Big data in China

Chapter 17 Big Data and Technological Change (1)
Farewell to the era of small data

We rely on the intuition generated by an independent data point to analyze and judge problems, which is a typical technical approach in the "small data era".However, this kind of intuition and data judgment based on a certain point can only solve daily problems. When faced with complex information flow, it often leads us astray in classification and decision-making, resulting in a series of mistakes --Can a certain information "point" make fuzzy predictions on general laws from point to surface?Is the case representative and extensive?What happened today, will it happen tomorrow?

If there is a lack of macro statistics on continuous data and multi-field data, as well as the development and use of related technologies, people may be dazzled by the latest and most recent "data points" and lose the overall grasp of the overall situation.

Big data technology is just the opposite of small data technology. It is more of a macroscopic technical thinking. It is the driving force for us to jump out of the "plate" and find answers with a broader perspective. It is to help us learn from various types of problems. The ability to synthesize and rapidly obtain valuable information from data.

Just like the operating system.If small data is Android (only for mobile phones), big data is XP.It carries more, faster speed, more accurate analysis, more capacity, and can trigger a technological change.

In terms of technical preparation, big data is also more extensive than small data alone, and almost exhausts all current Internet technologies, including massively parallel processing (MPP) databases, data mining grids, distributed file systems, and distributed databases , cloud computing platform, Internet and scalable storage systems, etc.

Take our life as an example, if you use big data technology to manage your daily life, you can't just rely on simple unrelated Excel documents for statistics, but you should make them interact, build a comprehensive database, classify Only by analyzing and summarizing can we improve our "life management".

Consumption data:

The statistics of our consumption is a very important thing in our life. I believe that almost everyone has a family account.But how to count and analyze consumption data, what tools and methods to use, the final results are quite different, and there may even be diametrically opposite, ice-fire comparisons.

The statistical method of small data is nothing more than a list. How much was spent in January and how much was spent in February? What should I do if the necessary matters exceed my plan?Can only accept the reality.But with this method, it is often the same in the second year, and the problem cannot be solved.

If we change our thinking, we can introduce statistical techniques of big data to ourselves.For example, according to the consumption plan, list different tables separately, re-book, classify and analyze: what money should be spent?Which ones should not be spent?What kind of money consumption is impulsive?Which ones belong to us being fooled by merchants?

The important thing is not statistics, but to find the reasons.We innovate thinking and repurpose technology.There is no epoch-making breakthrough in technology, but the functions used have changed.Therefore, we can find out the incentives for spending money indiscriminately, and then make changes according to the specific situation. We must make weekly statistics and analysis, and implement a consumption budget system to achieve the goal of rational consumption.

Time data:

Big data also has its efficient application in time management.How to allocate your limited time and energy reasonably?How do you make the most efficient use of your time?Small data thinking is to determine the work plan and then do it according to the plan. Strictly implementing the plan is the time management principle in the small data era.But the era of big data is different. You first need to determine your goals, then sort the things you have to do according to their importance, and then list the plans and execution plans.In the process of time supervision, you also need to adjust the urgency of different matters at any time, and change the plan flexibly to make the most efficient use of your time.

Work data:

In the era of small data, work data is a running account, and all you see are boring records.Big data turns it into a treasure trove of working data.Moreover, this is not only a process of recording work through data, but also our continuous summary and understanding of our own growth process.

Big data will not provide you with the final answer, everything it records is just for your reference - you can easily analyze the past, summarize the present, and gain experience in order to obtain better working methods and directions.

A friend of mine has been studying various applications of data technology for more than ten years. His biggest feeling now is: "Although the technological change has begun, many people have not realized that they should forget the importance of small data as soon as possible." Only by investing in the technical world of big data can we seize the future by investing in technical thinking. This shows that the reorganization of technology is actually building a higher platform, which requires the advancement of our thinking, and then we can enjoy the new functions of data technology with peace of mind. For example When we conducted street surveys ten years ago, we only needed a sample survey, which was small data; but now, we need to conduct a survey in hundreds of cities in dozens of countries at the same time, and then aggregate the data for analysis. Real-time analysis, and quick conclusions, that’s big data. Obviously, I’m not sure practitioners are jumping on the bandwagon.”

Now, what are the impetus for us to build "big data technology"?The first motivation is that we must be clear: "Do we really need a lot of data?" This is a question of "Why does big data exist?"If you can't solve this problem first, you will blindly change technology in order to have big data, and pay an ineffective price.Some people are full of expectations for big data, hoping to discover things that they did not realize in the past and reap surprising results, but in the end they find that "we don't need these things at all" or "this is just an existing fact".

Just like some companies have invested tens of millions of yuan in system development, vowed to step into the threshold of big data, bid farewell to the era of small data, and finally come to a conclusion that proves the "experience" of senior employees. It is unacceptable.Only "demand" is the biggest driving force, which is the main driving force for technological progress.If you have no demand, then small data technology is also very good.Just like if you only use your mobile phone to answer calls and send text messages, and never surf the Internet and make video calls, why spend a lot of money on Apple mobile phones?

The second motivation is related to the need to maintain data: who will maintain the large amount of data to ensure the quality of the data?
In other words, can our technology (technical personnel) guarantee the collection and collation of high-quality data?

For example, the department head of a company will receive promotional materials from a certain client every month, but the title of the recipient is not "Department Director", but the title he had when he was the marketing manager of the company some time ago.Although this is not a big problem, and he will still receive these materials on time, he still asks the other party to change his title.

The account manager apologized on the spot and said that he would make changes immediately when he went back.But the next month, when the executive received the materials again, he found that the recipient's title had not changed, and there were still a few big characters written: Marketing Manager.He was very disappointed and decided to stop working with this client.

where is the problem?Is it just that the other party didn't pay attention to this detail?Of course not, in the final analysis, the client company lacks the awareness of maintaining the customer database. When collecting and organizing data, the work is full of negligence and cannot maintain the real-time nature of high-quality data.In the era of small data, there is no need to care about these things, but in the era of big data, whether the data "outside the enterprise" is up-to-date and accurate is extremely important.

If the data your people collect is of unknown origin or contains serious errors, then the data will be meaningless; if the data is not maintained and updated in real time, it is not the latest data, and it will not produce any value.

The third driving force is our work passion and career planning.

Specifically, it is whether the strategic planning of the enterprise is perfectly combined with the career goals of employees.If not, employee motivation can become an issue.When you want them to enter the big data era and deploy big data technologies together with the company, their thinking is still in the "small data era", and for a long period of time, they must still use small data technologies.

This also tells us that in the technological innovation of big data, the human factor is always the most important.The direction of the enterprise is to strive to train our data scientists, and at the same time improve the ability of existing personnel to analyze data, increase their passion, and increase their awareness of analyzing and utilizing data.If each of our employees is very good at "data" and is extremely sensitive to data, and can often think about things and make judgments independently through data, your company will definitely be able to become stronger, and it will definitely become stronger.

The important thing is that data brings results to work, which can also make employees more motivated to work.These three points are very important for the technical application of big data.Chinese people have been looking forward to big data for a long time, and we have begun to see its shadow in various news programs and financial channels. However, if we want to completely get rid of the technical habits of small data and make big data truly grow, we still need A lot of effort.

Data Service Industry Chain

Drucker, a master of management, said: "The competition among enterprises today is no longer just a competition of products, but also a competition of business models."

For big data to be implemented, three conditions must be met: first, rich data sources, second, strong data mining and data analysis capabilities, and third, the establishment of a complete data service industry chain, that is, a business model.The business model guides how the company earns residual value, so it is very important to establish the company's position in the industrial chain and value chain.

Now, in the IT field, the threshold of analysis technology has been gradually lowered.Many companies have suffered setbacks in their big data strategies due to lack of data sources.They feel confused: "Big data is said to be an opportunity, but where is the opportunity?" They also lost their position in the data service industry chain because of this, but felt that it was not as good as before, so they came up with the idea that it is not as good as the small data model.

If enterprises want to lead in the big data era, they must obtain more data, and clarify their own business model, how to select upstream and downstream partners in the value chain, and how to conclude transactions with customers and provide value to customers.You must know that this is the foundation of big data and the core of the success or failure of big data strategy.The loss of many enterprises is precisely the confusion of business models and industrial division of labor in the new industrial chain.

In the era of big data, there are three kinds of big data companies active in the big data industry chain:

1. Data owner: the company based on the data itself.Possess a large amount of data, but do not have the ability to analyze data.

2. Technology Providers: Technology-based companies.For example, technology providers or data analysis companies.

3. Service providers: Thinking-based companies, that is, big data application companies that mine the value of data.

Playing the role of different industrial chains has different profit models.We can sort out and subdivide the business model of big data for readers' reference.

☆The business model of "data owner"
There are three types of data owner companies:

1. The reuse of big data is the driving force behind its development, among which big data is the core of their business. This kind of company has very strong big data technology capabilities and has three industrial chain roles: data + technology + service.Most of the time, their company's technology is used in their own operations.Such as Google, Amazon, Baidu, Alibaba and other world-class Internet companies.

2. Big data provides the basis for improving the production efficiency of the company, increasing business income or creating new income, and is not the mainstream business of the manufacturer.For example, operators, banks, etc. Currently, operators do not make profits through data reuse, but their main business is various network voice and data services provided by communication equipment.

3. Data middlemen.Such companies collect data from various places for integration, then extract useful information for use, and provide these high-value data to companies that need it, but they do not have the ability to create data themselves.For example, some survey companies, etc.

The business models of these data owners are:
2B: Provide the results of data analysis, mainly for enterprises or government departments.For example, Inrix sells the complete model map of traffic conditions to traffic planning departments, logistics companies, GPS manufacturers, etc.

2C: Provide services based on data analysis results, mainly for individuals.For example, Inrix provides users with free traffic information, but this is a free smart phone application that users can download themselves, but the company itself can get synchronized data.

2D: It is a brand new business model to directly sell data or information as assets and build a data asset sharing and trading platform.For example, Twitter authorizes its data to others through two independent companies; VISA collects and analyzes 210 billion transaction records of 15 billion credit card users in 650 countries to predict business development and customer consumption trends, and finally puts The results of these analyzes are sold to other companies.

☆"Technology provider" business model
The current mainstream business model of technology providers is 2B, of which there are 4 types:
1. Provide single-point technology.For example, Teradata provides big data analysis technology for Wal-Mart, a large retailer.

2. Provide overall solutions, mainly IT manufacturers.For example, the well-known IBM company provides a complete set of big data solutions; China's Huawei's big data solutions are based on the advantages of IT infrastructure in storage and computing.

3. Big data space rental.By renting out a virtual space, it gradually expands from simple file storage to a data aggregation platform, and organically combines with the cloud on the big data computing infrastructure.For example, in Tencent's "Open Cloud" strategy, small businesses also have the opportunity to innovate in the field of big data, providing cheap data infrastructure for big data entrepreneurs.

4. Provide E2E online big data technology or solutions, namely Bigdata as a service.In short, this is a new business model.For example, RJMetrics has a software. Customers only need to input specific data on the software side, and the company will optimize the data within 7 days. After backing up the information to a secure server, the data analysis results will be fed back to the company with a clear and concise interface. client.The software is priced at only $500 per month, but it can provide fast business intelligence online services for e-commerce.In addition, GoodData provides tools such as data storage, performance reporting, and data analysis for business users and IT enterprise executives. Among them, all data required for business intelligence analysis will be performed on the cloud.

Technology providers also have 2C business models, but they are relatively few at present, and there will be a lot of space after combining with the cloud, which will be a huge trend in the future.For example, some companies are oriented to personal household bills, household energy consumption, etc., or some big data solutions oriented to personal data.If you have the demand and the ability to pay the cost, you can get this convenient service immediately.

☆"Service Provider" business model
There are two types of big data service providers, one is the application service provider, and the other is the consulting service provider.

1. The application service provider provides external services, which is a service based on big data technology.

business model--
2B business model: services that provide data analysis results, mainly for enterprises or public government departments.For example, the Inrix company mentioned earlier.

2C business model: Provide services based on data analysis, mainly for individuals.For example, Flightcaster and FlyOnTime.us predict whether a flight will be delayed by analyzing the performance of each flight over the past decade and matching it to weather conditions.

2. Consulting service providers are the big winners of brainstorming. They mainly provide technical service support, technical (method, business, etc.) consulting, or provide some kind of consulting services for enterprises, similar to data scientists.

business model--
2B business model: Supported by a large amount of data, they use data mining technology to help customers develop precision marketing, and predict the behavior of relevant entities after mining and analyzing data.Their income comes from the share of the value-added part of customers.For example, the German consulting company GFK provides location-based personnel flow data, and uses time as the dimension to analyze personnel demographic data (gender, age) and action data in a specific area, mainly for retailers, government departments, and public institutions.This type of enterprise grows very fast and is good at data mining and analysis technology, helping some big data players such as banks and operators to develop new businesses.

(End of this chapter)