We were pregnant with a “mummy” heart, and we found that we only had to be a “miss”.
In the Chinese investment circle, the “platform” has a honeymoon period. When O2O is at its peak, the platform seems to be able to connect everything, and all industries can rely on the platform to integrate producers and consumers. There was a fashionable saying at the time, called “Uber for X”. But it turns out that when the honeymoon period calmed down and the bubble burst, 99% of X were dead. Looking at it, Didi and Meituan were considered to be doing well.
The platform model has a lot of investors and entrepreneurs. Zhang Hailong, founder and CEO of CODING, is one of the 6th alumni of the China-Europe Entrepreneurship Camp.
CODING founder and CEO Zhang Hailong
Zhang Hailong once worked on a platform project, “Code City”, connecting programmers and demanders, which can be understood as “Uber for Developers”. It sounds beautiful, freelance, flexible work, and it feels not far from the world of Datong. This project took about 50 million years before and after. I finally found that the entire circuit was not playing. I resolutely changed the track and focused on “CODING”.
Although the new project CODING is the same platform as the code market, it is very different in essence. The code market is similar to the “didi” in the vertical field. It does matching transactions, while CODING does the basic tools needed for enterprise-level services and is a pan-industry product. This is why CODING was eventually valued and acquired by Tencent as an important step in its 2B business development strategy.
In the view of Zhang Hailong, a pit-jumper, there are still many projects that continue to jump in similar pits. He hopes to share the lessons of his trial and error with everyone without reservation, and save a little resources for the world.
First, why is the platform concept attractive?
The platform concept is attractive to a large extent because it is very easy to achieve scale-services are basically provided by third parties, and the platform itself provides tools and supervision. For example, Didi has docked countless drivers and passengers, and Meituan has docked a large number of office workers and takeaway brothers. But Didi doesn’t own a car, and Meituan hasn’t cooked a meal. This concept is very fascinating at first.
As a platform user, you probablyLei won’t remember the name of any driver, and he won’t remember what the delivery brother is like? This is the first element of the platform model, service standardization. In other words, the goal of the platform is to make users hardly feel the difference between each service. This element is very important. It is because of standardization that consumers are loyal to the platform rather than relying on the true provider of the service.
Second, some important reflections on the “platform”
Reflection 1: The first element of the platform is service standardization
One of my early platform projects was called “Code City”, a matching platform that allowed software developers and demanders to quickly dock. After nearly three years, we finally found that the software development industry is actually not suitable for matching transactions. Because software development is non-standard, it is more dependent on service providers.
If the quality of the service varies greatly from one time to another, the customer will incur a cost of trust for each service, which will result in dependence on the service provider. In short, once the customer directly “sees the right eye” with a specific service provider and establishes a direct connection, the value of the platform as a connecting bridge is gone, and a single-order jump will occur.
When we wanted to understand this matter, we made some new attempts, such as using some tools, implementing some supervision, introducing project managers, etc. to standardize the software development process. However, these attempts do not fundamentally solve the problem of standardization.
Later, we tried to take orders by ourselves, and then thought of ways to split it, and found some scattered individual developers to solve the problem, which is equivalent to an outsourcing company, but we don’t need to hire a bunch of people. It turns out that this model still doesn’t work because the efficiency is very low. In the end we decisively abandoned the track.
So from the perspective of standardization, there are a large number of industries that are inherently unable to make platforms, such as decoration, weddings, tutors, designers, lawyers, etc., because they rely too much on people. However, these industries can still be made into yellow pages, because the need to find information still exists.
Reflection 2: The second element of the platform is the randomness of demand
When we realize the randomness of demand, we will find that some industries that meet “service standardization” are also difficult to make platforms.
For example, cleaning and laundry. These two services can be regarded as relatively standardized, and the level of service providers will not be very different. But here comes the second element, “randomness of demand.” We see that taxis and takeaways are both random needs. The time and place of each taxi you take is uncertain, and where you order the takeaway and which takeaway you take is also uncertain.
But cleaning and laundry are very fixed needs, you will not move often, so you may not need a platform to serve you, the one downstairs in the community is the best choice, even if the service is slightly worse. In fact, some training industries do the same, although trainingThe content of the platform is relatively standardized, but the training plan and training location are foreseen in advance, so the value of the platform is not great.
Around 2015, when we talk about platform models, we often think of necessary conditions as sufficient. For example, the factors that were more concerned at the time were “rigid demand, high frequency and low prices.” You may have heard of services such as cooking at home. When O2O is particularly hot, such projects can get money, or money from first-line investors, because high-frequency is just needed when eating, and it is cheap to make at home. But looking back, how unconscious this is.
First of all, it’s difficult to standardize cooking on your own, because taste preferences can lead to strong dependence on the chef himself. Secondly, its demand is not random. For example, I think this chef’s cooking is delicious, and I will choose to find him every time, so the value of the platform will be lost.
Reflection 3: Different industries have different understandings of the same element
Different industries have different understandings of the same element. For example, the standardization factor. Is the accommodation service provided by Airbnb standardized? This requires a deep understanding of the nature of each industry before judging.
Airbnb is a very special example. If you look at the “standardization of services”, the first element of the platform, you will find that Airbnb cannot work with this model. From the perspective of supply, each person’s house is different, the characteristics of each landlord are different, and the cleaning habits are different. If it is said that this thing should be provided to tourists as a standardized service, it seems impossible. From the perspective of demand, the expected experience of each guest for accommodation is also non-standard. Therefore, the Airbnb platform is actually a non-standard non-standard model.
From the second element of the platform, “randomness of demand”, Airbnb is different from renting. Although renting is also non-standard and non-standard, renting is a long-term business, and Airbnb is calculated by day. Live here, next time you can choose to live there, the needs are completely random. When the demand is in a random state, the value of the platform will become prominent.
A working platform model needs to meet both of the above mentioned elements. In general, the second factor (randomness of demand) is more critical than the first factor (standardization of services), which is why there are very few platform models that can really come out. Therefore, the platform is not suitable for all industries.
Reflection 4: Avoid being a “pseudo platform”
I believe that the real platform is matching, and the platform should not be heavily involved with either party (demand side or supply side). Otherwise, it can easily become a direct business. From this perspective, Jingdong cannot be regarded as a real platform model, while Taobao is.
After years of observation, I found that many people who wanted to be a platform have made direct sales at the end. For example, they are generous at home, their housekeeping services are quite large, and they cover many cities across the country. .
If it is a real platform model, your supply should be unlimited. This means you don’t have to pay a lot of money to bring in a lot of supply. For example, Didi and Meituan, recruiting a rider is actually very simple, because I don’t need this rider to have a high level of education and quality. As long as I am a normal person, I can train for a week after a little training.
It’s different at home. It’s not like finding an aunt to train for a week. So generous, the quality of your aunt’s service is so high that you will often miss an aunt. You will find that most aunts who use good gene at home are fixed aunts. Therefore, Haogen has made this business an extremely non-standard service, which is completely dependent on the person.
“Pseudo-platforms” are available in many industries and can solve problems, but scale is difficult and it is almost impossible to achieve “winner-take-all”. There are also some projects that will make you easily become Party B, such as a large number of “find x net”, rather than build a platform to do wedding dresses for others, it is better to make more money yourself.
Finally, let me summarize.
1) The platform is not suitable for all industries.
2) A true platform model needs to meet two elements: standardization of services and randomness of demand.
3) What a real platform does is matchmaking transactions, and the platform should not be heavily involved with any party.
4) A good platform model, in addition to choosing the right track, must also have strong execution and financing capabilities.
Three, more “platform” thinking
1. Xu Xiaonian: The core of the platform is not scale, but entry or exit barriers
Xu Xiaonian, lifetime honorary professor of CEIBS
What is a platform? First, there is interaction between various nodes on the platform, which can form a strongMetcalfe effect and bilateral market effects.
The Metcalfe effect emphasizes that the interaction between nodes of a network increases exponentially with the number of nodes. Internet companies focus on the number of active users and traffic, where the number of users is the number of nodes and the traffic is the interaction between nodes. What Metcalfe’s Law proposes is a theoretical assumption that only the most effective business model is to obtain income from online interactions.
The bilateral market effect is the mutual promotion of supply and demand, allowing a positive feedback between the two, which makes the market size grow exponentially. Uber, for example, is to maximize the bilateral market effect. There are more drivers for online car hire, so it is convenient to take a taxi. This will increase the number of people calling for cars, which will attract more drivers to join the platform.
Second, these node interactions can generate transaction value and customer value.
Finally, the core of the platform is not scale, but entry and exit barriers. Scale does not constitute a barrier. Scale is the result of barriers. Entry or exit barriers depend on the unique value you create for your customers.
However, the paradox of the platform has always existed. On the one hand, the bigger the platform, the better. It must be open, free, and low barriers. So, to resolve the platform paradox, always think about what unique value you create for your customers.
2. Chen Weiru: The core of the platform model is not subsidies, but whether there are network effects.
Chen Weiru, associate professor of strategy, China Europe International Business School
Platform thinking is different from traditional industry thinking. The two most important ideas are network effects and the structure of bilateral markets.
What is network effect? As more people in the world use the same product or platform, everyone’s effect increases. Consumers have a relationship with each other, and this relationship is to grow each other and help each other to achieve positive value-added. Of course, not all platforms have network effects, and platform models without network effects are traditional business models. Therefore, the core of the platform model is not subsidies, but whether there are network effects.
How to do network effects well? There are two ways: one is the same-edge network effect, and the other is the cross-edge network effect. It is difficult to make the same side network effect. At present, the only platforms that really have the same side network effect are Facebook and WeChat. Cross-border network effect refers to the introduction of a third side on the platform, which is required on both sides of the platform service. For example, Taobao’s introduction of Alipay’s third party benefits both buyers and sellers, which is a positive cross-border network effect.
At the same time, cross-edge effects may also bring negative network effects. Take WeChat as an example. When everyone is using WeChat, some people feel that the amount of information is too large, causing some people to choose to reduce their attention to WeChat or delete some people who do not contact frequently. In response to this, WeChat has responded by adding a “seeing” function, hoping to improve the quality of information dissemination through “seeing”. Therefore, when the network effect reaches its peak, it will turn into a negative effect. When negative effects occur, a product needs to be created to turn it into a positive effect.
In addition to network effects, another important factor in the platform model is the bilateral market structure. The bilateral market structure has changed the traditional production relationship from the original game relationship to the supply relationship.