Recently, the founding partner of Source Code Capital Cao Yi accepted an exclusive interview with a reporter from the Beijing News in Beijing. Photo by Wu Jiang, Beijing News reporter
Cao Yi, founding partner of Source Code Capital,
Investment period: 15 years,
Investment quantity: 100+, representative projects include ByteDance, Meituan Dianping, Lianjia Group, Qudian Group, Yijiupi, Meili United/Mogujie, Ziroom, Chehejia, Niudian Technology, Huishoubao, Zenjoy, Bluepay, etc.
In the spring of 2018, Forbes' "Global Best Venture Capital" list, known as the Oscar in the investment circle, was announced. People noticed two figures on the list: Sequoia Capital global managing partner Shen Nanpeng , and Cao Yi, the youngest among the selected candidates in previous years, born in 1984, founding partner of Source Code Capital.
Since Cao Yi left Sequoia China to establish Source Code Capital in 2014, people have been constantly describing him as "the young investor who looks most like Shen Nanpeng". Their common traits include accurate judgment of the track, quick moves, and outstanding performance among the same batch of investors.
But now is not the time to discuss how many shadows of Shen Nanpeng are in Cao Yi.
evaluates whether a new fund differentiated from old funds can survive. The venture capital industry has its own standards: whether the third phase of the fund can be raised.
source code first-phase fund and second-phase fund have begun to harvest fruits, and the fundraising of US$260 million in the third-phase US dollar fund and RMB 1.6 billion in the third-phase RMB fund has been completed. There should be no problem "surviving". So, a new question arises: how to face your new challenges as the source code enters the "maturity stage"?
"We have been surviving before, but now we can take a breath and look up at the sky; we were overwhelmed by what we were in front of us, and now we can spend a little more time thinking about relatively long-term things." During the conversation, Cao Yi was thinking from time to time like a "super CPU". "Data, algorithm , computing power " are the keywords he kept mentioning. If you want to catch up with the trend of institutionalization and even become the part of the industry that is running ahead, you must upgrade the system.
From "self-knowing" to online learning
When Cao Yi entered the industry in 2004, he experienced a long period of depression. At that time, the infrastructure of the VC industry was relatively poor. The Internet industry was in the early stage of recovery after the bursting of the millennium bubble. A number of VCs such as Sequoia Capital and Northern Lights were successively established.
Cao Yi found an internship as a P2P streaming technology investment analyst at Taiwan Fund C Squared Capital. During this period, he gained a lot. During this period, he met Wang Xiaochuan, CEO of Sogou , CEO of O, founder of PPS, Lei Liang, , Zhang Hongyu , and others, and accumulated personal resources for him to join Lianchuang Ceyuan in the future.
But this experience is also regrettable. The industry lacks traceable records and there is too little communication between peers. The way to explore methods and paths depends on "understanding by yourself". Cao Yi could only seize every opportunity to follow his boss and other colleagues to attend meetings, observing how they found projects, investigated, and judged.
now has methodological training within the source code every month, allowing young people to "see, see clearly, invest and help" in all aspects.
Cao Yi believes that the ultimate value of a company depends largely on what kind of magnitude and quality connection it has and how much exclusive and valuable data it has. Everyone has their own data. What he wants to do is to make these data online at a relatively low cost, so that everyone can "online" and accumulate data together continuously until a database is formed that is enough to become everyone's backing.
has two ways to accumulate data inside and outside.
internally, source code capital has established a record system with clear rules and strict requirements, and can find the ins and outs of each project from the beginning of the investor's contact with the project to the final decision making.
Cao Yi himself has benefited from such a recording method. Before joining Lianchuang Ceyuan in 2007, Wang Xiaochuan took him to meet Feng Bo, the founding partner of Lianchuang Ceyuan. Before leaving, he wrote an email to Feng Bo, which played a great role in his acquisition of the job. There is a two-page document in the email, which describes his judgment on the twenty Internet and P2P streaming projects he saw during his internship.
Later, this working method was written into the source code workflow. "It is still important to record this matter, which is conducive to later review," Cao Yi said.
For example, the source code invested a lot of resources in 2016 to discuss whether to invest in OFO, and the final decision was to give up. The system truthfully records the reason: the business model of shared bicycles is not defensive and is more suitable to be part of a larger collective such as AT and Meituan, rather than independently developing. However, there are limited buyers of large groups, so the room for project valuation is limited. The project valuation is around US$300 million and can enter the market, and if it exceeds US$500 million, the investment value is relatively small.
For external reasons, the very important starting point of source code capital when it is founded is to establish a core circle. Cao Yi wants to gather successful entrepreneurs in a circle through the "Code Club" and use his connections to leverage new opportunities.
In August 2014, Source Code Capital was established. Wang Xing and Zhang Yiming each invested 5 million US dollars. Together with dozens of LPs including Yao Jinbo , Li Xiang, and Li Yinan, gathered around Cao Yi to form the influence of the source code.
Not long ago, Cao Yi teamed up a dinner with 35 people. The participants came from the Source Code Investment Department, Meituan Investment Department, Toutiao Investment Department and Lianjia Investment Department. During the meal, everyone communicated with each other and talked about new industry experiences and new investment ideas. Cao Yi called this an "online learning".
Yijiu Chuang CEO Wang Chaocheng benefits from this "online learning". Every time the code meeting, Chaocheng takes time to attend. For entrepreneurs, on the one hand, Chaocheng has in-depth exchanges with Meituan’s business line head and even Wang Xing through the code meeting to reach strategic cooperation. On the other hand, Wang Chaocheng also had deeper cross-border communication with other entrepreneurs at the Code Conference.
allows each "CPU" to consciously optimize computing power
"provides relevant tools and methods, increase basic data input, implement iterative algorithms, and improve computing power." Cao Yi summarized his investment methodology in this way. This discourse system has been accumulated by him as a student of the 2002 class of Tsinghua Computer Department and a deputy to the "Future Road" version of Tsinghua BBS.
He believes that there are three elements for investors' cognitive improvement: data, algorithms, computing power, and three elements are the same as artificial intelligence. Data is broad, algorithms refer to the thinking methodology for forming decision-making, while computing power refers to the brain and physical strength of a person.
Cao Yi compared himself to a CPU. Now, the source code capital has about 60 "CPUs". 80% of the computing power of these CPUs is used for its own calculation iteration, and 20% of the computing power is used for "online" learning, which improves computing power in the system, thus making everyone's efficiency increasingly higher, time allocation becomes more and more reasonable, and state becomes more and more stable.
Cao Yi spent a lot of time on optimizing system algorithms. For example, in important online learning, weekly meetings, the mechanism for meetings continues to iterate. Can everyone submit projects? After submission, how long does it take for each project to discuss the most reasonable? The weekly meeting is 30 minutes, and the core proposition of the improvement of form is how to make full use of the computing power of 60 people in these 30 minutes.
Early, the weekly meeting restricted the number of projects submitted by investors. Each investor submits up to one project every two weeks. Starting from the second half of last year, the number of restrictions has been relaxed, and the quota has been moved to the time of speech. Each person can speak for a total of 250 minutes per quarter, and some analysts are specifically responsible for the timer.
By the beginning of this year, the rules were adjusted to investor assessments in the year, and look at the success rate of projects submitted in one year. For example, only four projects are submitted in a year, and all four projects have passed, and the utilization of organizational resources is much smaller than that of frequently submitting projects that cannot be approved, which is more in line with Cao Yi’s idea of “optimal computing power”.
Although this "optimization of computing power" approach initially put a lot of pressure on investment managers, it has obvious benefits. One of them is to force every "CPU" to continuously improve its "computing power", do its homework in advance, and use organizational resources on the edge. This is exactly what Cao Yi is proud of.
Expand partners to make up for shortcomings
For the source code of the "system", there is another key factor in improving algorithms and computing power - the addition of new partners. When the source code of
was established, Cao Yi was the only partner.One person has his own advantages, such as being able to define company culture, investment strategies, and post-investment style relatively efficiently in the early stage.
But one person also has pressure and temptation. How to solve the problem of Yirentang? When you are in a bad state or in a hot mind, who will check and balance you? This is the stressful part. The temptation is that if one more partner is the fund size may be larger.
Two years ago, Cao Yi was frequently asked by LP. His answer was that good partner groups are hard to come by. Don’t lay hidden dangers for development just because you need to prove something to LP in the short term.
Therefore, more than half of the source code funds were Cao Yi personally. He wants to make the characteristics of himself and the source code distinct enough and then ferment slowly. This has achieved outstanding results in source code. For example, it has reached a top position in the Internet finance industry, but due to the limited tracks that individuals can cover, source code has also missed some opportunities.
For example, starting from the Pre-A round, Cao Yi has had contact with VIPKID's three consecutive rounds of financing. Every time he felt it was good, but he also "lose the courage to shoot." Afterwards, Cao Yi summarized the reasons. Education was not the main track that he and the team at the Source Code were good at at that time. The investment in the track was not enough, which led to the delay in making a decision.
Based on the judgment of the development of industries such as education and medical care, Cao Yi made the decision to expand the team of source code partners. In the second half of 2017, after former Kingsoft CEO Dr. Zhang Hongjiang joined the investment partner, Huang Yungang, former managing director of Jingwei China, also joined Source Code as a partner. Huang Yungang's fields of expertise include mobile Internet, trading platform, corporate services and online education, which complement Cao Yi.
Now, the proportion of projects from Cao Yi among the source code invested in projects has declined. Ten colleagues in the investment department are regarded as the engine of their respective projects, and Cao Yi's role has gradually changed to a post-investment manager and a fund product manager. He focuses more on thinking about fund management portfolio and organizational establishment.
As a manager, Cao Yi is fading his youth. Shortly after the establishment of Source Code, Zhang Xingchen, vice president of Source Code Capital Investment Department, wanted to know Cao Yi's evaluation of him and see how to work better, so he took the initiative to ask Cao Yi. Cao Yi blushed at that time, and he couldn't answer the words immediately, and replied, "I'll go back and think about it and tell you." Now, facing such a problem, Cao Yi is at ease.
In Wang Chaocheng's view, Cao Yi was gentle but firmly defended the principle. Once, another investment institution hoped to take a look at the research report of Source Code Capital on Yi Jiupu. Cao Yi decisively replied, "No, this is the core asset of Source Code Capital. I won't share it, I'm sorry."
Thinking about problems from a longer dimension
After experiencing the quiet period, the period of rapid development in the VC industry and the return to rationality today, Cao Yi also has a deeper understanding of the institutionalization of VC.
In Cao Yi's view, based on the position of the invested companies in the track, the fund must be ranked in the top three in the track, so that its own existence will be valuable.
"If you do things that everyone is doing in a bunch of ways, there is not much more me, and there is no less me, and there is no sense of participation. But if you can become a pioneer in a certain industry, define its logic, and explore its value, things will be much more interesting," Cao Yi said.
In 2015, Cao Yi entered the industry for 11 years and felt that he had some spare time before he began to explore the big framework of source code investment. By 2016, the opportunities that source code is looking for and looking forward to in the top application layers of nine vertical fields were established.
In the track of source code heavy holdings, there is a type of projects that have faced considerable controversy recently - projects involving Internet finance. Ziroom is facing doubts, and Qudian fluctuates. Among the projects invested by Source Code Capital, some entrepreneurs with technology origins, such as Toutiao Zhang Yiming, Qudian Luo Min , and Ziroom Xiong Lin, have all faced controversies about copyright infringement, campus loans, and Ziroom formaldehyde housing.
how to iterate over its own algorithm when filtering projects? How to iterate over the value issues in the investment gene?
Cao Yi thought for 3 days when facing the reporter's question. He finally failed to answer directly and handed over a "middle version" answer.
He clarified that one of the visions of investing is to make people live better and how to make better allocations in demand and supply. He also admitted that "everything has two sides of a coin. We must make a more comprehensive assessment. When there are more good aspects of society than inadequate places, we must make up for the shortcomings and make the good aspects of society more and more amplify." After the
Qudian went public, it broke the issue price and its market value fell. In response, Cao Yi shared with other entrepreneurs, "This is part of the investment, you have to accept it, nothing."
He is still firmly optimistic about the overall track in the long run: the value of Internet finance lies in continuously promoting the development of the financial industry through technological and inclusive methods, and bringing money from financial institutions into the capillaries, and there is still a lot of room for development.
Recently, he also gave some suggestions to Luo Min and several CEOs of invested companies in the Internet finance field: be patient and do it deeper, heavier and slower.
is patient and thinking about problems from a longer time dimension, which is also Cao Yi’s expectation for himself and the source code. Previously, Wang Xing asked him how to think about the situation in the venture capital industry five years later. Being forced to think about long-term problems is a painful thing for young investors, but Cao Yi does benefit from it.
Q&A on the same question
VC industry has experienced a period of rapid development and has now returned to rationality. How can we seize the opportunity in the future?
Cao Yi: This industry has experienced a big explosion in the past, but the era of single-handed, skimming the water, and making big money is over. In the future, some craftsmen who focus on vertical fields may be able to get good returns, but for most institutions, if they require development, they must make breakthroughs. The trend of VCs from manual workshops to machine production, from guerrillas to army combat requires institutions to have stronger functions in all aspects of fundraising, investment, management and withdrawal, and the fund brand must be good. Only in this way can you attract good funds. Only good funds can help you choose good projects, attract good talents, and build an industry relationship network.
What do you think of the culture of source code?
Cao Yi: When the source code was established, I wrote several keywords for the culture I hope to have: self-drive, truth-seeking, extreme, openness, and curiosity. Everyone should drive themselves to do things well, rather than wait for the requirements and standards of organizational design to be pressed on before taking action. Now, through algorithm upgrades, I feel that this sense of cultural unity is forming.
Beijing News reporter Liu Suhong Special correspondent Wen Yi
