Here comes the factory manager | From personal heroes to software engineering, what have we experienced

It has been more than 30 years since the birth of the first real "commercial software" WPS in the Chinese world. When it comes to software, the first image that pops up in many friends' minds should be a bald programmer sitting alone in front of a computer and struggling to type on the keyboard. Once upon a time, individual heroism really occupied the mainstream of software and even Internet development. The story of creating a software, a website and "getting rich" and reaching the peak of life is always talked about by everyone. In fact, today, we have already stepped out of the era of slash-and-burn cultivation, and software engineering and development management have become more important. As a result, we have also gained more stable software and Internet services. In this iterative process, we have experienced Which, what are the untold stories of modern development engineering management?

Wang Yingqi, CEO of ONES, was a guest on "Relish", and together with investor Huachuang Capital partner Wu Haiyan and industry practitioners, they talked about the evolution history of software engineering, as well as the understanding and thinking of China's software industry.

During the program you will hear the following:

• Products full of personal heroism in history
• Where did the individual heroism of the year come from?
• Memories of the Times: Outsourcing Trends in the Software Industry
• As users continue to increase demand, the division of labor encountered bottlenecks
• The separation of front-end and back-end makes the division of software development clearer, and the technology stack is divided
• With the increase of users and scale, higher requirements for software stability and performance
• Programmers: From code farmers to designers, how to understand the source of innovation in the software industry
• Look at the Era of Development Whole Process Management from an Engineering Perspective
• Changes in Entrepreneurs' Cognitive Changes in Software Engineering Development Management
• Entrepreneurship enters the deep water area, where is the depth of the water?
• The accelerated digitalization process brings development opportunities to ONES
• "Social fear" programmers, how to sell software?

Guest introduction:
Wang Yingqi, founder and CEO of ONES, has participated in the core development of large-scale software such as Kingsoft WPS and Kingsoft Duba; founded Punctual Technology in 2011, and its products Punctual Alarm Clock and Punctual Calendar have over 100 million users worldwide; The fund Morningside Capital serves as EIR and participates in the management consulting work of more than ten companies in his personal capacity; in 2015, Wang Yingqi founded ONES, which is committed to providing enterprise-level R&D management tools and solutions.

Wu Haiyan, managing partner of Huachuang Capital, is responsible for investment business and fund management, and takes the lead in investment in enterprise software and financial technology. He has led investments in companies such as Tongdun Technology, Tiger Securities, 2D Fire, PingCAP, TigerGraph, Xiaoman Technology, Wish, What is Worth Buying, Xiakitchen, and Daily Fresh.

Anchor Zhu Feng, a teacher of Gao

The following is the program content, edited by CGCVC——

Zhu Feng: Today I want to talk to you about software. We talked about a lot of Internet and service products in the previous episodes of "The Factory Manager is Coming". Software is something that everyone can see, touch, and install when they touch the computer. Another teacher, Gao, who didn't come today, has a history of cracking other people's software and putting it on his website for people to download. From that time (the era of personal development), we first came into contact with software, and the software at that time was more likely to be written by individual developers.

Mr. Gao: I didn't crack the concept at first.

Zhu Feng: At that time, I felt that a lot of software was free software.

Teacher Gao: Free shareware, there was no difference between free and paid in the early years.

Wu Haiyan: This can be called own software.

Zhu Feng: I just thought about the word for a long time.

Mr. Gao: When we were young, I wonder if anyone used a software called TT to practice typing. As long as you have a floppy disk, the software is yours.

Zhu Feng: In fact, that software is also authorized. You may not care so much about the copyright issue.

Wu Haiyan: There is still a period of time when dongles are running all over the sky.

Zhu Feng: Yes, the dongle came later. Then there are talents like Lao Gao, who specialize in cracking. The most ruthless software in that array of encryption may be antivirus software, which will exhaust many ways to engage in encryption.

Teacher Gao: Antivirus software is the earliest software to support subscription, because it needs to be constantly updated and upgraded.

Zhu Feng: In my impression, after purchasing the software, you have the right to upgrade it regularly. Do you remember Lianbang Software? The Lianbang software store will copy the upgrade files for you, but will not charge you any additional money. This is not a subscription system, but just bought the rights to enjoy the genuine version. Later, due to the development of payment technology, subscription payment was created only after subscription payment conditions were established. There are also KV series written by Wang Jiangmin, which are constantly upgraded. You may have to buy each upgraded version separately.

Wang Yingqi: Back then, the subscription of the software was based on the version number. You needed to buy each version, and it issued a large version every year. Subscribing every year is such a logic.

Zhu Feng: We also do a lot of encryption technology, the most famous one is KV 300L++. There is a logic lock in KV 300L++, and if you find that you have cracked the software, it will lock the hard disk boot. Jiang Min was later fined up to 3,000 yuan for this.

Mr. Gao: Remember, WPS doesn't seem to have an anti-piracy mechanism?

Zhu Feng: Let me add some background materials. Back then, WPS was sold as an add-on to Chinese character operating systems, such as UCDOS and CCDOS. After the installation is complete, WPS is included in the bundle, it does not exist in the form of independent software. I forgot which one had the copyright at the time, their "competitor" CCED was going to buy it.

Wang Yingqi: CCED is used for watchmaking.

Zhu Feng: Actually, WPS corresponds to Word, and CCED corresponds to Excel. The experience of Qiu Bojun, the founder of WPS, actually reflects the image of the lonely hero of the individual developer we want to talk about.

Wang Yingqi: I was a user of WPS when I was in middle school and high school. After graduating from my undergraduate degree in 2004, I went to Jinshan. This was my first job. At that time, I participated in the research and development process of the latest major version of WPS, which was truly compatible with the entire set of Microsoft Office files and the development of the V6 engine.

At that time, I learned that there is a widely circulated story about Qiu Zong, saying that when Qiu Zong was doing the first big version, he was in a Shenzhen hotel, eating instant noodles, not resting well, dragging his sick body, This version is done.

Zhu Feng: Please make it all by yourself.

Wang Yingqi: I heard that is the case.

Mr. Gao: I would like to ask a question, before the software is made and sold, how does Mr. Qiu Bojun live?

Zhu Feng: Eat instant noodles.

Wang Yingqi: I understand that Mr. Qiu started by himself, and later investors from Hong Kong invested.

Mr. Gao: I remember that the WPS interface reads Hong Kong Jinshan.

Wang Yingqi: Yes, the investors are from Hong Kong. However, Mr. Qiu wrote the first version of this story in a hotel in Shenzhen, and I have not verified it.

Zhu Feng: I remember that after the 1.0 version of WPS came out, almost no one used it. Those of us who study computers may have used it.

Wu Haiyan: After WPS, Kingsoft has a series of software, and later Kingsoft Internet Security and so on.

Wang Yingqi: Kingsoft Duba, Kingsoft PowerWord, as well as Yingba player, typing pass, and later games, there is a love of swordsmen.

Zhu Feng: Yes, and then there was Xishanju, its game department.

Wang Yingqi: The name Xishanju comes from Mr. Qiu's hometown, Xishan Village.

Zhu Feng: Today is too archaeological! Back then, a lot of software development was done by individuals. Regardless of whether it is true or not, one person wrote the first version of the software in a hotel. To imagine the whole thing today is a workload that is difficult to understand and quantify. At that time, it was possible to use a very low-level programming language to complete such a large amount of work. At that time, individual heroism was very much admired by everyone.

Wang Yingqi: Back then, regardless of the development language, the complexity of the equipment, and the complexity of the user group, software development did not require so many people to do it. This is the early stage of software complexity. And the people who could program at that time had very good IQ and skills.

Wu Haiyan: At that time, only a small number of people could program.

Zhu Feng: The threshold is relatively high.

Wang Yingqi: At that time, it must be a scientist type and a person with a very high IQ to be able to do this. So at that time also gave birth to personal heroism.

Wu Haiyan: I also have a story. In 2006, we invested in an IT training company outsourcing to Japan. When the company told investors stories, one of the most famous stories was that they trained a security guard. After the training, the security guard went to Japan to earn a monthly salary of 300,000 yen. This became their famous student case.

Wang Yingqi: The emergence of more advanced programming languages makes the entry barrier for programmers lower and lower.

Zhu Feng: During that time, there were many outsourcing companies in Tianjin. I was quite surprised at the time, maybe I was still in the cognition of personal heroism, thinking that writing something is difficult, or that it will take a long time and a high price to complete it.

But at that time, I suddenly discovered that a large number of outsourcing companies to Japan have emerged in Tianjin and Dalian, and the threshold for recruiting people has been lowered. I asked why, and got the news that they will write the work module very specific and detailed through the software engineering method. Everyone is responsible for their own small piece of work, and it can be done without a particularly high level of training.

Mr. Gao: For me, outsourcing to Japan is a very mysterious field. I am very curious about what systems or software these companies are doing. Do you know?

Zhu Feng: I don't particularly know. When I arrived in Japan, I found that the software used by the Japanese is quite old.

Wu Haiyan: In the 2000s, there were several listed companies, specializing in outsourcing to Japan and Europe and the United States. Like Pactera at that time, it was a software outsourcing company employing thousands or tens of thousands of people, which was still a trend at the time.

Zhu Feng: Why was there a lot of foreign software development work in China at that time?

Wu Haiyan: As an investor, when we looked at the software market in 2006, a very important reference was Japan. First, the number of young people in Japan is decreasing. Secondly, among young people graduating from Japanese universities every year, the proportion of IT or engineering graduates is very low, and it is getting lower and lower. So there is a huge shortage of IT staff, and they have to look outside for IT talent.

Zhu Feng: Why is the IT talent gap getting bigger? In China, it's the exact opposite. Many people learn computer science and can enter big factories to get relatively high salaries after graduation. Why didn't this happen in Japan at the time?

Wu Haiyan: This has something to do with the stage of economic development. Now this situation has also appeared in Europe and the United States. We should be the same age. When we were young, the school had about 10 classes, 9 of which were science classes. Now whether in Europe, America or Japan, engineers in a broad sense are not only computer majors, but the proportion of science students has dropped significantly.

Wang Yingqi: I think there may be a reason. There may be a mature market behind this, and there is a problem with the industrialization of the software itself. Now, software is not science or research, but a fully industrialized engineering field, even in engineering, it is already the industrialization of relatively low-end engineering. Research and R&D become production and become a software production process. Once marketed, several trends may emerge. First, it requires a lot of output; second, it requires a lot of people, so the salary should be relatively low.

There is a company in India called Infosys, which is a very large IT outsourcing company in the world. Large companies including Airbus, Citigroup, and Volkswagen are all clients of Infosys. Why put IT outsourcing companies in India? There must be a mismatch in the salaries of domestic talents. This is also a clear manifestation of industrialization.

Wu Haiyan: There is also a famous IT outsourcing company in India called Tata Consultancy. Another reason is that the whole society relies on software or information science, which makes applications more and more popular, and the demand for software development naturally increases significantly, resulting in insufficient supply.

Mr. Gao: I think there is less demand for innovation in software development now. When a field all needs to be driven by innovation, the role of people is the designer, and the mind is very important. Today's software is a tool for solving efficiency problems in a given process. When there is not much innovation, people change from designers to means of production, and it is necessary to reduce costs. Software production does not require other materials, only people need to keep typing out the code. So to reduce human costs, software companies can survive. In the market, the money everyone earns is the average social profit. If the cost is higher than the average social cost, the company will easily fail.

Zhu Feng: Why are programmers in big factories still so expensive?

Teacher Gao: I think Dachang may have survival anxiety or development anxiety. To ease anxiety, be sure to innovate. So they expect better people to complete the innovation.

Zhu Feng: Big factories just happen to be able to afford this price.

Teacher of a certain Gao: Then it was inscribed.

Zhu Feng: Back to software development. In fact, in the era I lived through, I did not do software called C/S (local installation) now, but more system software called B/S (Browser/Server, browser/server mode). At that time, I felt that I was able to do everything by myself, from database to code to front-end pages.

Teacher Gao: In the company, people like you are the most terrifying.

Wu Haiyan: This is called a full-stack developer.

Zhu Feng: This was quite a common phenomenon at the time.

Teacher Gao: You are talking about the era of the rise of the Internet.

Zhu Feng: It is the era of the rise of the Internet, and this matter can be used to map the entire history of software development.

Wu Haiyan: I have to say, what you are doing may be a simple website, and you can do it all by yourself. Today's website, it is difficult for one person to do it all.

Zhu Feng: There are very few now. At that time, there would also be many people working together, but the way of dividing tasks was different from today. At that time, the task division method was, you made this channel, and I made that channel. Unlike now, you do the front end and I do the back end. It's actually a horizontal split, not a vertical split.

Wu Haiyan: Around 2010, I met some entrepreneurs with non-R&D backgrounds. They need a research and development team to make a website to provide services. For example, selling e-commerce products requires a R&D team to build a website. The founder once discussed with investors that it takes 50 R&D personnel to build a website. They think that after three years, the website will be ready and mature, and the R&D personnel can be laid off.

I said that you must be naive. There will only be more and more R&D personnel. There will be no so-called website that has reached the end, no further R&D is needed, and the R&D personnel can be laid off. If the company is still alive after three years, there may be 500 R&D personnel.

Zhu Feng: In my impression, the first website of JD.com was made by Liu Qiangdong himself. The earliest system was based on Windows system and was done with .Net.

Teacher Gao: The SQL Server used in the background is still a pirated version, and was later beaten by Microsoft and lost tens of millions.

Zhu Feng: This is also a rumor. Back then, when I was building a website, tasks were divided horizontally by channel, not vertically like today. In the past seven or eight years, the division of positions has become more and more sophisticated and complex, including back-end engineers, front-end engineers, client-side engineers, and mobile-end engineers. Everyone only does their own thing, not like the general manager of the year, who wrote it from the beginning to the end. Whether it's Internet-based or client-based, it's hard to do that. We found that there is no way to achieve the kind of personal heroism we wanted back then.

Wang Yingqi: From the beginning of the software to now, there are probably several stages of development. First, the software is tool-based, and it is a personal tool, with a single scene, a single role, and a single application to do one thing. Such as WPS documents, tables, calculations and so on. At its core, software itself is a binary mapping of the real world.

Second, after the simple scenario is satisfied, the software we are talking about now is no longer a software tool with a single tool, but a system software. In fact, it is to implement a certain scene in the real society, or some complex logic, into a software system. At this time, multi-scene, multi-role, and multi-person interactions in the real world are very complicated. The core is that the software requirements themselves become complex and require a lot of manpower.

Third, because of the advancement of programming languages, programmers no longer need to face memory itself, no need to face garbage collection, pointers... Programmers only need to write logic, and software production becomes relatively Simple.

The environment faced by software production is becoming complex and simple at the same time. Looking at the overall trend, the number of man-days required for a single software project is still increasing. Humans have biological limitations, such as limited expressiveness and inefficiency in multi-person communication. How to manage these people, organize them, and make things happen together has become a big proposition.

The software management itself, in fact, how to cut horizontally and vertically into small units can be handled by a single person or a team. This is the software industry and software management methods that we are talking about now.

Zhu Feng: From programming to managing people.

Wang Yingqi: The software itself is still partly creative. In addition to creativity, it has become an accumulation of people and days. Just like when we see a good building, its birth requires an architect or a small team to complete the drawing, and more people will be the work of "moving bricks" in the future.

Whether it's the Bird's Nest or the Grand Theater, it still needs a lot of workers to work. The same is true of software development today. Product designers or technical architects are important. But after they have designed the underlying structure, they still need a large number of engineering personnel, through man-day, through large-scale system organization to complete. This is a very big difference between the modern software industry and the previous era of individual heroism.

Zhu Feng: It's more engineering, not building a small kitchen at home.

Wang Yingqi: You can build a small kitchen yourself. But if you want to build a big hotel, you still need an engineering team.

Zhu Feng: Especially the big hotels that cannot fail. Over the years, we have invented various software engineering management methods, such as CMMI, waterfall development, agile development, TDD, etc., and invented many terms. What do you think of these nouns? Mr. Gao, have you experienced all of these?

Teacher Gao: Doing CMMI certification is usually for selling things, in order to meet the access requirements in the market, indicating that the process is compliant. In my professional experience, these have indeed been used. In order to optimize the software development process, we will draw on many external existing methods, including certified methods. But in fact, it is not very useful. It is like drawing on a special industrialized system to make special handmade things. It seems to be done step by step, but the cost will increase a lot.

The terms agile, DevOps, and CI/CD are mixed together, and people don't know what they are and what is the difference between them. What impressed me the most is that I was doing continuous integration in the company at that time, probably in 2007, 2008. At that time, there was no ready-made reference in China. We copied some ideas from Google and started to organize this internally.

Continuous integration is a bit like borrowing from the traditional construction industry or industrial way of thinking. Simply put, it is a bit like Lean Development, which was said more later, to continuously revise and test the output, rather than going to Check once in a long time. The effect is that the deviation will not be very far from the original assumption. The final landing requires a variety of inspection tools, including subversive ideas, such as anti-humanity trunk development. The practical effect is not bad, but because it is too anti-human, it is difficult to stick to it.

This type of work is usually effective at the push of the company's CTO. If the CTO leaves, it's a waste of time. Because the whole process requires a coach, and if the coach doesn't have enough energy and experience, it simply can't move forward.

Wu Haiyan: Software management and development is relatively dependent on people, and people's management concepts are needed to promote implementation.

Teacher Gao: All participants need to have the same understanding, and the coach needs to constantly correct and check everyone's behavior every day. Once there is no coach, it may fall back. We started doing pretty well. There will be traffic lights on the integration trunk, and whoever destroys the trunk will block everyone. It looks pretty good, but everyone gradually stopped looking at the lights, and still led the code to their own branches for development. We are all human beings, but no one can resist humanity.

Wang Yingqi: I share some understandings in ONES practice, we just happen to be working on collaborative management tools. I agree very much that people are often unable to fully control themselves, especially in the process of multi-person collaboration, the efficiency of communication between people makes many things impossible to do. I mentioned CI/CD (Continuous Integration) just now. My first job at Kingsoft in 2003 and 2004 was to do the "Daily Build" of WPS. At that time, Jinshan had "Night Build", which built and packaged every night. This was my first intern job. At that time, software management or project management was actually the rule of man. Although everyone has tried various management methods, many of them are not fixed.

Why is ONES a collaborative management tool for software development? We believe that after software industrialization, software production is not a manual workshop for carving, but an industrialized assembly line, similar to a car production line. From the large aluminum plate, the mold is first pressed, and then spray painting, glass installation, sofa installation, and wheel installation are interlocking assembly lines.

The software industry also needs an assembly line now, how to form a version plan, an iteration plan, and how to test when the demand comes. The work done by everyone in this process is very local, and the project manager can "hot swap" at any time. With the software development process in place, many of our customers use ONES as a production line. After customers choose ONES, they will make some adjustments to meet the production line required by the software they develop, and then start using it. This addresses the process of matching people with things.

Mr. Gao: This is an episode of a program to persuade computer majors to quit.

Zhu Feng: Haiyan, when you are investing, will you examine the company's development and management process?

Wu Haiyan: Yes. In the process of looking at the company, we have accumulated our own experience. R&D is such an important thing, we will definitely look at it. How the company organizes the research and development team, how to do research and development, how to manage research and development, will definitely see. We even take time to understand how HR recruits people. Because a well-managed company, even a start-up, must have an effective approach at every step. Effective R&D management is necessary because R&D spending by startups is often the largest piece of cost. It won't stand out among many startups if it's not managed effectively.

Zhu Feng: Have you seen any changes in the entrepreneur's perception of software engineering development management in start-up companies over the years?

Wu Haiyan: It's really not the same. There have been many changes in the past ten years. First, the type of company we look at is different from other organizations. In the early years, most investors invested in the consumer Internet, mainly for business model innovation, such as e-commerce, community, social networking, or new products, apps, and O2O services. Essentially, they use R&D to support new business models. This kind of development doesn't need to be a lot of originality, it doesn't need to be surprising, but it needs to be implemented more quickly. At the same time, many founders did not come from research and development, which is a feature of the consumer Internet era.

In recent years, whether in the medical, financial or software industries, many entrepreneurs have professional backgrounds themselves. In some industries with logic and development context, the needs of entrepreneurs for R&D are very clear. Their understanding of R&D is very different from that of the founders of the consumer Internet era, and their investment is also different. In the past two years, the founders of the companies in the intelligent manufacturing and robotics industries invested by Huachuang have technical backgrounds. In these industries, founders manage R&D teams, project processes, and code in a logical way from the early days of their business. In terms of attaching importance to R&D investment, entrepreneurs in recent years have been much stronger.

Zhu Feng: A teacher of Gao has a profound experience in entering the deep-water area of entrepreneurship.

Mr. Gao: We talked about the deep water area in an episode before. At that time, the shallow water area was the consumption of To C, and the deep water area was actually To B, especially for traditional industries.

Why do you say deep water? Mainly, decision-making mechanisms in traditional industries, including budget and financial systems, do not embrace the Internet. It is difficult for us as Internet people to adapt. Recently, I found that not only the water depth, but also the oxygen tank.

ONES does R&D process management, which is the same as the industrial PLM (product life cycle management) that I have been in contact with more now. It is mainly to manage the entire process from design to manufacture of products, starting from CAD drawings, to design and production links, including production line data, all integrated into one system.

I was wondering, can ONES also do industrial PLM? Because PLM is also a workflow, it can be flexibly customized, but the content embedded in it is different.

Wu Haiyan: PLM generally refers to the management process, process, production standard and process.

Wang Yingqi: Industrial PLM involves a lot of hardware, which may be related to BOM (Bill of Materials) and finance. We have some customers who do hardware, actually use ONES and PLM together, ONES manages the software part, and PLM manages the hardware part.

Wu Haiyan: There are very few industrial fields that do not need to write code. There must be a part of the amount of code that needs to be combined with the workflow.

Teacher Gao: What is another feature of the so-called deep water area? Traditional industrial customers are actually far away from IT. In the Internet, or the R&D process management done by ONES, users and customers are the same group. We talk about customers and users, and customers pay. In Internet companies, users are programmers who understand R&D and technology. But in traditional industries and industries, users and customers are separated. Although users are also technicians, they do not understand IT at all.

Wu Haiyan: Actually, what a teacher said is one of the reasons why Huachuang dared to invest in ONES so early. There are two types of software, one is human resource management software and financial software used by functional departments, and the other is software used by IT personnel like ONES. The use of software is a learning process. It is necessary to change the original work habits, and even the work process and work standards will change with the use of the software. But IT people manage their jobs with a new set of software that happens more naturally and easier to use.

Wang Yingqi: To continue the logic I just said, software is actually a simulation of the real society and the physical world. The more complete the mapping, the less human interpretation is required. If you sell an original wrench, it's cheap. If you sell an electric wrench, you'll need a lot less per day than if you were selling a traditional wrench. You need tools to evolve, and as you evolve less and less action to twist, you need fewer people.

Now, there is a very big proposition that all good companies are software companies, in every industry. For example, People's Daily is our client.

Wu Haiyan: How big is their R&D team?

Wang Yingqi: Hundreds of people are using it.

Wu Haiyan: People's Daily needs hundreds of R&D personnel?

Wang Yingqi: The software systems they need to maintain on a daily basis are called two micro-ends, WeChat, Weibo and the client.

We just said that all good companies are software companies, but there are many definitions of software companies. Some companies use software as a product directly for customers. For example, the game customers we serve, they sell game products to players; another example is SaaS companies, which also sell software. There are also some companies that sell products to customers through software. For example, we also have customers from FMCG companies. They use software as an internal IT process to connect with the e-commerce supply chain. It is also a software company, but it does not sell software. It may be selling ice cream, selling dumplings.

Why do our customers have so many needs? Not that they understand software well. On the contrary, in these industries, their understanding of software is not so good, and the project management itself is not so professional, so they need more intelligent tools. It's an analogy that you need a power drill without such good drilling techniques.

Zhu Feng: So is this a driving force behind your decision to do ONES?

Wang Yingqi: The source of my motivation for ONES is that I can and only do software.

Zhu Feng: ONES is now a star company. I know that you have received many rounds of financing in recent months. (Note: nearly $100 million has been raised in the last three months) What do you think you did right?

Wang Yingqi: Since we started our business in 2015, in the past five or six years, we have only done one thing, that is, we have a very large market judgment, and only we have and persisted. We believe that software companies need R&D management tools, which is our judgment. This is the judgment we made when we founded the company six years ago, and we have always adhered to that judgment. Six years later, there are software teams in all kinds of companies, and they all need tools, these market demands, to support what we have done right in the past five or six years.

Again, what ONES supports us from 0 to 1 is our cognition and our persistence in the past five or six years. But today we can change from 1 to 10, to 20, to 30, to 100, thanks to the changes in the entire industry. The first and biggest industry change is the digital transformation of the entire Chinese society, which is happening to all people and all enterprises.

Wu Haiyan: Technology drives the development of every industry. Now, whether it is policy promotion, industry awareness, or the launch of the Science and Technology Innovation Board, it is all good for this. There is a growing consensus that scientific research drives innovation.

Wang Yingqi: All industries are engaged in digitalization. Our customers include the Air Traffic Control Bureau, the supercomputing center, the welfare lottery, the banking, insurance, securities, and the fast-moving consumer goods industry as mentioned earlier. We have recently signed a lot of clothing companies, and they do internalization Process management requires a lot of software.

If these industries want to complete digital transformation, they need a large number of software engineers. This is a market change that we could not foresee when we started our business. Because of the changing needs of the industry, there is a huge demand for engineers. At the same time, it can be seen that the supply of universities is not enough, and the supply of good majors and good students to become engineers is not enough, so there are many, many talents who have transferred from other industries. In this context, good R&D management tools are needed.

Wu Haiyan: This is the situation of outsourcing to Japan in the first few years of 2000. Japan faces a huge shortage of R&D personnel, and we will also face it one day. Today our demand is up, and the supply is not increasing so fast.

Wang Yingqi: The imbalance between supply and demand we are facing today is very obvious. It can be seen that among all companies, the salary of colleagues in the software department is the highest. It is because of the lack of supply, the problem of supply and demand. Why are so many people flocking to this industry? Because the industry needs so many people. In this context, the large-scale team management needs have given us very, very good opportunities for development.

We were ready to do this in the early years, and when the industry needs a lot of talents, we have the opportunity. Why we are a little lucky today to be paid attention to by everyone, the important thing is the industry.

Wu Haiyan: China's quasi-listed companies basically list their investment in R&D as pride. For example, "We are a technology-driven company, what percentage of our revenue do we spend on R&D each year". From some large banks in China to small fast-growing startups. They list the entire proportion of R&D expenses in their income as an indicator.

Teacher of a certain Gao: I was actually thinking about a question just now, which can be regarded as a small challenge. Discuss startup opportunities with DevOps-like tools with others in the past. This is actually a small track, and you (ONES) actually include this track. I always had a view at the time that big companies didn’t really need (these tools), because big companies did it themselves. Big companies refer to big companies of the size of Byte and Didi. They must be wealthy. I did it myself.

I just heard you say that companies in many industries have R&D teams. I was actually a little surprised. This is not the same as what I have observed in the past. Maybe it has something to do with the type of company I see. I have seen that most traditional industries still use the way of procurement for informatization. Even if it is Internet-based, it may be the way of procurement. Otherwise, there will not be so many ToB companies doing SaaS.

So I had a feeling at the time that for such a tool, the market is not necessarily that big. Because big companies don't need it, or they do it themselves. You can only do mid- and long-tail, but mid- and long-tail companies don’t necessarily live long. What you said just now is very different from what I think.

Wang Yingqi: When we were doing this project at the earliest stage, we actually had a perception that once the product was complicated, it would require a lot of people to do it. When many people do it, they must need two things. The first is management methods and management tools, which are simpler.

One more thing, it's called automation. The Daily Build we just mentioned, including CI/CD, is actually automated. Now we are talking about DevOps, which is DEV and OPS, development tools and operation and maintenance tools. There is a very important logic here, which is called automation. Why automate? Automation brings several benefits. First, it can save manpower. This is not the most important, the most important thing is that it can be solidified into code by solving the problem when it encounters a problem, and the entire operation and maintenance tool or compilation tool will quickly lead it to a final correct one. place. So it's a very, very important thing.

This DevOps thing, no matter how big or small, must be something that software companies should have, and this is what we see today. And as everyone's perception of software management, software release, and software quality changes, it has become a must-have tool.

Second point, you just mentioned the so-called big companies, medium-sized companies and small companies. We believe that some large companies, including BAT, are actually traditional Internet companies. What is a traditional Internet company? I just feel that building a wheel by myself is a very cool thing, and it is still a process of dazzling skills. The new generation of Internet companies doesn't do that anymore. Why? Because the tools are very mature, there are already very mature commercial companies, so they don't need to do this. Because no matter how big or small a company is, they have a question - where is the best talent? Talent must be placed on the main business, which is a very important logic we have seen.

Internet companies after 2010, no one is keen to do this kind of thing (build their own wheels). So, they definitely need to procure external tools. So what do their in-house IT staff do? Doing the software that serves the needs of customers, docking with e-commerce, docking SCM, docking logistics, they have done these things.

There is also a very important logic that may be seen only after entering the deep water area of the industry. China's software management maturity is actually very, very insufficient. Far from enough, beyond our imagination, they need good tools to help with software management.

We now have a very large company mission - to help businesses launch products better and faster. Customers who buy our products, their software releases are faster and of higher quality. Therefore, our pricing makes it cheaper for our customers to buy us than outsourcing or self-development. This is still an economic issue.

Mr. Gao: Having said that, I have hope for our country's software industry.

Zhu Feng: A certain high teacher is full of oxygen again. We know that China Venture Capital has invested in many enterprise service companies, and ONES may be a very unique one among them. It actually serves enterprise service companies. ONES was established relatively late, around 2015. But you pitched early. I feel that the To B market was in its infancy and was not generally optimistic at that time. How did you choose Ying Qi to promote such a thing?

Wu Haiyan: We voted for software earlier. Before Ying Qi founded ONES, Huachuang started investing in software in 2014. It is not accurate to say that we invest in enterprise services, but we actually invest in enterprise software. From 2014 to the present, Huachuang has invested in more than 40 software companies, and has experienced the spring, summer, autumn and winter of this industry. We knew Yingqi at the end of 2017 and completed the investment at the beginning of 2018. At that time, the ONES product was just launched and there were some test customers.

Wang Yingqi: We had contact at the end of 2017, and at the beginning of 2018 (accepted investment from Huachuang). At that time, ONES had only a few angel clients.

Wu Haiyan: At that time, Huachuang had invested in many software companies. We firmly believe in the future of third-party independent software, and don't believe that every business has to do everything themselves. Huachuang does it by example, and is also a user of many SaaS software.

We also agree with the general trend mentioned by Ying Qi just now. The most important thing about our investment in ONES is about people. Ying Qi used to make software at WPS. R&D management software is a complex software, a big software, not a simple tool software. In addition to process management, it also involves a lot of the DevOps professional testing and integration just mentioned, and many details in the process.

In China, there are not many teams with large-scale software or complex software development experience. Those who participated in the WPS stage in the early years are likely to be suitable. The team background that has only developed small apps in the past is not suitable for such a complex software. Ying Qi's experience is rare, and the team has been working hard for more than two years at that time.

In addition, we have a mutual friend who is also an entrepreneur, and he highly recommends Yingqi. We trust Yingqi's background and experience to be able to do this, and have the endorsement of trusting friends, and we have a common belief in the future of R&D management, and a common understanding of the trend of R&D management. Several factors add to the Together, let us have the confidence to invest in ONES at a very early stage.

There is a small interlude in the investment process. Huachuang makes early investment, and the single investment amount is not very large. But when the investment in ONES was almost settled, we told Ying Qi to invest more. He said it was good, and more votes were welcome. At the last minute, we increased the investment amount.

Zhu Feng: Having invested in so many companies that do enterprise software, do you think To B entrepreneurs have any commonalities and characteristics?

Wu Haiyan: There are still many differences. Compared with entrepreneurs in the consumer field, or those with non-technical background and innovative business models in their early years, entrepreneurs in the software industry have several special differences.

In the software industry, many entrepreneurs are CTOs turned CEOs. Before starting a business, in his past career, he was usually a technical leader. There are almost no entrepreneurs in this industry who have non-technical backgrounds. This is very different from Internet entrepreneurs. Such entrepreneurs' understanding of R&D and products is natural and inherent in their genes. But how to do a good job in a company is completely different from managing a R&D team in the past. The challenge he faces is to raise money, build a sales team, expand customers, and achieve customer success.

Zhu Feng: It is useless to just bury your head in writing code.

Wu Haiyan: Yes, he wants to make money, which is not the same as the responsibilities and tasks of being a CTO or R&D leader, and it is completely expanded.

Mr. Gao: We usually call this management capability, not operation. The word business covers all of these things.

Wang Yingqi: I started writing programs since I was a child, studied computer science in college, and wrote software in Jinshan after graduation. When I first started my business, I also made an App. Before starting a business, I would and only make a software. After starting a business, I found that making software, making products and making companies are completely two propositions, and To B is different from To C. The CEO of the To C industry is the company's largest product manager. A long board pierces the sky, pulling traffic back and business back. To B is product, technology, sales, operation, financing, everything has to be done, it is really a very obvious short-board effect. Once you have shortcomings, this company will definitely fail.

Zhu Feng: Many programmers are a bit socially afraid. For sales, it may be a very big thing. And you?

Wang Yingqi: I was lucky because I stayed in Jinshan long enough. Jinshan is a company that is willing to give everyone a chance. Writing code, making products, doing business, doing marketing, doing operations, and leading projects, I have done everything. But Jinshan is not a strong sales company, so I have not done Sales. I also had the opportunity to make financial investments in between. ONES has been in business since 2015 and 2016. We started to recruit sales, and I also started to sell software. Until 2018, I got back the first one or two million yuan of the company's orders.

Wu Haiyan: At that time, the company did not have a sales team, so Yingqi could only brush her face.

Wang Yingqi: Therefore, the software sales of enterprise services cannot be sold entirely by sales. It must be the CEO who sells the software.

Mr. Gao: Just like my current company, we have an independent marketing center full of sales staff, but I still have to go out and sell software. I can't sell it, and they can't sell it either. Only "social bullshit" sales can't understand the product and can't explain it clearly. You must be the operator of the software product to quickly sign the order. What I'm breaking down now is that it's all me before and after sales.

Zhu Feng: So your sales ability has become a shortcoming now.

Mr. Gao: I am not a building-sweeping sales person, I am more like a pre-sales. The sales brought in customers, and I chatted, and it might happen. But if I don't go, it probably won't be delivered.

Zhu Feng: Did the sales over-promise?

Teacher Gao: In order to make an order, they can't wait to say that the things they sell can go to the moon. I have been relatively broken in the past two days, because I fell into such a pit in a row. Got caught by this over-promising phenomenon.

Wang Yingqi: I can share some thoughts on software sales. As a CEO or entrepreneur, how do you overcome the psychological barriers of selling software and doing sales? In the early days, I also felt that it was difficult to tell people and sell things.

Everyone has been to good restaurants, and some chefs will come out to introduce how to eat the set menu, in what order, and how to match it. Because he knows how to do it and how to enjoy it. His logic is - I'm good at this thing, for you, but you have to follow my logic, which is a very good Sales logic.

When I was selling software in the early days, it was the same mentality. When I usually go to a client company, I tell their project manager first, and then just after I finish, a test manager comes. He says I want to hear it again, and I tell it again. Just finished speaking, they said that we have a new programmer who thinks your product has some problems, can you tell him again? I say it again.

Zhu Feng: Maybe the three you are talking about are not the same.

Wang Yingqi: Yes, there will be differences. But there is still a lot of repetition. My mentality at the time was that I was telling them how a good dish should be eaten. This is what I used to get over the mental hurdle at the time.

Wu Haiyan: Tell them what the industry best practices look like. The difficulty behind is still to build a team. Because you can't always rely on you to sell alone, learn to work with a professional sales team, including how they can sell software more effectively. This may be the next hurdle to cross.

Wang Yingqi: Yes, this is one thing we are doing now. Haiyan is very, very right. What we need to do now is how to complete the replication and turn a private chef of an online celebrity shop or boutique into a chain store to sell to more customers.

Zhu Feng: Cultivate more of your clones.

Wang Yingqi: They are actually much more capable than me.

Zhu Feng: I am very happy today. I had an appointment with Ying Qi for more than two months, and it was delayed due to the epidemic, so I made an appointment to record. We were particularly nervous, because of their rounds of financing, our rounds of programs could not catch up.

We were quite confused at the beginning of this show. We know what ONES is, and some streamers are using ONES, but how to tell everyone what ONES does has cost us a lot of brains. After thinking about the direction of a lot of topics, we finally decided on the direction we are today - from the personal heroism of software to the engineering management of software.

I hope today's program will give you a better understanding of software engineering management and the best practices of modern software development. If you want to try ONES, you can register on the ONES website.

Thank you Ying Qi, Mr. Haiyan and Mr. Gao. Today's factory manager is here to chat with you all here, thank you for listening.