Zhongtai has recently become a hot topic, combined with my cognition of Zhongtai, I try to answer these questions with you: 1. What exactly is the middle stage of an Internet company? 2. What are the types of medians? 3. What are the difficulties? 4. The current situation of The Internet company Zhongtai 5

1

It is said that The centralization of Alibaba was proposed by Jack Ma after his visit to Supercell, requiring the mode of “big centralization and small front desk”. The goal is also very clear: the small front desk is closer to the front line, easy to make quick decisions, agile action; Support will do the rest.

 

What exactly is the medianization of Supercell?

 

First of all, Supercell has always advocated the organizational culture of “Less is more”. The overall number of employees in the company has remained extremely lean, with no more than 400 employees at the end of last year, while the operating profit in 2018 was around 510 million euros (3.9 billion yuan). In this meritocratic culture, CEO Ilkka Paananen wants employees to be free of processes and management mechanisms and give each small team enough decision-making power. And because resources are scarce, each small team can focus more on the resources at hand and choose to do what is most important. (Supercell’s cell also means the same thing.)

 

(Diagram of Supercell’s Mid-Stage concept)

This is the centralization of the organization.

The value of centralizing organizations to games is clear. Game is originally a creative industry, it is reasonable to give decision-making power to the front desk of the business, delegate power to the middle-stage structure, and the CEO mainly to complete the support and integration of everyone’s resources.

Second, Supercell is said to have achieved a degree of technology centralization, where the front office business team can take advantage of existing technology modules and quickly and flexibly try and error. (It remains to be seen, given the game’s high complexity, how far it can be moderated.)

 

This is the centralization of technology.

The core purpose of technology centralization is cost. Cost can be viewed from two dimensions. One is the cost of wasted resources, that is, the company has three or four lines of business doing the same thing. Can the technical implementation integrate the reusable parts? The other is the time cost, where the new business is agile enough to explore using existing technology modules.

In fact, the environment to achieve the centralization of Taiwan is quite demanding. Even Paananen himself says he doesn’t think Supercell’s experience works for all companies and is unique in its own way. In his opinion, every employee of Supercell is able to set up a company outside and be their own boss, so the elitist culture and medianization are maintained in the company.

 

 

2

On the contrary, domestic Internet companies have been mentioned in recent years, or in the practice of Taiwan. Many of the small and medium enterprises I know are following in the footsteps of big companies in building their mid-taiwan capabilities.

 

However, it is obviously not a panacea for improving efficiency, nor is it suitable for a variety of companies in various fields. To exaggerate, many people talk about the centralization of Taiwan, every word looks the same, but it is not the same thing.

 

In addition to the organizational and technical centers just mentioned, there are data centers and business centers. Let’s go from difficulty to difficulty.

 

First, the technology Center (basic services center).

 

Technical centralization refers to the convergence of common technical capabilities under the responsibility of the same team to prevent duplication of wheels. It is the easiest centralization to achieve. The core value is cost reduction, that was mentioned.

 

The basic services of each company, represented by account system, have been changed in Taiwan. Taobao, Tmall, flying pigs and other businesses, express, special cars, hitch cars and other businesses, meituan take-out, wine brigade, group buying, is bound to do through.

 

Second, data center.

 

As the name implies, on the surface of the data center is the business data can get through. However, in practical use, and divided into a variety of.

 

Basic data acquisition, data warehouse building and sharing of data analysis capabilities are actually the domain of data technology center, which is the integration of technical teams that do data related work to support each business. The core value is cost reduction.

 

The data integration, data sharing and collaborative application of various business lines belong to the scope of the middle stage of business and are led by business objectives (for example, the premise of Ali 88VIP membership is user data integration).

 

Sometimes the two are coupled, sharing both technical capabilities and business enablement. However, in China, the cognition of data analysis is still relatively shallow (think that data are pulled statements), which often has poor effect. The former has indeed landed in many Internet companies.

In some companies, the data center just pulls data together in a superficial way without solving any problems or adding value to the integration. Xie Chunliang, director of Middleware architecture of Ali Cloud, once said that the data center of many companies is to process the data into a “large screen”, and to make a “large screen” for display is to realize the center. That is, if the data center does not serve the business, or if it serves the business center, it has no value.

 

Third, business in Taiwan.

 

Since technical modules can be abstracted for reuse, can business modules also be abstracted for reuse? The marketing strategy that had been used when taobao, can move to fly pig to use directly? Can the private car strategy be applied directly to the express train? This is the concept of the business middle stage.

Obviously, it is more difficult because of the higher complexity of the business.

 

The middle stage of business is the ultimate result pursued by major companies: agile promotion of front business, stable support of back business. However, the real situation is often contrary to one’s wishes, difficult to be satisfactory, truly achieve good business in Taiwan, very rare.

 

Fourth, the organization of central Taiwan.

 

As mentioned earlier in Supercell, it is a mid-stage structure that is grafted on the basis of the first three mid-stage structures, and then adds “decentralization mechanism”.

Alibaba’s business support ability in Taiwan can support the rapid innovation of new e-commerce related businesses; Bytedance’s growth capacity is moderated. Toutiao’s growth force can be quickly invested in Douyin, or even invested in various overseas products in various countries. Meituan has a trial-and-error team that uses existing technical capabilities, user traffic and other resources to quickly trial-and-error and find new directions, including food delivery.

These are the embodiment of the organization. Organizing the Center is not so much about the center’s capabilities as it is about the space for intrapreneurship. Its core value should be “whether the old business can help the new business”, which is a test of organizational ability. Some companies are in-house and insular, no different than this one investing in a new team.

 

3

There are several key contradictions in the difficulties of the Central Taiwan:

  • The mid-taiwan team is far from the business

  • It is difficult to prioritize the demands in the middle

 

First, the mid-taiwan team is far from the business.

 

Why is the difficulty of Taiwan: technology – data – business – organization? Because in the future, the more need for business team involvement, the more need to have business awareness to achieve the middle stage. And Taiwan team, nature is far from the business.

 

The cooperation between each company and the Center team has great inefficiency. The front desk business team must communicate with the center team repeatedly to clarify the background of business requirements. In large companies, cross-department cooperation will increase costs. In addition, for the mid-stage business team, it is far away from the business for a long time, with no growth and inaccurate judgment of demand, which makes it difficult to gain the trust of the front desk business team.

 

Over time, it has become a pure support team in Taiwan, passive state of demand, become “chicken ribs in Taiwan”. The bigger problem is that the value of a mid-stage team is to integrate requirements and abstract capabilities, which can be particularly hard to do without a strong understanding of users and the business. If there is a conflict between the needs of multiple front desk businesses, the central team may not be able to do a good job of coordination, and multiple parties need to repeatedly communicate and wrangle.

 

There are two common solutions to this problem:

 

First, the middle office responsibilities are directly assumed by the largest business party.

 

Like the shell of the Taiwan module, is advanced by the second-hand house; Most of gm’s passenger driver product processes are also initiated by Didi Express. The largest business parties are close enough to the business to make their own decisions about most requirements and be more efficient.

 

In this way, instead of “trying to find a middle-taiwan team to collect people’s needs”, it has become “big Brother to share his own capabilities”, and it is still up to the respective teams to make decisions and implement how to use them.

 

Second, let the Taiwan team do business closed loop.

 

This applies only in a few cases where the middle – station business itself can be self-closed – loop. Common examples include companies’ account systems (including registration and login), Alibaba’s payment and order processes, and Didi’s maps and navigation. These businesses are not closely related to other businesses and can be decoupled. They can complete the cognition and decision-making from users to businesses by themselves, so they can be independently done by themselves. When other businesses want to use them, they can connect with the general interface.

 

And most middle-office teams are responsible for businesses that are not so easy to self-loop. For example, it is difficult to abstract the order receiving process of different business lines of Didi and let the middle team directly do it by themselves in a closed loop. For example, the commodity part, order process part and payment part of different business lines of Ali can be abstracted. However, in terms of user experience, it is difficult to abstract due to the great differences between different front businesses.

 

Besides, the priority of the demand in Taiwan is difficult to sort.

 

The real problem is that mid-stage teams are faced with countless demands and priorities. It’s hard to have a standard for how to deal with it.

 

You can say just use business metrics, such as which front end business needs create more business value. This is certainly possible, but it runs counter to the role of “Centralization is for the front office business to be more efficient, especially for innovative business” mentioned above. Because it’s quite simple. When you have few users in your innovation business, you’ll never get support from the middle office, so make a yarn.

 

As a result, if the ranking is strictly in accordance with the quantitative benefits, the personalized needs of small businesses and innovative businesses will never be ranked, and they will have to finish by themselves in the end, which is “de-centralization”.

 

There’s no good solution to this problem right now. When I talk to people in every company, I will almost encounter complaints from small businesses or marginal businesses that they are not supportive enough and their demands are blocked. In addition to doing their own, can only hope that they need, the central department has provided.

 

 

4

Based on my interview and understanding, I can make a simple description of the current situation of the Chinese Taiwan companies.

 

Since Tencent proposed the centralization of Taiwan more than half a year ago, TEG has been leading the implementation of some integration of the underlying technology, mainly account system. Data consolidation is also being done, but there should be more low-level consolidation like data warehouse construction. In terms of business, search, advertising and information flow are all being integrated one after another, basically in the way of appointing a certain company, led by browser, wide point communication and FCC (information flow and content community business line) respectively. Wechat remains an enclave, independent of any Chinese influence.

 

Tencent is supposed to be the most difficult, as we all know, with mountains and horse racing. Some of the new business is being done by countless teams in a completely de-centralizing way. A friend of mine is in a group that works on new projects that are very secret internally for fear of being a target, but not so secret externally. It’s magical.

 

Ali’s business in Taiwan should be the best. This diagram, which is circulated on the Internet, better describes the situation of ali Middle platform architecture.

In the middle “business platform”, each Center undertakes self-closed loop business, which is the second solution to “far from business” mentioned above. In the e-commerce environment, the user needs and experience of all links have a high degree of similarity, and the organizational culture of Ali is sufficiently efficient, so alibaba’s Central Taiwan department will not have too many “central Taiwan chicken ribs” and can deeply participate in the business.

 

Alibaba’s User Center, for example, is a unified account management department responsible for building basic capabilities such as User portraits and User tags, as well as front-end experience (such as “My Taobao”). The Item Center is responsible for the product details page, with both the information logic on the back end and the experience on the front. Whether it is “my Taobao” or “commodity details”, the central department is to provide a complete set of basic capabilities, by Taobao, Tmall and various front desk business departments to personalized use (can be simply understood as skin change).

 

As mentioned in the middle office of Shell, it is led by the second-hand department, which abstracts the business modules faced by second-hand and other businesses (housing management, CRM system, broker guidance/follow-up ability, contract signing, etc.) and provides them for everyone to use. The Taiwan department is not isolated, but strongly coupled with the business department.

 

In addition to basic services, Didi will have a platform product department and some support departments (basic information, risk control, etc.). Many business modules of Kuaidi are actually middle stage (such as providing order receiving process and marketing tools). Although online car hailing is a core business, it is relatively decoupled from other businesses (single bike, designated driving, hitch ride, internationalization and car service). There are not many modules for external Localization, and they are mainly some basic services.

 

Bytedance’s media center is dominated by three departments: user growth (focus on advertising), technology (focus on algorithms) and commercialization (focus on advertising). Recently, bytedance added live media. It is obvious that these businesses have strong reusability and sufficient business independence in all directions. This would be similar to Alibaba’s Zhongtai, allowing it to penetrate the business while maintaining its independence. Toutiao’s organization pays great attention to efficiency, so it is supported uniformly and the results are good. For example, the user growth team controls the budget, is involved in many businesses as BP, and has decision-making power over many product features, as opposed to some companies where user growth is really just marketing.

 

Meituan rarely mentions its Own China strategy or capabilities. Internally, the basic service center and data center are built. For the middle stage of the business, the differences between Meituan business models make it difficult to abstract (group purchase – takeout – wine travel – logistics – new retail – bicycle – online car hailing). However, The organizational construction of Meituan is very friendly to new businesses. Every new business is supported by the old business (of course, the premise is that the data itself is good enough), whether it is the support of the underlying technology business, human resources, or traffic support. Therefore, it can be said that Meituan has a certain organizational culture in Taiwan.

 

After jingdong put forward the idea of centralization, it has also set up technology centralization, data centralization and some business centralization (such as search centralization and marketing centralization). Similar to Ali in structure, but relatively inefficient, the business side and the cooperation is not smooth, there will be all kinds of problems mentioned above.

 

On the whole, there are several characteristics in the middle stage of each division:

 

  • Basic services (account, payment, security, risk control, etc.) ability of the middle stage, relatively very complete. Technology in the basic more or less in the construction of the penetration of the degree of business have different.

  • Most of the data centers are under construction, but they still focus on providing “collection, storage and data collection” capabilities. Although the r&d cost is reduced, cross-department collaboration usually increases the cost of collaboration. Such a center is far from the data center that can provide insights into business decisions. The real general data analysis work is the closed loop within each business, their own completion.

  • Business depends on the situation in Taiwan, some companies between the natural business is not suitable to do in Taiwan. For suitable companies, the most common way is to adopt the “big brother” approach when it is difficult to decouple the closed loop. The core business provides support, such as shell and didi. The other is the chance for China Media to take control of its own business, such as Alibaba’s e-commerce media and Bytedance’s growth media.

 

Finally, I would like to make some summaries of the question of centralization.

 

  • In Taiwan or to follow the “specific problems, specific analysis” of the truth. There is no universally applicable zhongtaihua strategy, some businesses do not need zhongtaihua, and some Zhongtaihua will be counterproductive. Some companies have only one business at all, or there are several businesses have no repeated wheel phenomenon, that hard to learn ali’s middle stage, must be a monkey. See Ali Xie Chunliang’s interview, including I with Ali’s friends to understand, Ali’s middle stage is also the “pain extremely think change” process, in stumbling into growth, not the boss on the order of all in place.

  • The goal of Zhongtaihua is to reduce costs in two ways: first, the merger of duplicate wheels between existing businesses to reduce costs; Second, provide capabilities for certain businesses, especially new ones, that allow rapid trial and error at low cost.

  • For some large companies, it is not a technical issue, but an organizational issue, that is, should each line of business accept what others offer, or should I do it myself? And whether the new business can be supported by the existing capabilities of the old business.

  • For product managers and operations, it is important to join a self-closed loop team rather than a mid-stage team with pure support. It’s easy to become a “chicken in the middle” if you stay away from your business. For example, in many companies, in the middle of basic services, there is no such business post as product and operation, it is all technology.

 

 

I hope I can help you.

 

 

Reference: enterprises have followed the example of Ali to build zhongtai, in the end is blindly follow the trend or do not do will die?