So for small and medium-sized enterprises, in fact, they are also very concerned about a big pattern change, which can bring them an opportunity to expand the market scale and even lead the industry. Therefore, this newspaper cooperates with CBN Brainstorming Program to pay attention to how innovative SMEs grasp the opportunities brought by the era of big data, and what bottlenecks and challenges they face. The guests who participated in the discussion were: Zhang Shaofeng, CEO and Vice President of Beijing Dianfen Information Technology Co., Ltd.; Liuli, CEO of Bawangcai, small secretary for ordering food; Zhang Yiming, CEO of Beijing ByteDance Technology Co., Ltd.; Ding Zuyu, CEO of Yiju (China) Holdings Co., Ltd.; Tang Jun, Chairman and CEO of Hong Kong and Macao Information; Yan Mingchuang Investment CEO Kuang Ziping; Yi Lee, co-founder and CEO of Venture Commune; Kong Fanren, Chairman of Zheng Qi Gumu (China) Consulting Company. 1 Big data is changing from a technical hot word to a social wave, affecting social life, but what can big data change? "Big data is not necessarily used to connect people and information. It may be used in medical care and finance. Usually, it is to better obtain the information of individual users. The first feature is that you can use new content at any time, and the other is more important. You can use everyone's interest to analyze everyone's behavior and recommend content that all users are interested in. The process of reading by users is for themselves, but the result is actually that I am for everyone. " Zhang Yiming: For me, big data has changed my ability to connect users and information. I can use big data to change the way users get information. As far as the company is concerned, our company can use big data to connect users' information. We can analyze and process massive information, analyze massive user behavior, and connect information with users. There was no such opportunity before, and only after more and more user behavior data can be analyzed can we have this ability. At the same time, it can also give users a better discovery. Big data is not necessarily used to connect people and information. It may be used in medical care and finance. Here, it is more typical to obtain the information of a single user. The first feature is that you can use new content at any time, which means that we have obtained a huge amount of data. This is called big data, but another more important thing is that we use everyone's interest to analyze everyone's behavior and recommend content that all users are interested in. On the client side, a large number of users' data behaviors are mined. The user's reading process is for himself, but the result is that I helped everyone, that is, better access to information. Zhang Shaofeng: Big data has changed my life and work. I can use big data to change the operational efficiency of enterprises and people's lives. What is big data is to integrate data from different industries and fields, and the data can undergo qualitative changes. Because there are more data, an industry has an extra latitude. With the increase of latitude, the data value changes exponentially. I saw the impact of big data. I can provide services for other enterprises based on my big data, such as precision marketing. Let me give you an example. Yintai Department Store is called OTO. Online and online integration will get it through. What does it want to do because of its competition with e-commerce? Users want to go to the mall for free and lay a free wireless point. He hung up a few percentage points as soon as he went online. Do you know what this user's preference is? Yes, you pushed him a coupon and said that there was a sexy underwear on sale on the second floor. Do you want a top brand? Maybe the fifth floor has a taste you like. You like fresh fruits and vegetables, which is a very strange dish, because we have to analyze users' preferences on the whole Internet. Chai Ke: Big data brings new data services and products. I think that although we are an entrepreneurial enterprise with a small size, more than one million girls actively participate in informing the relevant physiological data every day, or they are all Asian people, so I can find a rule through the disorder of the data, which is more valuable for the research and development of products and drugs for Asian people. There are nearly 2000 Asian women in our company now. I believe our data is still convincing. Maybe in the end, I also think that the era of big data is not an era. Data services and data products are the times that our entrepreneurs have begun to explore and have taken shape. 2 For entrepreneurs, is big data an entrepreneurial change or a flicker? "Big data itself is not a fool. But any revolution, when it comes, people's expectations will exceed its actual speed. You think it will get faster, but it will actually slow down, but it has really changed and will be destroyed faster than expected. Now, in fact, big data has not yet arrived. In fact, big companies are still fooling people. " Zhang Shaofeng: I think many companies are bluffing at this stage, but big data itself is not bluffing. Companies such as IBM and Oracle Bone Inscriptions are bluffing. They introduced the concept of N-size data to our customers in order to sell their equipment. For example, a recent customer of mine was told a lot about cloud computing by an internationally renowned brand company, and my customer was fooled and bought a lot of its equipment. As a result, the chief operating officer asked me what I could do after I bought it. Isn't this a bluff? In fact, you have to figure out what application you are making, and then buy those devices and software in turn. Because the application aims at value, you must first think clearly what value I want to create, and conversely, what equipment and software I need to create value, so I think IBM, oracle and Microsoft are all fools, and SAP is all fools. They are not big data companies. Let me give you another example. In the past, the boss in Oracle Bone Inscriptions said that everything in the cloud was hype. The cloud is nothing more than the previous CS mode and BS mode. Later, Google couldn't help it, and Oracle Bone Inscriptions made one. Because we don't need high-end packaging, we can't sell equipment. I think that at the beginning of any revolution, people's expectations actually exceeded its actual speed. You want to change quickly, but it actually becomes slow, but when it really changes, it will be destroyed faster than you expected, just like e-commerce, so I think big data has not arrived yet. In fact, big companies are still fooling people. In fact, they are not big data companies, they are product companies, and they want to make a product. We don't care if we are big data. Just solve the problems of our enterprise. Big data is really a flicker. What I want is how to solve my problem, and then one day I say OK, and I package it into big data, but I still believe that if the matrix described in The Matrix becomes a reality, then the future must be the era of big data. This is an irreversible law that I firmly believe in. Zhang Yiming: I agree with Shao Feng very much. Solving the problem will naturally bring some improvements to the scheme, instead of taking existing machines and existing software. When we do something to solve the needs and problems, we will naturally think of using lower costs. We used to do big data. Our typical weather forecast is to use supercomputers and 500 CPUs, which are no longer used. IBM has been promoting supercomputers for a long time. Who will buy this kind of CPU computer now? Tang Jun: I don't think this can be said. They are called equipment suppliers. Why is there data in the past with the advent of the era of big data? Why is it called big data now? In the past, it was not called big data, because now the computing power is enhanced, so a large number of data can find order from disorder through calculation, which must be based on the increase of computing power. So these vendors have actually brought changes to big data, which is undeniable. Now we really can't do it without them in our application. For example, when we were doing gold futures, you said that you found the correlation between the data of the past ten years and found a trend. What is needed is that the calculation speed is very fast, and then it helps you to react quickly and make a judgment. Yi Lee: This is actually the underlying support of big data. In today's world, big data standards and the right to speak are in the hands of IBM, SAP and EMC. Because it is very simple, why do you think domestic banks, such as China People's Bank, should purchase SAP software and IBM mainframes? That's true. Now you have to call up one year's data at any time in one second. How can this data not crash? How to deal with unstructured data? Frankly speaking. Including the artificial intelligence that Microsoft and Google are pushing now, I say that Arabic is real-time, so you can translate it for you without any interval. Frankly speaking, this is not what ordinary people can do. Let me repeat, flicker is not flicker, and bubbles are not bubbles. Now there is a saying that the investment industry is very popular. What is a bubble? You're inside, that's a bull market, you're not inside, that's a bubble. What's in it now? How can it be foam? You have to say it's okay here. This is obvious. Then the second one says that from the three latitudes of big data now, it actually means that data must be collected and stored after collection. This is something that small and medium-sized enterprises can play. This is what big enterprises and related enterprises of Chinese Academy of Sciences are doing. The third is analysis. If I can predict the sudden cold after half a year, I can even predict the glacier after ten years. But to put it bluntly, where does your data come from, and then do you have the ability to analyze? Then, in a sense, SMEs have found their place in their respective industrial chains. Liuli: Although I think this big data is very exciting and has indeed brought many changes, I think there is indeed a bubble in it. Why are you excited at the meeting? I think big data gives people an illusion, from predicting people's behavior to predicting the future. I think that although big data plays a great role, there is no need to exaggerate it. We need to always start from the demand, instead of saying that I am doing this big data. The purpose is to be a big data company, our big data. As our main business develops, it will naturally become big data. This is my opinion. Is it easier for small companies to play big data, or are there more opportunities for big companies? "SMEs can only obtain data from large enterprises and other places, and then help them interpret and analyze. Many small and medium-sized enterprises will explore that if there is a mechanism to enjoy it, these data will still exist in small companies, which may be more useful than the data of large companies. It can be seen that small companies must think of applications in order to collect data. " Yi Lee: Big data comes from big people. Look at the Shanghai Elevated Road, all kinds of elevated roads in Shanghai. Why are there signs on it? It shows red, green and yellow. How did it come from? Because induction coils and weapons are buried under elevated roads, they can be collected. Does this mean that SMEs can play? This is what big enterprises and related enterprises of Chinese Academy of Sciences are doing. Small and medium-sized enterprises can only get data from big guys and then help them interpret and analyze them. I estimate that many people should be "very accurate" in using mobile apps. This is made by a small company in Hefei. Why can you provide data for flight delays and so on? That is to give it up and talk to the Civil Aviation Administration of China. The Civil Aviation Administration said that this data is idle anyway. Then let's talk about a share or a business cooperation. Several people in the company will be on fire in Hefei because the data he has in the computer room of the Civil Aviation Administration of China is not very accurate. This conclusion is that in the era of big data, it is as simple as that for entrepreneurs to stop playing with big guys. Zhang Yiming: Yes, I think it's good for innovative companies to get data from big companies or government agencies, and then as a society, the standard exchange of big data must be developed or developed, and it's better for the government to have this data disclosure plan, but I think it's best for startup companies to stand up by themselves, that is, the application itself is collecting, I'll recommend information to you, and you can tell me your information yourself, so that it can be virtuous, and it may not be as big as Chinese data at first. Chinese data, but users have been used. The more you use it, the bigger your data will be, so I think the application and collection of forward circulation is very important, which is also an opportunity for startups. Liuli: Four small possible data are not big enough, 40,400,4000. As long as there is a mechanism to share these data, I believe many small and medium-sized enterprises will make an exploration and have a sharing mechanism. Data will still exist in small companies and may be more useful than data from large companies. Zhang Shaofeng: In other words, if a small company wants to collect data, it must consider the application. The application itself, whether 2B or 2C, is valuable. It depends on services or applications to get data, not the data you give me. I have joined 800 companies, maybe one is not that big, and together it must be bigger. Chai Ke: The products and services of small companies can obtain data more effectively. Just like a message released by a health department in a country today, all girls must tell me every day whether you have had your period, and no girl will tell him, because you have to rely on excellent services to get close to girls' lives in order to gain trust and data application. A large number of entrepreneurs are standing at this point in time. How to find "big data" for gold? "In the era of big data, entrepreneurs who want to start a business in this field must first start from the needs of users, find out the pain points from the needs of users, and solve these problems with the hands of big data instead of deceiving themselves." Yi Lee: As far as the service of big data is concerned, if we call it a revolution, it is only a small revolution to the original big enterprises of IBM, ICP and EMC. For it, it is a slap in the face. But for small and medium-sized enterprises, it is a revolution, a revolution of small enterprises, so that it is possible to really make a professional big data. The safest way for innovative SMEs is to get rich quickly and sell quickly. After accumulating strength, they can do more things, which is more reliable. The second is that if we talk a little longer, we have made it clear from data collection, storage and analysis. In fact, the most opportunity is analysis. For small and medium-sized enterprises, if there is an opportunity, we must focus on analysis. For example, catering workers should do this well and don't want to do anything else. If they have cooperated with Yintai, they have to do it. You don't want to do so many things. You just have to stay focused. You can only concentrate. There is no other way. Tang Jun: I agree that there are basically no collection opportunities, because the collection is limited after all, right? So we can only rely on historical data, but analyzing this is definitely a great opportunity. Because what do you think this is? Compared with wisdom. I think if we talk about the concept of big data, there are still many opportunities to use the data collected by predecessors to analyze and find your application on the basis of this analysis. This is my thinking about all entrepreneurs or those who want to start a business in the field of big data in the future, that is, to provide a suggestion that they have sufficient opportunities to find their own entrepreneurial opportunities. Kuang Ziping: Big data is to find a rule in disorder. This is big data, something you can't predict. But through the analysis of your big data, you found something predictable. This is my understanding of big data. But I don't agree with the analysis I just chose. I agree that the simplest thing is to do analysis, but the problem is that it is difficult to do analysis in China, because the enterprises with the largest amount of data in China are not as open as those in the United States. Some large enterprises have a lot of data at hand, but they are generally not open to third parties in China, so if you really want to put pen to paper on the analysis side, you must first make a unique contribution on the collection side. Of course, you have to do a lot to stop you from collecting, so you have no choice but to stop doing that. For example, if someone is willing to give you his cycle, which is open, or if someone is willing to give you something to participate in, you can do so. Therefore, my suggestion is that entrepreneurs who want to start a business in this field in the era of big data must first start from the needs of users, find out the pain points from the needs of users, and solve these problems with the hands of big data, instead of deceiving themselves. Don't start with big data. Just because I have big data now, I want to make a big data, and then I will find end users who can do it with me. First of all, you should know the industry you want to enter, the needs of users and the problems that many users want to solve in the past. Now that I finally have the means of big data, I will solve it, so in this case, we will not follow suit, and everyone will not do advertising push and accurate advertising push. Not 100% of users want you to push advertisements accurately, so I think it is necessary to know your industry very well and what their needs were in the past, but there is no way to satisfy them and use new ones. Ding Zuyu: Now I think big data is an idea. It can push you to make some innovations, but if you want to make money and sell it, you must rely on bluffing. In other words, you will only spend money to buy it if you convince me by bluffing. In fact, in the end, the concept of big data was over-packaged in their applications, and then you can get simpler data analysis than before, such as selling 10 yuan, and the concept of big data fooled me and made me dizzy. I gave him 20 yuan or 100 yuan. I think a suggestion for small and medium-sized enterprises is to think about the person who pays the bill at last, the person who pays the bill for you. Of course, I suggest thinking more about real estate, because many people pay for real estate. If you can provide a good real estate solution, you can live, live for a lifetime and be comfortable. Then look at other industries, cars, etc., one by one, and then you can dock with him who can dock with you. Kong Fanren: He is an idea and a way of thinking. For some companies like Tang Jun, it may be like a small pit, and one foot will pass. For some enterprises, it may really be a trap, so what we want now is the value of data, but we really don't care much about whether it is big data. Whether big data is an opportunity or a flicker, I think we can learn from big data and think with it, but don't be obsessed with big data. 5 What are the bottlenecks of innovative enterprises in the era of big data? Is it a talent problem or a user privacy problem, or a data acquisition channel problem? "For startup companies, big data should have a small application. This small application should be a successful application, not a big application. Using big data to make big applications is not something that entrepreneurial companies do. How to directly link this application to the final profit model. " Tang Jun: Big data, big data, without enough data, it is actually impossible to make big data, and it is difficult to make a judgment or decision just now, so everything is based on big data, but for our general entrepreneurial companies, it is difficult for them to collect enough data, just like Google. So the problem of data acquisition channel is a big bottleneck. Ding Zuyu: I think it's a data application. I think there are some problems in the first three aspects, but in fact, we have all thought of ways, such as the problem of talents, access by ourselves, and user privacy. There is no privacy now, and then the database, anyway, has its own channels, but it is not today's big data. We can use the hat of big data to cover it, but I think big data should have a small application for startup companies. This small application is what we think is a successful application, not a big application. Making a big application with big data is not something that entrepreneurial companies do, but this application can be directly linked to business or the final profit model today. I think this is the biggest problem for all start-up companies. Zhang Shaofeng: I chose talents. Why? It is precisely because I have been doing data mining for ten years that I knew when I worked at IBM that a data warehouse is particularly prone to failure, that is, technicians talk about technology and business people talk about business. In the final analysis, we need a talent with both technical thinking and business thinking. Such people are hard to find. You can tell your customers a lot. I can help you with the data. What are you doing with all this data? I'm not sure. I think this is the bottleneck that restricts the development of startups. This is a man with comprehensive thinking, which is not easy to master. This is one of my views. Chai Ke: I also think it's a matter of talent. Now there are also professional gynecologists in our team, but they are far from enough, especially those gynecologists. We adjusted their annual salary from 300 thousand to 600 thousand, and no one wanted to come. First, because they don't trust small companies, you are in an entrepreneurial organization and he is very stable in the hospital. Second, because these doctors with medical experience are older. There is also his disdain for internet things, so I think it is a challenge for people in our company at least. Zhang Yiming: I think the bottleneck is the product, because I think talents, privacy bottom line and data channels are really problems, but the problem is not necessarily the bottleneck, because I just said that with products and good products, there will be a lot of user data, and there will be channels for data acquisition, good products and good data. Only by building a good platform can talents come, not by recruiting talents from top institutions. For example, Google is acquiring a company called WAZE (Crowdsourcing Map). He is a small start-up company. He collected a lot of data through his own application. Both Google and facebook are grabbing data from this startup. This is a good example. You should collect data in a unique way and provide applications. This is the product.