Nail Leaf Army: Timidity, Restraint and Progress

How a startup project of a large company continues to grow.

Wen 丨 Guan Yiwen

Editor 丨 Song Wei

It's hard to start a business within a big company

But what's harder than that?

It's not easy for a project to start a business within a large company because it tends to face more constraints and attention.

One of the reasons for the failure of the social products that Ali once regarded as the "CEO project" was "eutrophication", and there was too much attention and high expectations within the company.

The failure of the exchange made Alibaba executives realize the importance of space for start-up projects, which led to the birth and growth of the next entrepreneurial project of the "communication" team, the enterprise mobile office platform "DingTalk" - the company did not interfere too much, the project had independent office space and culture, and had enough human rights, financial power and decision-making power.

DingTalk has thus become a special existence in the Ali system, with the distinctive personal style of the founder Chen Hang (flower name: no move). Chen Hang is considered an "idealistic madman" by DingTalk employees, he has a hot temper and directness, and his product requirements are extremely strict, and he has recommended DingTalk every time, such as bun stall owners, kindergarten parents, journalists, etc.

Now, it's harder than starting a business within a large company to take over the project, and it's still such a project with the founder's personal style.

Ye Jun (flower name: not poor) is facing such a more difficult thing.

When he arrived at DingTalk in September 2020, the external situation was also more complicated: because of the epidemic, DingTalk's daily active users (DAUs) tripled in one year, followed by the challenge of product experience and use; Ali was no longer the coolest company in the past, and some companies chose to use ByteDance's collaboration and management platform "Feishu", because the management concept and methodology represented by ByteDance behind Feishu seemed to be more advanced in their eyes.

The internal situation is more severe, Ali began to brew a business responsibility system from 2020, the person in charge of each business can no longer use as much money as the previous "rich second generation entrepreneurship", and every dollar of the team must be used on the blade; the founder Chen Hang left and the DingTalk team fully integrated with Alibaba Cloud is full of huge insecurity and uncertainty.

Ye Jun recalled to us that when he first took over DingTalk, the first priority he put on was to stabilize the team, the team is everything, "It is not easy, I am taking over the entrepreneurial team, if there is another person to pick me up later, I think it will be easier, because I will hand over a very clean team." ”

"In a short period of time, more than two dozen P9s have gone, and the year after I led the team, the people who left should be the sum of the people who have walked in the history of DingTalk in the past few years." P9 is the backbone of Ali's management work, in Ali means that the rank is only one step away from the senior management position, usually their title is called "senior expert" or "director".

The departure of a number of DingTalk executives made Ye Jun feel very sorry, and the departure of several made him feel tricky, such as DingTalk's former chief architect and former CTO Zhu Hong (flower name: Yisu) chose to follow Chen Hang to start a business. The departure of the core personnel is very tricky, because only he knows which person in the team can solve the problem to which thing, "a little bit blind at once." Ye Jun said.

For Ye Jun, who has just taken over DingTalk, retention and communication means a lot of time. On the scale of management, organizational efficiency and unity of thought are often more important than getting everyone to get a balance in the short term, so he quickly arranged for someone who was more identified with the business strategy to take over the technology, rather than promoting from the original people.

"The problem I want to solve is not only the stability of the team, but more importantly, the success of the business strategy." He said.

Ye Jun has been with Alibaba for more than 16 years, and has worked in departments ranging from Alibaba Software, cross-border e-commerce, enterprise intelligence division to government affairs DingTalk. Daniel Zhang, chairman and CEO of Alibaba, commented on Ye Jun's "one line of work, one line of love."

Alibaba's park is full of Ye Jun's products, such as Xiaomi inside and outside, wireless projection code, Arirang and so on. At the beginning of the outbreak of the epidemic in 2020, the country's first health code was born in Hangzhou, which was developed by Ye Jun. A person familiar with him commented, "I don't know how to use people who have original military achievements, and I can quickly gather people with trust." ”

The flower name of "not poor" is taken from a sentence in the "Tao Te Ching", "Great profit is like a rush, its use is not poor", which is relatively similar to Ye Jun's personal style, he is not extreme, in the relationship between people, neither do things too full, nor do too insufficient, used to find a balance in the middle.

Ye Jun said, "Only in this way can I have the opportunity to achieve endless sustainability and endless innovation." ”

Ye Jun will use "timidity" to describe himself, in more than three hours of interview, he will mention "security" many times, not only to give the team a sense of security, but also to give customers, ecological partners enough sense of security.

Chen Hang has a strong product ability, and the management team is quite wolf, suitable for a project from the 0 to 1 stage, can quickly rise, do large-scale; when this product achieves a certain scale, needs fine management, and begins to consider commercialization and integration, Ye Jun's steady and progressive style may be more suitable for dingtalk at this stage.

Ye Jun needs to prove that this large company startup project, which has already occupied more than half of the market share, can continue to grow under his leadership.

Catfish in the Service Industry for Chinese Enterprises

It's time to change

In 2004, Salesforce went public as a pioneer in the SaaS (Software as a Service) industry and currently has a market capitalization of more than $200 billion among global enterprise services companies. It was also this year that many PC-based companies in the Chinese market began to explore SaaS, such as XTools, Jin target, etc.

Compared with the always-in-full-fledged to C Internet company, the to B field has experienced a long silence, until 2014, the collaborative office services company "Now Target" received two rounds of financing totaling $70 million from tiger funds, which was still a huge amount of money in the to B field that year.

Zhai Jia, managing director of Sequoia China, who has invested in the long-term investment to the B field, once told LatePost that investors who are accustomed to making quick money on the C company used to disdain to invest in B, but in 2018 he observed a significant change, and as the investment opportunities in the to C field became less and less, they began to recognize the value and opportunity of investment in the to B.

Around the same time, DingTalk began to strike around like a catfish, further activating China's corporate services market. At the end of 2017, DingTalk announced that its registered users exceeded 100 million, and it was the first to hold a product launch conference on Shenzhen Satellite TV at the doorstep of Tencent to shape the DingTalk brand.

In 2020, DingTalk expanded its market boundaries with two battles.

One is for the government-only market. Since two years ago, the "Zhejiang Zhengding" was developed for the Zhejiang Provincial Government, so that 1.4 million public employees in the province have used DingTalk for office, including the education system, the medical system and other enterprises and institutions that need to coordinate with government affairs have also begun to use DingTalk.

One is for the student market. Under the background of the global epidemic, schools need to rely on DingTalk's online classroom function to live stream classes online, including students, teachers, parents have begun to use DingTalk, a large number of people, which is a market opportunity that Chinese enterprise service companies have never encountered.

Capital's preference for the to B sector is even more pronounced after the 2020 pandemic. Zhai Jia believes that the impact of the epidemic on the B field is more continuous than that of to C, because the past way of working is inefficient, because of the epidemic, users have been forced to try some office information services, "This kind of improvement is addictive to enterprises and individuals, (they will) be dependent." ”

The pandemic has given all to B products a rare opportunity.

At the end of 2019, the enterprise WeChat and WeChat opened up, single chat, 100 crowd chat, circle of friends and other functions were fully launched, the epidemic year, the enterprise WeChat partners from the previous year's 20,000 to 80,000; Feishu initially focused on overseas markets, after the epidemic, the strategic focus shifted from overseas to domestic, in February 2020, Feishu was open to all enterprises and organizations for free.

DingTalk's data has also grown rapidly during the epidemic, before the epidemic, DingTalk's DAU was only 30 million, and in more than a month after the fermentation of the epidemic in 2020, this data has tripled and quickly exceeded 100 million. Such a growth rate is very rare for a to C product, let alone a to B product like DingTalk.

But the problem also came quickly. After March of that year, a group of users quickly lost, users poured in like a tide, and then receded like a tide, "We are all confused." Ye Jun quickly analyzed with the team for only one reason - the functions provided by DingTalk are too simple, and the value is still relatively shallow.

DingTalk is one of the very few To B products with To C scale in the Chinese market. As of August 2021, DingTalk is already a product with more than 500 million users and more than 19 million organizations and enterprises, and in Ye Jun's words, DingTalk's market size has been absolutely leading.

But it is such an industry-first platform, until the outbreak of the epidemic just did a very simple video conferencing function, at that time DingTalk did not even have wireless screen projection, online documents and other functions. Before the epidemic, DingTalk only emphasized scale, and many product functions did not, and even if there were, they were extremely simple and rough.

When the epidemic suddenly came, DingTalk took this set of "Millet Plus Rifle" as simple video conferencing system to the front line, "all rely on Alibaba Cloud, when the video conferencing system on bandwidth and server consumption is huge, aliyun most of the other projects have stopped, all kinds of machines are transferred to support DingTalk." The system was so rudimentary that Ye Jun and his team optimized two-thirds of the way a year, the same video conference, and now consumed one-third of the resources they had before.

The reason for the success of DingTalk, in addition to entering the to B field early enough, Ye Jun summarized that DingTalk has been able to quickly start the "three axes" of scale in the past few years: online approval, online attendance and salary, online communication, many companies only use these three functions. But today, the advantages of the "three-plate axe" no longer exist, because these features are also available to other competitors.

Moreover, bosses will never pay for these few functions alone, they will ask Ye Jun, "What problem can DingTalk help us solve?" All walks of life need customized functions, but DingTalk did not have such capabilities at that time, and the market needed a DingTalk that could provide users with deep value, rather than shallow services.

Nail in the coffin when it has to change.

After stabilizing the team

Organize your business and define your strategy

When a project changes the person in charge, the adjustment often takes several steps: the first priority is to stabilize the team; the second step is to sort out the business and concentrate on the work, including shutting down and adding some products and businesses; and most importantly, unifying the thinking and clarifying the strategy.

Ye Jun did just that. After the team, he closed several to C products, such as social products Dot Worm, Real such as me; many of the software and hardware developed by DingTalk in the past have also been stopped, involving SaaS products such as intelligent finance, intelligent personnel, OKR, etc., as well as time attendance machines, punch card machines, etc. The F2 video conferencing all-in-one machine unveiled at the just past conference will be DingTalk's last self-developed hardware.

The principle of closing the business is very simple: DingTalk should concentrate on doing to B, so it should "shut down and turn" the products that are biased towards C; DingTalk should be ecological, so Ye Jun believes that "we must give partners enough security".

Ye Jun's most important adjustment to DingTalk is to unify thinking and shift strategic goals from large-scale to deep value. At the same time, it shifted from serving small and medium-sized enterprises to serving medium-sized enterprises.

When he just proposed to be a medium-sized enterprise and update the indicators, there were a bunch of people to challenge, and even Ali's partners would look for Ye Jun on DingTalk and question, "Are you going to abandon the small and medium-sized enterprises of DingTalk?" Ye Jun didn't answer, nor could he answer.

"I was walking on thin ice." He said.

Ye Jun chose to answer the question with action. In the past, DingTalk's only goal was to do large-scale, the number of daily active users (DAU), DingTalk was once regarded as just a tool for the boss, because once the operation was successful, all employees of the whole company would use DingTalk, and the fission speed was the fastest.

Now that the core indicator of the DingTalk team is no longer DAU, Ye Jun has switched the indicator to revenue and the number of active enterprises.

For example, in the past, a company had a thousand people, of which 10 people used DingTalk, and the DAU of DingTalk was added to 10; the algorithm of the number of active enterprises was more complicated, for example, a company with a thousand people had to have more than 500 people using DingTalk, or more than 300 people using DingTalk and the average number of use scenarios exceeded 8, and it would be listed as an active enterprise.

After identifying the new indicators, Ye Jun did not switch immediately, and he thought of a strategy to help the team change their concepts and habits.

Ye Jun set up the "A1000 Team", which provides services to the top 1,000 large and medium-sized enterprises in various industries, and also added new industry lines including manufacturing and education.

He pioneered the pilot implementation of new core metrics in these newly formed teams, namely the number of active enterprises and revenue, because the new team did not have the inertia and burden of the past, and the trial and error rate was faster. For the entire DingTalk team, when Ye Jun took over, the indicators were still dominated by scale. It was not until half a year later that the number of monthly active software on the enterprise nearly tripled, and the pilot had a certain effect, ye Jun began to implement new indicators to the entire DingTalk team.

"There is inertia in the organization, security is the focus of team stability, before the work and brains in people's hands are not turned around, I want them to feel that their work is safe, and what they are doing before is continuous." Ye Jun said.

More than a year has passed, during this period, the number of DingTalk users exceeded 500 million, the revenue has also increased a lot, the activity of high-live enterprises has increased far more than that of ordinary enterprises, such as a 1,000-person enterprise, from 300 people to 800 people, and the average number of scenes using DingTalk has changed from the original 2 to 4.

The attitude of the whole company has also changed from the original uncertainty and doubt and worry to support and recognition.

How to make the innovative business of large companies bigger

When Ye Jun first took over DingTalk, he reached a consensus with Zhang Jianfeng (flower name: Xingyi), the president of Cloud and Technology, and the general direction was to adhere to the strategy of integrating Cloud DingTalk and deepen the value of DingTalk.

However, doing deep value and how DingTalk should do it specifically is a big problem handed over to Ye Jun.

DingTalk's customers come from thousands of industries, and each industry's demand for service is different, "Which value customers will pay?" In fact, we don't know. Ye Jun saw the current situation of China's enterprise service industry, he believes that the extreme point is to "do projects", often investing huge manpower and material resources, only serving a small number of companies, which makes it difficult to scale.

DingTalk does not want to "do projects", what it wants to do is "inclusive", and hopes to find ways to let more enterprises in China benefit from lower costs.

In Ye Jun's vision, there are two ways to do deep value, namely, organizational digitalization and business digitalization - the organization is more agile and the management is more efficient because of digitalization; digitalization goes deep into the core business links, so that the production and operation data of the enterprise can be fully circulated, reducing costs and increasing efficiency for all walks of life.

Among them, DingTalk's role is very restrained, just the basic capability platform (Paas, Platform as a service, platform as a service), only do documents, audio and video, conferences, projects and other basic products, and the rest are left to ecological partners to do. This involves two big aspects, namely the part of the basic product, how does DingTalk do itself? How to open the ecological part?

Ecological openness, Ye Jun divided into two steps: the first step, the implementation of low code, to solve the problem of applications being created in large quantities, while fully open to ecological partners; the second step, the implementation of cool applications, to solve the problem of high-frequency use of applications and the connection between applications.

Low code, popular understanding is to lower the threshold of application development, so that an enterprise employee can make the digital application products that the enterprise needs, for example, users want to shoot short videos on Douyin, and the low code of DingTalk is the editing tool "cut".

After vigorously promoting low code, Ye Jun soon found that low code alone is not enough, and DingTalk still has to operate more enterprises that do low code capabilities and SaaS products, which requires the entire ecology to allow companies that create professional software to cooperate deeply with DingTalk.

Over the past year or so, a large number of applications that meet individual needs have been developed on DingTalk. As of the end of December 2021, the number of DingTalk apps has grown by more than 5.5 times, the number of low-code apps on DingTalk has exceeded 2.4 million, and the number of all developers on DingTalk has exceeded 1.9 million. DingTalk has more than 4,000 ecosystem partners, including ISVs (Independent Software Vendors), hardware, service providers, consulting and delivery.

However, low code solves the problem of a large number of application creation, the problem of interconnection between applications gradually becomes prominent, in order to meet the needs of the deep business digitization of thousands of industries, DingTalk further proposes to do BPaaS (Business PaaS) on the basis of low code (APaaS, Application PaaS).

BPaaS is changing the way software is developed, delivered, and used. DingTalk believes that the delivery and use of software in the future must be fragmented and high-frequency. Based on this concept, DingTalk for the first time deeply opened the key high-frequency scenario - the work group to ecological partners, and jointly created a new product "cool application" together with ecological partners.

The cool application is similar to the Mini Program on WeChat to some extent, and also allows users to directly call up the application components in the DingTalk working group, without downloading and installing the application, but it takes a half-screen pop-up and does not need to jump away from the original working interface. DingTalk ecological business applications can also be pushed to groups, single chats, calendars, documents and other high-frequency use scenarios in the form of cards, compared with the traditional "people looking for applications" experience, Cool applications have realized "applications to find people", reflecting the concept of event-centric collaboration.

Ye Jun said at the press conference that DingTalk provides a high-frequency scenario of IM (instant messaging), which has paved a new path for China's software industry - using high-frequency Internet products to drive low-frequency software products. He said that on the boundary issue, DingTalk will continue to do a good job in horizontal innovation, and other software will do vertical innovation, "horizontal and vertical have intersections, but do not compete with vertical software." ”

Four battles, challenges and minds for the coming year

Although DingTalk only does the basic capability platform, the basic products including documents, audio and video, conferences, etc. will still be made by themselves. In the next year, Ye Jun will fight four battles for DingTalk's plan: polishing the three hard-core product capabilities of audio and video conferencing, documentation, teambiton project management, and exploring the commercialization of DingTalk.

Each battle has its own time frame. The document battle is expected to end within three months, "the document function is now basically the same in our three companies (enterprise WeChat, Feishu), and after three months, DingTalk's documents are characterized by connecting everything, and our integration will be very high." ”

Audio and video conferencing is the next most important battle for DingTalk. What Ye Jun hopes to solve is the most basic problem when meeting - to ensure that users can also hold high-quality meetings in a weak network environment. In his observation, the quality of Tencent Conference in the weak network environment is much higher than that of DingTalk Conference.

To this end, Ye Jun released the high-speed rail project, hoping that users can use DingTalk to stably open the audio and video conference when they travel on the high-speed rail. DingTalk also wholly acquired the audio and video platform service company Pai LeYun, the founder Zhao Jiayu will serve as the number one position of the DingTalk Audio and Video Division, Ye Jun hopes to apply the most advanced codec technology of Pai Le Yun to the audio and video conference.

Audio and video conference this battle, Ye Jun to the team's expectations are half a year to a year, this time next year he thinks that this battle has ended, at that time, users will obviously feel the immersive experience, users may be involved in the meeting "to do", "chat", "breakthrough", "video", "form" and other functions can be integrated in a card, "DingTalk's comprehensive ability advantages will become more and more obvious. ”

2022 will also be the first year for DingTalk to officially open commercialization. In the past, DingTalk was bent on pursuing the growth of scale, basically did not explore commercialization, and Ali did not give DingTalk too much profit pressure at that time, after Chen Hang would mention many times in the interview that Ali's support for DingTalk was "almost costless".

Last year, DingTalk spent most of the year in pilot commercialization, and the revenue in 2021 is the sum of DingTalk's past seven years. Ye Jun hopes that within a year, the scale of dingtalk's professional version, exclusive version, and exclusive version can be verified, and the overall commercial revenue will at least double within one year.

At present, there are three main models for the commercialization of DingTalk, including subscription-based payments such as three specialized versions, talent inventory, and low-code items, which is also the most important source of income of DingTalk, far more than half; the second is hardware sales revenue; the commission income of DingTalk's open platform can be understood as the Apple App Store model in the B field.

Feishu has a large team of 6,000 or 7,000 people, and it is hard to imagine that DingTalk, which serves nearly 20 million organizations, has less than 1,500 people. This is Ali's active choice, Daniel Zhang, Zhang Jianfeng in the communication with Ye Jun team size, they stressed that this is an innovative team should have the persistence.

Ye Jun also agrees to maintain the current number of people, which is conducive to forcing DingTalk to use a more open platform play, and working with partners in the ecosystem to build DingTalk into the entrance of enterprise digital transformation. Once there are many people, it is inevitable that it will slowly become a "project-based" enterprise service company, constantly making various products, "I think this will become a traditional company." ”

"There is a characteristic of the to B industry, if you really have to serve a good company, all the services you have to do yourself, because you do shallow meaningless, customers will have a lot of customization needs 365 days a year, I think so that every big customer you have to invest two or three hundred people." 」

In Ye Jun's computing system, DingTalk has nearly 20 million organizations today, and it is impossible to choose a "project system" approach. With the arrival of Ku Wei, vice president of Alibaba Group, who is responsible for DingTalk's key customer strategy, he will form solutions for different industries through cooperation with large enterprises in each industrial chain.

However, the products are handed over to the ecology to do, how to ensure the experience? Cool application is one of the countermeasures, trying to make the experience as unified as possible; DingTalk has also established a team of nearly 100 people to solve ecological quality problems, including operation, research and development, products, etc. 100 people in DingTalk is already a big team.

More importantly, Ye Jun mentioned that to B products have an underlying logic - the value is greater than the value. DingTalk's high-frequency products will require the same experience as to C products, and some basic products will be directly made by themselves; and for some non-high-frequency products, such as a salary system that is opened once a month, and a year-end bonus system that is opened once a year, the tolerance for the experience will be relaxed. "I'm willing to give up some experiences and give them ecology at some point, because the ecological value can exceed the value of the experience."

It is precisely because of the limited manpower that many things are not left to carry out. F2 video conferencing all-in-one machine will be the last self-developed hardware launched by DingTalk, Ye Jun believes that to do a good job, at least a team of one or two hundred people is needed, which is too heavy a burden on DingTalk, and DingTalk is more willing to choose to let ecological partners participate in doing it.

DingTalk has not made efforts in internationalization, and at present, DingTalk's overseas team is less than ten people, and the focus is mainly on the domestic market.

ByteDance is accustomed to "vigorous miracles", likes saturation attacks, feishu will release posters every time it cooperates with a somewhat famous company, and the airport also has a large-scale feishu advertisement. Ye Jun shared that DingTalk was a customer company, and Feishu specially sent more than 60 people to station around the company, "We were all shocked." ”

Ye Jun is more of a "small input, large output" style of play, knowing what he can and cannot do, this rational and accurate judgment of ability is also more suitable for the current DingTalk.

DingTalk still faces many challenges. For example, the most direct challenge at the moment is that DingTalk is still very complicated and bloated to use, but the experience problem is not an overnight thing, and it takes time to slowly improve.

DingTalk also needs to have a mental breakthrough. DingTalk is the largest head product in the industry, it is the largest, but also can serve the most industries including education, government affairs, manufacturing, catering and other industries. However, the enterprise WeChat binding WeChat can provide customers with private domain capabilities; Feishu is to some extent sought after by some startups that advocate Silicon Valley culture, which DingTalk does not have.

In the past year, One of the biggest changes in Ye Jun's concept is that when making products, the mind is more important than the function, and the function is more important than the KPI. This means that you can't make a certain function for a KPI and eventually lose the user's mind. Instead, it starts from the user's mind and makes a product that is easy to use.

DingTalk was once a product that gave customers a sense of glory, when DingTalk's slogan was "Cool Company, Use DingTalk", and today's Feishu's "advanced team, first use Feishu" is the same as this. It's just that when the breadth of the service industry is getting wider and wider, and the depth of the service industry is getting deeper and deeper, DingTalk can no longer emphasize the advanced and cool of a few people.

There is no era of enterprises, but DingTalk can still become a product of the times, it abandons some of the past problems, continues Ali's "inclusive" entrepreneurial spirit, believes that digital technology can bring change and innovation, enterprises can use DingTalk well. DingTalk can drive enterprises to change, help enterprises find and produce their own innovation, which is also the value of DingTalk.