“Kill the tough village and fight in a daze.”
Editor’s note: This article comes from the WeChat public account “ GGV Jiyuan Capital “(ID: GGVCapital), author GGV Jiyuan Capital.
“Ah, the company computer has crashed again!”
“Let ’s find IT brother.”
IT brother, one of the most mysterious brothers in the company, but their daily work is more than just repairing computers. IT operation and maintenance can ensure that the enterprise “continuous network, system is not paralyzed, data is not lost”, is an indispensable part of enterprises in the information age. In the past two years, in addition to IT operations management (ITOM, IT Operations Management), a new field has emerged-intelligent operations and maintenance (AIOps, Artificial Intelligence for IT Operations).
Simply speaking, AIOps uses big data and artificial intelligence technology to analyze the data generated by IT operation and maintenance, so as to achieve fine-grained alarming, fault location, fault prediction, and index anomaly detection. Qingchuang Technology since 2016 is a technology innovation company focused on this. They are also the first software company in China that focuses on enabling artificial intelligence and big data technology to enable traditional enterprises to operate and maintain.
In this issue of Entrepreneurship Insider, we have invited Yang Chen, founder and CEO of Qingchuang Technology, who will talk with Kuang Yingwu, senior investment manager of GGV Jiyuan Capital, to talk about this seemingly obscure but everyday Closely related fields.
Lily:
Mr. Yang, first of all, can you please explain the intelligent operation and maintenance in a language that laymen can understand?
Yang Chen:
Yes. Let me give you a common example. Now that we are in a huge epidemic, everyone will care about their health. But the problem is that people will definitely get sick when they eat whole grains. When we get sick, we can go to the doctor and do various examinations, and these examinations will bring a lot of data. The ordinary people cannot understand these data, and we have to ask a professional. Doctor.
The essence of operation and maintenance is a doctor who can quickly help enterprises and users find business problems in operation and maintenance data. Traditional operation and maintenance is often more convenient when problems are found, but there is no good way to analyze and judge problems. Intelligent operation and maintenance can use big data and artificial intelligence algorithms to determine and solve these problems.
Lily:
Which industries are the main objects of its application? What is the size of the company?
Yang Chen:
In fact, it is a bit inappropriate in terms of scale. A more appropriate way is to see how far the enterprise ’s digital transformation has progressed, or how much its business depends on information technology to run. Generally speaking, any industry, as long as its business and information technology are veryClose relationship, relying on information technology to support the smooth operation of business, it will have the corresponding operation and maintenance needs.
It is at this time that it feels that the human cost of operation and maintenance is so large that it is difficult for enterprises to bear. How can we reduce costs and improve operation and maintenance performance? This is the key reason why intelligent O & M needs to be deployed. Therefore, regardless of finance and telecommunications, including the energy industry, intelligent operation and maintenance may be required.
Lily:
In 2016, Gartner officially defined the concept of intelligent operation and maintenance, and predicted that by 2022, 40% of large enterprises in the world will deploy intelligent operation and maintenance platforms. Qingchuang Technology was established in early 2016, and quickly launched products, showing that Qingchuang’s global layout is relatively early. I would like to ask Mr. Yang, why anchored this track so early?
Yang Chen:
Qichuang has a special place. Our entire founding team, including myself, have deep operation and maintenance management genes. We have 15-20 years of experience in this field, and have served the world including domestic Some very large customers. In this process, especially in 2014 and 2015, when I was at BMC, I saw that a large number of companies we served, such as UnionPay, Bank of Communications, and Huawei, had various O & M problems.
The main problem is that due to the increasing business volume and the complexity of the business, the operation and maintenance work load is very large, and the traditional operation and maintenance management tools are not equipped with some intelligent algorithm capabilities due to the backward architecture, and rely heavily on manpower The increase in efficiency is very low, which puts a heavy load on the enterprise.
The company itself felt very painful, but it was powerless. At that time, as a large international manufacturer, we were unable to help the company solve such pain points, which was very abnormal. This field should have corresponding innovations in order to cope with this vigorous digital transformation.
Large enterprises still have such problems, not to mention some SMEs. They are tortured because it is more difficult for them to find excellent operation and maintenance personnel, and it is more difficult for them to bear the cost of these operation and maintenance personnel. So around 2014 and 2015, we decided to use some new technology to change this industry, that is, at that time, we found several like-minded partners to aim at this track together.
In fact, at that time, intelligent operation and maintenance could not be called a track, because Gartner officially defined this track as intelligent operation and maintenance at the end of 2016. When we first started a business, because there was no concept of intelligence, we put forward a very preliminary idea at that time, called to build an IT operation and maintenance management software that can analyze and thinkThis coincides with Gartner’s prediction.
Kwong Ying-wah:
For operation and maintenance, I will give you a more intuitive example. We all have computers, personal computers will have various failures, if in the company, we will go to the company IT. Today, whether we watch videos, remote offices, or meetings, these systems are actually run on the server. A server is an architecture that is many times more complex than a personal computer. It may involve databases, storage, networking, backup, and even new technologies, such as containers.
Operations and maintenance personnel are required to repair the database or the server is broken, or to maintain the server in a relatively healthy state. In the past, O & M required O & M personnel to go to the server, pick up problems in various data, and then fix them. Intelligent O & M is to make it easier for O & M personnel. It can judge systematically, or use data to find the problem, and help O & M personnel better fix the problem.
Lily: When there was no intelligent operation and maintenance in the past, how to solve this problem?
Yang Chen:
Let me give you the most classic example, everyone will use online banking, but no consumer may think that when you transfer a sum of money to another bank account, this business application actually goes through countless systems and transaction links . How did you solve this problem? Once you find that a transfer is slow or wrong, you will complain. At this time, the bank is messed up. It needs to gather many departments, such as server department, network department, security department, and equipment management department. Of course, there is also operation and maintenance. Departments, and perhaps upper-level application departments, with a total of more than 10 people, gathered in the same place as a wartime meeting room, everyone began to investigate.
The efficiency of this investigation is very low, because no one is willing to take responsibility, but we have no way to see end-to-end, the impact of the entire transaction link status and the underlying infrastructure on its support. Therefore, diagnosing and analyzing problems are often very time-consuming and labor-intensive, and the emergence of intelligent operation and maintenance effectively solves this problem.
First, it can collect the data generated by the technology of the entire transaction-related link on a platform in real time, and then use the algorithm to observe the abnormal patterns of these data, and then compare the abnormal patterns with the normal patterns to help A person who is not very skilled in finding the crux of the problem. This makes the root cause that it took ten people to spend a few days, or even more than ten days to diagnose. Now it may be found by a person within a few minutes, and even the skill requirements of this person are very low, so the ability to reduce costs and increase efficiency is Very strong.
Lily:
Which players are currently offering such services on the market, and what is the position of Dynatron?
Yang Chen:
Our earliest layout and productization in 2016, so as of today, we are definitely the leader in the domestic market. Whether it is the maturity of the product, the accuracy of the upper-layer algorithm, or the richness of the end-to-end scene, we are second to none. Of course, we are not the only player in the market, because Gartner continues to promote the market after the end of 2016, and tells everyone about this trend. Intelligent operation and maintenance must be the standard embodiment of the next generation of operation and maintenance, that is to say, 5 years After the year, there will be no traditional operation and maintenance, all will be intelligent operation and maintenance.
Therefore, in the second half of 2017, especially in 2018, there was a swarm of bees. There are actually some competing products in the domestic market, but so far these competing products are still far from Qingchuang. In comparison, in fact, we are still more competing with players in the international market, such as IBM, Splunk, HP and so on.
Lily:
International giants have more adequate resources, manpower, and even a better customer base. How do you face the challenges posed by giants?
Yang Chen:
I have been working in a multinational enterprise for almost 20 years. I think the giant has one of the biggest problems in China, that is, it will not really care about the needs of the Chinese market, and it still spends more energy on these European and American traditions. in the market.
This brings up a problem. When the products of all international giants land in China, the pain points for Chinese customers and the scenarios involved often have mismatches. Compared to this mismatch, Qingchuang may be able to bring greater value. First of all, we better understand the needs of Chinese customers and understand the pain points of China’s operation and maintenance management, so we will cut the knife International players are more accurate.
Second, a very important factor is the word smart. In essence, intelligent operation and maintenance or any kind of intelligent industry cannot do without the richness of data samples and the scale of data volume. For these two points, the Chinese market is very friendly, because we can get very good data samples from our customer base, which allows the product’s algorithm model to be trained with high quality, which makes us accurate The accuracy can quickly climb, surpassing some international players.
In our industry, or other fields (including visual processing, language processing, and various AI fields), China actually has world-class leadership, which has a lot to do with the richness of our data samples and the domestic demand market. Because we both understand customer needs and have good scenarios, plus our underlying algorithms are more accurate than international counterparts, so we are not afraid of the challenges posed by giants. Perhaps the only thing we lack now is our scale Not big enough.
Because the giants have been in China for a long time, they may have more sales force and more people, but I believe this will be changed in the near future. In addition, after 2014, the state also has certain support policies for technological innovation companies like us, and also encourages our major customer groups, such as traditional banks, insurance companies, and manufacturing industries, to try more like our technology-based Innovative enterprise cooperation also brings us more opportunities. The maturity of this piece of technology is definitely in the future, so this is why I think we have the confidence to defeat these international giants.
Lily: Yinghui, what do you think of this market? It has traditional giants, but at the same time it is a new opportunity. What are the opportunities and challenges here?
Kwong Ying-wah:
First of all, I think that the overseas market is quite different from the Chinese market. When helping us optimize the efficiency of corporate operations and maintenance, the communication between the US teams mainly relies on emails, while in China, a lot of it is through WeChat. This is reflected in the product design, possibly these foreign operation and maintenance management tools, how to do things including process approval through mail, but this model cannot be used in China, so this must be based on the Chinese market Do some customized development.
In addition, we actually see some systems that are commonly used by everyone in China, such as the railway ticketing APP12306. In the past, everyone went to buy tickets on holidays, and found that the system could not log on, or the ticket issuing system was particularly slow. People also have a lot of pressure, and a large amount of traffic is concentrated at a time point, resulting in a particularly heavy load on the system. How to repair it at this time is also an advantage of Chinese local companies, which can meet some O & M needs with Chinese characteristics. Lily:
For example, a company like JD.com, before there is no intelligent operation and maintenance, what size should his operation and maintenance team be? If there is intelligent operation and maintenance, how large should the operation and maintenance team be?
Yang Chen:
I give an example, a domestic first-tier insurance group that we have served.We used to do the intelligent operation and maintenance of his database. From a large team of 32 people, it only takes 2 people to work full time in a year. This is a very large cost change. Of course, it is not that the 30 people disappeared. The key to the evolution of information technology is that digital transformation requires a large number of IT talents. If these talents are deeply trapped in this well, It is not possible for companies to generate income. In other words, the same IT talent, if I put him in, for example, the design and development of business, it may be more effective.
Kwong Ying-wah:
I think there is a particularly good analogy. You can look at the total number of employees in your company and the number of IT staff. For example, GGV, we have less than 100 employees worldwide and 4 IT staff. Our ratio is already high. Relative to a team of 100 people, it may be common to allocate 2-3 IT, 1000 people is 20-30, and 10,000 is 200-300 IT. For some large enterprises with more than 100,000 people in China, thousands of operation and maintenance teams are often needed to ensure the smooth and normal operation of the system. This is an exaggerated number.
Lily:
I noticed that there are many financial institutions in your customers, such as UnionPay, Bank of Communications, China Development Bank, Founder Securities, etc. What are the financial industry’s IT operation and maintenance management needs? To serve these financial customers, are there any other entrepreneurs need to pay attention to?
Yang Chen:
The financial industry is basically in need of IT management. To give a few classic examples, the first one, for example, all systems will be alerted when there is a problem. We see that there are some credit card financial companies, and it may have millions of alarms at the same time. Where is such a large amount of alarms, where will there be so many people to deal with them? So there must be a rigid demand among them. Can we sort out the ones that really need attention in a large number of alarms in the first place? Identify related alarms and alarm scenes, so as to directly find the root cause during post-processing.
This is a very classic intelligent operation and maintenance scenario, because traditional operation and maintenance can only organize these alarms into a big package, but in fact they have no ability to find out which alarms are worthy of attention in the first place, but rely on people Judgment, so it takes a long time for investigation and analysis.
Another, when a business fluctuates, it needs to be cleared first, what caused this fluctuation, is it the system? Is it a safety issue? Or is it caused by an error in a piece of code in the application itself? This is also important for intelligent operation and maintenanceWhere it works. In general, the most important thing for intelligent operation and maintenance is to improve the ability of traditional operation and maintenance personnel to understand the various data appearing in the business operation process, rather than relying on the experience installed in the operation and maintenance personnel.
Experience itself has limitations, and experience is very dependent on people. If the number of people is not enough, experience is difficult to work. In addition, experience requires mutual cooperation to achieve, and cooperation between people is often not very efficient. So these are the problems that intelligent operation and maintenance can effectively solve, and it is also the reason why financial institutions are particularly demanding in this regard. Because the business of the financial industry is fully digitized, its business cannot be problematic. It’s called Mission Critical. Once something goes wrong, the social and economic impacts are huge. So it will hope that this revolution will come first in terms of operation and maintenance to improve its effectiveness.
Lily:
You have many years of experience working in a large company. From a large company to a startup company, what are your biggest challenges?
Yang Chen:
The challenge is very big. Although I have worked in some international companies for many years and I have also done many years of management work, the biggest feeling is that managing an entrepreneurial company and managing a large workplace team are completely Different. A start-up company is a blank piece of paper, and even the employee manual has to be written by yourself. As you thought, all the things that should be agreed upon are completely absent, and everything must be rebuilt by hand.
The second is team building. In fact, I think that entrepreneurship is essentially a result of belief, and it must be a common belief of the entire team. This is relatively unclear in the workplace. Especially for people like us who work in multinational companies, you tend to work in the region, so in a state where the entire enterprise organization has been very strong, everyone will not be The vision, or the so-called common belief, is so clearly understood.
But in a start-up company, you find that you have to rely on faith when you are going through any hardships and any difficulties. It is up to everyone to believe what they were originally doing. So this is actually completely different from when you were working, and you also encountered many difficulties in the process of starting a business, including the difficulty of acquiring customers at the beginning, coping with unfair competition, etc., which are situations that you will not encounter in large companies.
Lily:
Financial customers have the highest security requirements. How did you manage these financial customers? What’s wrong with QingchuangWhat?
Yang Chen:
There are two main reasons for this. First, it is indeed related to my years of experience in business development. I have a relatively deep understanding of some of the head financial customers themselves. I used to deal with these customers when I was at BMC or HP. They will also be surprised that you give up work in a foreign company and suddenly become an entrepreneur in a small company with less than 10 people. What must have ignited you, even though he said to you now, including your ideas about products, Maintain a suspicious state, but he has a certain degree of trust in you, so this is helpful.
Second, you have to be willing to work hard, such a large financial company, It is impossible to let you replace BMC at first. He must pick some problems that cannot be solved by large companies, or temporarily I can’t see the place of hope for you, see if you can let him see hope and change in these pain points, and whether you can grow a little crop on this saline land.
Therefore, it is very important to set up a business with one size fits all. You must be in the first place to really reach the pain points of the customer group in the positioning of the product. Even if you are not the best, but as long as he sees hope in this pain point, he will surely realize that your path may be the future, then he will give you more opportunities. Several important customers in our initial period were basically acquired in this way, which is to do “salty ground work”.
Lily:
This shows me the spirit of Chinese entrepreneurs. When we face the challenge of giants, we do n’t use the methods of the previous generation of entrepreneurs. We do n’t use relationships or pull channels. We really use them. Products and technology impressed him, and he planted fruits on the land that big companies were reluctant to cultivate.
The official introduction of Qingchuang mentioned that the current product repurchase rate can reach 100%, how is this done?
Yang Chen:
There are several prerequisites for a product repurchase rate of 100%. The first is related to the business model. I also discussed with Yinghua about SaaS before, whether medium and large-scale enterprises can use the subscription system. In fact, whether customers use a subscription system or not depends on the products or services you provide, and whether they have a sufficient price / performance ratio to beat project-level companies that can continue to be redeveloped or customized. If the price / performance ratio is not high enough compared to a project-level company, then he will not buy your product at all, so why not talk about repurchasing.
Second, can the customer use it?Your SaaS, or subscription period, enjoy some additional value added by crowdsourcing. Because we are an intelligent operation and maintenance company, in essence, we not only let customers use products, but also in the process of using, let multiple customers continue to converge on the product platform for data analysis best practices.
In this way, for any customer, he actually enjoys the best practices used by many customers in the same industry, so that he is willing to continue to buy your products. Because of this, he can continue to enjoy the experience or best practices brought about by sharing.
Once he bought out this product, he only got one product with no vitality, but those “best practices” that are constantly being supplied are his real value demands. So I think this mode setting is very important, and it is also a very important reason why we have insisted on going to this day. From the present point of view, although we did it earlier, we can consolidate this model and let it grow healthily and effectively, so this is also a part of what we think about when designing business models or products.
Lily:
British, all enterprise-level applications are on the commercialization path, but the amount is actually not fast. What is the real reason that prompted you to make this investment decision?
Kwong Ying-wah:
We see that some customers have been using this product for two to three years, or even longer. We also see that this product is constantly iterating, improving, and optimizing, and customers continue to pay. The company dared to tell the head customer that signing a one-year contract means that the customer has a very strong autonomy after the one-year expiration, they can choose to continue to pay, or no longer use it.
But we found that all such customers chose to renew the service contract with the company, so the company’s products were verified on the customer side. We also see that the company ’s large customers are selling, the number of sales is very small, and the sales efficiency is very exaggerated, so we think that after investment, we can use these investment funds to help the company supplement the sales force and sales team, you can see this good product service More customers.
Lily:
General Yang, I know you started as a software engineer in the late 1990s. What were the more common career goals for software technology developers at the time?
Yang Chen:
I estimate that most entrepreneurs, especially C-end entrepreneurs,Are much younger than me. In the late 1990s, as an engineer, the best choice should be to go to a foreign company, but there are also some who might choose to be pre-sales consultants, after-sales technical consultants, or supporting technical consultants. But the basic path is from the technical side, and then gradually contact the project side, the so-called project side is that the overall interaction with the customer will be more and more, from the project side will slowly move from after-sales to pre-sales, and then some do business . So generally speaking, most software practitioners or technical engineers who enter the industry will experience such a career path, which is very different from the engineer in 2000. It is estimated that most engineers at the end of 00 will go to Internet companies.
Lily:
For the engineers who graduated this year, what do you think is better?
Yang Chen:
If you are a young person graduating this year, I think joining a startup company is the best choice. In terms of trends, first, China today is an era of mass entrepreneurship and innovation, but in fact it is because China, a huge domestic demand market, has brought a great deal to technology startups, especially enterprise service companies. Huge survivable soil.
For young people, one of the biggest benefits of startups is that they can get very good exercise. Because in big companies, it is always just to make screws, and entrepreneurial companies give you a stage. This stage will continue to grow, so your development space on it will also become very large. Now some seemingly unknown science and technology enterprises, especially those on the enterprise service track, may have huge development in 10 years, so I think young people should be brave to try some different life paths.
Kwong Ying-wah:
Because I belong to a halfway career, I think it ’s ideal for everyone to find something that interests them and that they can maximize their value. Some people may be suitable for large companies, relatively stable, and learn from the basics, but I think that for those who are ambitious, have strong learning ability, and are willing to see their positive impact on society as soon as possible, in fact like Qingchuang A growing company is a very good platform, because the company is growing, it will give these young people more opportunities to contact with new technologies, new industry trends, etc. It is a very good platform to expand their horizons and learn to grow.
Lily:
This new crown epidemic outbreak has a global impact. Will it affect Qingchuang?
Yang Chen:
From our business ontology, the impact will not be very large. Because in essence, we mainly serve medium and large-scale enterprise customers, which are less affected by this particularly large seasonal fluctuation, so for us, it is more affected by the time window, such as the epidemic During this period, some projects may be delayed because there is no way to meet and communicate with customers, but the demand will not disappear.
In terms of organizational management, the biggest impact is that we are forcing ourselves to become more effective. In the entire remote office process, we made OKR more intensive. It turned out to be a two-week period, but now it is a single week. Part of the team, especially the production and research team, has become a daily review, so the efficiency of the entire team, especially R & D, has improved.
There is another thing that has changed us a lot. I quite believe that a sentence called Slow down, Move fast, this is a proverb that I used to tell us when I was in BMC. Any time you need to make a decision to move back, but often you need to slow down to think long-term. Startups are sometimes like a fast-running, nonstop machine. This epidemic has enabled our entire team, especially the management team, to slow down, think more about some strategic issues, and see some of their own shortcomings. , So these are my thoughts during the epidemic.
Lily:
How did Sinotron Technology cooperate with GGV?
Kwong Ying-wah:
Our earliest contact with Innolux was at the end of 2018. In fact, it was a long time away from our real investment. As a global platform, GGV has invested in many corporate service companies in the United States, so when we look at corporate service companies, we have a complete set of methodologies and evaluation criteria. Therefore, at the end of 2018 and early 2019, when we analyzed the data at that time, we always felt that we still wanted to wait a bit. Later, in the second half of 2019, we realized that the company’s growth was very good, and a large number of 2018 customers chose in 2019 Annual renewals, and even additional purchases, the company’s products are also improving, so we think this is the right time node, we must invest in helping the company grow better.
Yang Chen:
What impressed me the most about GGV is your analysis of the entire track. I think GGV is a typical plan and then moves. By meeting with Ying at the end of 2018, to 2019In the second half of the year, we met again, including later when I met with Ji Xun. It felt completely different. I could feel that they really did know a lot about this market. As Yingying said, basically all the innovative companies related to IT operation and maintenance management have been seen. The feeling is that there is already a lot of knowledge reserves. We very much hope to find a like-minded partner and have a deeper understanding of this track .
Secondly, when I saw Ji Xun, I was very impressed. Generally speaking, preaching always talks about higher operating income, but Ji Xun interrupts me later, and he did an analysis for me. He said that your growth rate is actually too high, because this time The issue you need to consider the most should not be the growth rate issue. He gave me some insights, and I was very touched. I think he is very sincere and knows the industry very well. He knows what problems entrepreneurs should encounter at what stage, what issues to consider, including some ideas about our future in the sea, and also gave me a lot of input .
All of this makes me feel that GGV does not just give you a question of money, but that you will have one more partner on this road, even if we do n’t need money at that time, but because of the team ’s Analytical skills, recognition of this vision, and some thoughts on future development. In the end, we still chose to cooperate. Today, I also feel that it is a very correct decision. We have found a very good partner.
Lily:
Qichuang has been established for 4 years. Do you have some experience and lessons to share with young entrepreneurs?
Yang Chen:
There are a lot of lessons, because we are constantly jumping into the pit, and we just feel that there are new pits when we just climb out of the pit. I actually like Zeng Guofan’s saying, knot hard village and fight in a daze.
For entrepreneurs, especially young entrepreneurs, this is a very useful sentence. Strong hardship means that your products must be solid, do n’t do projects one by one for income, this Not a long-term plan. Fighting in a daze is that you have to live. At any time, no matter what competitors you encounter, and what difficulties you encounter, first of all, you must persist in survival and go down step by step.
Of course, maybe this experience is only suitable for enterprise service, because enterprise service is a rational decision-making market behavior, so if you really put the customer ’s success first, if your product can solve his problems, time Is your friend, after a while you will find that the so-called compound interest growth will eventually let you goOn a relatively large stage. This is the experience I think can be shared with entrepreneurs, especially To B entrepreneurs.