Shulou(Shulou.com)06/02 Report--

Big data, what is big data? How big is the data called big data? The popular data analysis came to us one after another, saying that enterprises would not be able to analyze data for a long time, but what kind of data is big data and what kind of data is the largest?

If you haven't come into contact with big data, then you don't know how big big data is and what kind of data can be called big data. Then, according to the port of data collection, the quantity level of big data is different between the enterprise side and the individual side.

Big data's development and learning is difficult to a certain extent. To get started with zero basics, you need to learn the Java language to lay a foundation. Generally speaking, it takes about 3 months for Java to learn SE and EE. Then he enters big data's technical system, mainly studying Hadoop, Spark, Storm and so on.

What is the age of big data before big data?

What is big data?

How big is the data called big data?

For many people who have not come into contact with big data, it is difficult to know clearly how much data can be called big data. Then, according to the port of data collection, the quantity level of big data is different between the enterprise side and the individual side.

Here I still want to recommend the big data Learning Exchange Group I built myself: 251956502, all of them are developed by big data. If you are studying big data, the editor welcomes you to join us. Everyone is a software development party. Irregularly share practical information (only related to big data software development), including the latest big data advanced materials and advanced development tutorials sorted out by myself. Welcome to join us if you want to go deep into big data.

The enterprise side (B side) data level of nearly 100, 000 can be called big data; the personal side (C side) big data should reach the level of 10 million. Collection channels have no specific requirements, PC, mobile or traditional channels can, the focus is to achieve such an order of magnitude of effective data to form a data service. Very interesting, you can see 2B and 2C, the difference between the two types of big data by two orders of magnitude.

Some small companies have data on a scale of only ten thousand to ten thousand, but after collection and analysis, they can also sum up the principles of this group, and can also guide enterprises to carry out a certain degree of user analysis, acquisition, or service work. But this is not big data, but general data mining.

In the case of Tian × × sharing just now, it was said that at the beginning of this year, there was a middle-aged man who sold fruit with a principal of 50 yuan. He did not know big data, but he knew the fruit harvest like the palm of his hand: he knew where it rained, how sweet the fruit would be, and where consumers would like to eat the fruit with this sweetness. In the end, it sold 137stores with annual sales of 470 million.

This is indeed a small kind of data mining, but it is not data analysis. Although big data analysis was born here, big data is oriented to a larger amount of data, with the help of a broader analysis method of knowledge database. The data sources of most data companies are massive, and its collection and analysis are not limited to individuals, but to a very wide range of groups.

What is big data's industrial chain like?

When I was interviewed, according to the relationship between the upstream and downstream of big data's industrial chain, I proposed to divide them into three different categories:

Big data collection co., Ltd.

The so-called "finding data" can be divided into two types internally:

A large number of data sources can be generated in the process of normal operation.

Through cooperation with telecom operators and financial enterprises, access to data sources.

Big data analysis co., Ltd.

This kind of company basically has its own set of models, but most of the database models come from the same several mechanisms, including statistical models, deep learning algorithms and so on. It is also based on the applied analysis module developed by IBM and cloudera and so on.

Big data sales co., Ltd.

Although it is said to sell data, but the sale is not a single data, but a complete set of data-based solutions, such as precision marketing and so on.

How do these three types of companies work together and apply big data to our lives? The easiest thing to understand is the ads placed on Wechat moments now.

When Tencent promotes advertising to every user, it has made an accurate analysis of users. By collecting people's use habits on Wechat, and then analyzing users' spending power and consumption habits, form a set of accurate marketing plan, and generate some targeted advertisements for advertisers.

Lancome ads, for example, are never promoted to male users or luxury car ads to fresh graduates. Big data's analysis model is used in the entire Wechat advertising system, and there is general feedback that the conversion rate of ads placed on Tencent is higher than that on platforms such as NetEase and Sina, thanks to Tencent's big data foundation.

The Investment value of big data Company

How to understand the investment value of big data?

Big data is so popular now, its commercial value is obvious, but there are not many people who can really realize it.

To realize the commercial value of big data, the first requirement is to reach the data level of big data. So at present, the three BAT companies have the most advantage in terms of data volume. In the PC era, Baidu has a very strong advantage in data, but in the mobile era, Tencent and Ali have achieved anti-surpassing.

Tencent has Wechat and QQ, which accounts for 90% of the data generated on mobile devices. Alibaba uses its consumer data resources to be more vertical. Then for small and medium-sized enterprises and start-ups, the focus of realizing business value is how to make use of other people's big data resources to better serve their own businesses when they are small. This requires deep judgment and excavation.

Therefore, when judging the investment of a data-related company, it is not only the development of the existing business, but also whether it can accumulate effective data and accumulate high-accuracy data in the process of its continuous development. to achieve real-time updating of data. Only in this way can enterprises better set up competition barriers.

For example, in the field of developer services, such as talkingdata Aurora and so on, Fosun Kunzhong attaches great importance to the project, that is, the current business of the project is to provide services only for developers? Or in the service, to their own accumulation of effective data, the formation of long-term barriers?

2B is the breakthrough of big data industry.

I mentioned earlier that BAT is a monopoly on big data collection, and it is very difficult for startups to achieve huge amounts of data (tens of millions or even hundreds of millions of C-end users) on the C side. Currently, there are only 15 app with more than 100 million monthly active users in China, and the top 10 app are controlled by BAT, such as Wechat, QQ, Taobao, UC browser and so on. If BAT is bypassed, only the relatively traditional telecommunications industry, financial industry and so on can have massive data on the C side.

It can be seen that if you want to invest in companies in the field of big data, it is very difficult to start from the C side. So, I think if you want to layout in the big data industry, 2B is the key: on the one hand, 2B development is relatively late, BAT has not formed a monopoly; second, the threshold for development is relatively high; third, the demand for data is relatively small, reaching the level of 100, 000 can serve big data's analysis, so if you want to invest in the field of big data, the main area you should pay attention to is 2B.

In the 2B area, there are three different categories:

The first category is the B2B trading platform; the current trend is basically the e-commerce trading platform in the vertical field of the industry, and the core competitiveness is to break the information asymmetry and opacity between buyers and sellers. So the key point for companies in this field is not to record trading volume, but every valid data information. In this field, we have invested in Huimin, mainly serving the trading platforms of small and medium-sized merchants and their suppliers, such as various "find" series projects, and so on.

The second category is now very popular enterprise services, mainly to SaaS, such as customer management CRM, human resources plate HRM and so on. Under the premise of getting user permission and ensuring data security, they accumulate data of enterprise users and employees by serving a large number of enterprises. For example, Licai Network and so on.

The third category is services for developers, such as cloud storage, statistical push of running data and instant messaging in app.

Fosun Kunzhong mainly invests in these three types of 2B projects, because the 2B business model of these projects can effectively accumulate big data. This is why Fosun focuses on both big data and 2B enterprise services-because in 2B enterprise services, you can find the best and most effective big data.

The future investment target of 2B industry

If we predict the future of this industry, I have the following views.

Enterprises with rich data sources will become the hottest investment targets in the whole industry.

In big data industry, the difference in analysis algorithm, resulting in the difference in accuracy and practicability of analysis results is the difference between 93 points and 95 points. The difference caused by the quality of the data source is 60 points and 90 points. In particular, a constantly updated big data is an effective way to verify the accuracy of this algorithm and constantly optimize big data's analysis results.

The project that binds the demand side that is most in need of data will win.

At present, the people who are most willing to pay the bill in big data's field are basically customers in the financial field, banks, insurance companies, and so on. They want to carry out multi-faceted analysis and services to users, so they are very willing to buy. The next layer is the emerging Internet companies, in order to more accurately obtain users, improve the conversion rate, but also more willing to pay, such as Meituan-Dianping and so on. Next, there may be a transition to the consumer goods industry.

Based on big data's business opportunities

The harvest of intelligent hardware and artificial intelligence is still a long time.

In fact, the mode of combining big data with intelligent hardware is still very challenging, and the main reason is the order of magnitude of big data. At present, the shipments of smart hardware can not match the order of magnitude required by big data. At present, the largest domestic shipments are millet bracelets, followed by 360 children's guards. The remaining shipments of smart hardware are often in the order of hundreds of thousands to hundreds of thousands of thousands. This is still a hundred times different from the 10 million and 100 million level requirements of big data at the C end.

The field of artificial intelligence is relatively good, such as the rise of Japan. Google, Amazon and Softbank Corp. overseas are already investing in some investment targets, but they are still projects in some conceptual areas, not services that can be commercialized immediately. Including Google's unmanned cars, even though they have accumulated millions of miles of safe driving mileage, there is still a process to apply. There are also a small number of angels and early institutions in China who are optimistic that they will start to invest in this field, but it may have to wait at least five years for it to blossom and bear fruit.

Therefore, the investment in this field should be patient, and there are more promising areas, including voice semantic recognition, AR/VR, drones and so on.

Why are SaaS projects so popular?

In fact, many people do not understand what is the difference between SaaS model and traditional software services, why is it an industry based on big data?

There are many differences between SaaS and traditional software services. The most fundamental difference is that their entire architecture is different: SaaS is based on the public cloud, standardized module services, and data is also stored on SaaS's public cloud platform. The traditional software services are basically deployed in the local area network. This architectural difference determines all other differences.

For example, because the SaaS architecture is in the cloud and adheres to the principles of standardization and universality, the implementation process is very fast. At the very least, the site construction work before implementation is much less, so the acquisition of users is also accelerated accordingly. It takes a long time to accumulate hundreds of customers in traditional mode, while it is not difficult for SaaS mode to accumulate thousands or tens of thousands of customers in a short period of time.

For example, the payment model is different, the traditional software has pre-implementation fees, annual update fees, special customized service fees, troubleshooting costs and so on. Generally speaking, the cost is high and the payment is complicated, and only large enterprises can afford to spend. On the one hand, SaaS reduces the initial deployment cost, and the system and architecture can serve multiple users. Its charging model is basically a monthly fee or an annual fee, which is only a few hundred yuan a month, and many small and medium-sized enterprises can enjoy services.

Quan A link

Q: is it feasible for Xiao San to participate in the SaaS investment of big data and AI?

A: I think the only way Xiao San wants to invest in this area at present may be through equity crowdfunding. The industry threshold of these projects determines that it is best for Xiao San to invest in such projects through a professional crowdfunding platform.

Specifically, there are several reasons:

The threshold of the project is high. Our small scattered contact projects are often through our own circle of friends, but the entrepreneurs of such projects are basically professionals, and our small scattered are out of reach.

This kind of project requires high professional experience of the founders, and it is difficult for Xiao San to carry out such professional projects. On the other hand, the professional crowdfunding platform has done a background check on the project before it is promoted to the small and medium-sized investors. To achieve a protection for small bulk investment.

All in all, the threshold of this kind of project industry is high, and the requirement of professionalism is high. Xiao Sanxiang must find a responsible crowdfunding platform like Angel Guest.

Q: everyone says that we are in the Internet age, and you emphasize that we are in the data and information age. How do you understand this?

A: we are already in the big data information age. Big data does not conflict with the Internet, and it is precisely the emergence of the Internet, especially the mobile Internet, that has greatly improved big data, who can be collected effectively. So the era of big data came arm in arm with the mobile Internet.

What is big data, exactly how much data can be called big data, do you know, if you want to learn big data technology, then work hard, on the future road, know how to analyze data, you can grasp the future!

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.