's Huilongguan, and at that time it was designed to solve the needs of the last mile. There is no way to give specific data, so let's use numbers instead. At that time, Didi Express was 1 in Huilongguan area, and I achieved 5. How long did it take me? One month.

What he solved was the last mile, which is a very important strategic issue. However, he didn't expect that Mobike and ofo would come and ruin the whole thing. He didn't expect that the last mile would be solved by bicycles so quickly. In this case, there was a rapid 5-fold growth from 0 to 1 in a short period of time. I will use this picture to explain to you my analysis method at that time.

The first point is the basic quantity, your business scale, brand penetration, share, and reputation, this is your scale.

If I want to understand the people of Huilongguan, I need to understand the local buses and taxis. First of all, there are basically no taxis in that area, and buses have three pain points: crowding, waiting, and congestion. Every day when I get out of Huilongguan subway station, it's like getting out of a vegetable market, and I have to queue up. These people are stuck there and waiting in the hottest weather. This crowd represents the basic volume.

Second point, variables. To achieve growth, you need to take advantage of trends. For example, if I want to sell a hot-selling product, I might take advantage of Double Eleven . And if we, who are in the travel industry, want to achieve growth, we will spend the subsidies on weekends or holidays.

The third point is regulation. When you discover these two points, you have to develop regulatory measures, such as using educational models to formulate passenger strategies, driver rewards, pricing, and how to dispatch cars. Going further into the market, for example, defining new user education, effective channels, marketing activities, and scenario stimulation.

The story begins. Didi has several pain points, namely, the long pick-up time, the difficulty in finding a car at the subway station, and the fact that when you take a car, there are three empty seats, which is a waste. The form of carpooling can solve this pain point to a certain extent, so minibuses came into being.

When this product is launched, we need to find specific scenarios and the entire brand communication method. The best thing in the seed stage is ground promotion to verify my retention, which is my core purpose.

First we did a brand exposure. When Huilongguan Hualian opened its doors, I discussed BD with them. I sent out 100 flyers, but there wasn’t a single order. I felt that this was a very poorly made product. Thinking back, it might be the wrong time.

The next day, I arrived at Huilongguan and went to distribute orders at the entrance of the subway station in the morning. I found that people would take them, but there was no conversion . When I distributed them in the evening, the conversion rate reached 2%. I don't look at the person who took the order, I look at the person who lost my order. I asked the users why they threw it away, and they said they had no need, their home was just across the street, they were waiting for someone, they were annoying... basically all kinds of questions, and I basically understood it as no need.

Later I found that there was a bus stop at the entrance of Huilongguan. It was crowded with people, most of whom were waiting for the bus to go home, so I went there to send it. The conversion rate is 7%, which has tripled.

After I posted it, users started asking me what a minibus is. Where are the minibuses? What is the minibus entrance? How much does a minibus cost? And so on. At that time, I found that my DM was very poor. I started by making small cards that were beautiful and foldable, but users didn’t understand them. I need a simpler way.

So I asked our sales staff to wear special T-shirts with place names written on the left and right sides of the chest and prices written on the belly buttons. They should hold a sign in their hands that said Didi minibuses with lights on at night and a horn on their waists, which would keep blaring, “It’s only two yuan to go to Huilongguan, there are air-conditioned seats and they should hurry up.” It was this kind of down-to-earth scene that attracted people at the bus station and made it easier to get orders, which increased to 9%.

At the same time, I found that many users still don’t understand, because everyone has different understandings of the word “minibus”. Some people think that a bus with more than 30 seats is a medium-sized bus, and a bus with more than 10 seats is a minibus. In fact, a minibus has 7 seats, and I need to make users aware of this.

At that time, I found a minibus advertising vehicle, turned around in Huilongguan and kept running. At the same time, I asked all the salesmen at the station to stop pushing and just do one thing - to queue up. The girl in the T-shirt was standing. The bus was coming soon. The driver came over with a horn, smiling, and shouted: The Didi minibus has arrived at the station. You can go to Huilongguan at any time. There are large seats with air conditioning. It's two yuan per person. Get on quickly.

After a user was brought in, the car started. It was summer then and the driver was asked to adjust the air conditioner to the maximum, 10 degrees lower than the outside temperature as it was hot outside. After getting off the car, everyone took a bottle of water. Everyone was very happy and said thank you to the driver, and the driver said you're welcome.

In addition, our group of people lined up very neatly at the time, and the bus stop was very scattered. Later, we found that after the bus arrived, our group started to get on the bus, and people at the bus stop also started to get on the bus. The contrast was not very strong.

One thing we did immediately was to provide a 15-second pick-up service. We asked all the minibuses in Longze and Huilongguan to turn on the double flash lights. After Longze came out, there were 5 or 6 buses with double flash lights. You can quickly scan the code and enter your destination. This time is significantly shorter than that of express trains and buses, and is shortened to 30 seconds. This has increased the conversion rate of our entire ground-based marketing. We no longer need to send out orders. Instead, users can directly find our brand and scan the QR code.

At the same time, the seed period was over, and I spent a month observing the monthly retention. After the monthly retention was over, I found a very effective channel, such as the full-site advertising package for the Longze Huilongguan subway station.

I mentioned above that we are a startup team and the company will not give you a penny unless you prove your value in the seed stage and prove your ability to generate revenue quickly. The company later said that they would reimburse you 400,000 yuan, plus some other targeted framework media for advertising. This is a successful case of internal entrepreneurship, and it has now been handed over to other departments.

Fourth, user education. For example, if we run an advertisement in a community, the effects of copy A and copy B differ by three times. Everyone will definitely know which one is better as soon as they hear it.

Copywriter A: What is Didi Minibus? Didi Minibus is... (and then put those slogans there)

Copywriter B: Dear community owners, hello. In order to facilitate travel for community owners, our community and property management have specially cooperated with Didi Minibus to open a community special line to the subway station. You can then scan the code to join the group and take the ride. This is a very effective method.

As for advertising space , the best way is to go and take a look first. For example, if I want to rent an LED advertisement in a city, but I don’t know whether this LED is good or not, what should I do?

I would often ask pedestrians/stall owners at LED newsstands or in the small area near the LED if they knew what the advertisement was about. If they told me that gynecological hospitals broadcast it every day, I would know that this thing is effective. If they don’t know what is being broadcast, then they shouldn’t do it there.

So if you want to grow, insight becomes extremely important.

<

p style=”text-align: center;”>The author of this article @李森 is compiled and published by (Qinggua Media). Please indicate the author information and source when reprinting!

Product promotion services: APP promotion services, advertising platform, Longyou Games