Big Data Entrepreneurship Needs To Cross Several Hurdles?

In May 10, 2013,

10th anniversary evening - Ma Yun's retirement speech, Ma Yun said: This is a changing era.

Some people have not yet figuring out the PC, the mobile Internet is coming; the mobile Internet is not yet clear, big data is coming.

The age of change is the age of young people.

Ma Yun said this sentence is very critical, he not only mentioned the big data, but also used a sentence to explain the Internet evolution from the PC era to the mobile Internet era, and then stepped from the mobile Internet era to the era of big data.

Several key points are important: in the PC era, the world has spawned a large number of Internet listed companies, including Google, Amazon, Sina, Sohu, New Oriental and so on.

In the mobile Internet era, China's entrepreneurial boom has come to life. Not only has a large number of mobile Internet (including hand travel) companies listed in the US, it has also created numerous entrepreneurial miracles.

The mobile Internet has not only brought convenience to our life, but also pushed the entrepreneurial boom to the peak of history.

Now the problem is coming.

Is the entrepreneurial boom more lively than the mobile Internet era? How to start a business in the big data era? What are the thresholds for big data entrepreneurship?

First answer the first question: in the era of big data, should the entrepreneurial boom be more lively than the mobile Internet era?

As far as I know, No.

Walking on the Zhongguancun Venture Street, you can receive 100 financing BP, maybe 99 are APP and O2O projects, but 90% of the 99 households will attach great importance to big data.

How to start a business in the age of big data? Please understand the threshold of big data.

Threshold 1: Data

Big data big data, no data how to play? So where does the data come from?

BAT enterprises such as Baidu, Tencent and Alibaba have accumulated a lot of data themselves, so they play big data mostly because they are "making a big noise".

Of course, we can also say a few examples of BAT enterprises playing big data, such as Baidu's "Baidu migration", "Baidu actuarial", "Baidu public opinion", "Baidu big data prediction engine" and so on. All these are Baidu's big data products applications; Alibaba's words, "Ali cloud", "Ying Hua Bai", "Qian Bai Bai", "sesame credit", "ant gold suit" and so on, all should have big data technology.

And Tencent, Tencent wide point, Tencent cloud analysis and WeChat also cited big data technology.

If you don't have data, how do you play?

First, you can purchase data through third parties, for example, data hall has a lot of data to sell and share.

Secondly, you can crawl back some data to store with crawler.

Moreover, it is authorized to use big data tools to accumulate data by giving enterprises, developers, webmasters and so on.

New ventures in this area include Talkingdata, friends alliance and DataEye.

Finally, free government, business, and institutions are used to open the data.

For example, the API interface of High German data and the API interface of micro-blog commercial data and so on.

In general, solving data sources is the necessary threshold for big data entrepreneurship.

The key is to see what your project is.

Threshold two: Hardware

In Beijing, I visited a big data start-ups, and they had not yet received financing.

I went to their office to find something particularly sad.

Their employees are crowded in a small room, and two larger rooms are used to amplify the data storage server.

The storage capacity of big data is amazing, which poses new challenges to computer rooms and hardware devices.

This is not the same as the mobile Internet. You make a APP, use computers to develop, servers use cloud servers, and buy on demand.

But big data is no good. You can't store your own data on other people's cloud servers. On the one hand, they are security factors, and on the other hand, there are property rights factors.

Hardware is also one of the threshold of big data entrepreneurship, but not the biggest hurdle.

Incidentally, the big data start-ups I have visited have now completed a million dollar A round of financing, and now their office area is very spacious, congratulating star map data.

Threshold three: talent

I think the biggest threshold of big data entrepreneurship is talent.

Unlike APP, big data entrepreneurs can't play with one person or even a few people.

Start up a team of 10-15 people and start recruiting people. Such teams include Hadoop engineers, Algorithm Engineers, data modeling engineers, architects, NoSQL engineers, BI engineers, etc., all of whom are highly skilled and highly demanding.

How expensive are big data professionals? In the United States, the salaries of professionals in R, NoSQL and MapReduce have reached about $115 thousand a year, which is not much cheaper in China. There is no annual salary of 300 thousand, and you can hardly recruit a big data talent.

According to the China Federation of commerce data analysis Specialized Committee statistics, in the future, China's basic data analysis talent gap will reach 14 million, and in the BAT enterprise recruitment positions, more than 60% are recruiting big data talents.

That is to say, the big data talent of technology is very big, his choice is very wide, or he has already entered BAT business, or he has a high salary in a good company. You have to dig this kind of talent, besides money, stock, option, welfare, etc., are the price that must be paid.

In 2015, -2016 was the most lacking in big data talents for two years. For a simple reason, colleges and universities that had just opened big data subjects had not graduated yet, and the demand for large data talents in the recruitment market was far short of demand.

In addition to BAT enterprises, communications enterprises, power enterprises, financial banking industry, medical industry, industry, game industry, etc., which industry is not recruiting big data talents?

Threshold four: Technology

Speaking of talents, we must say

Now.

Big data technology is not enough for you to understand C++ or R language. Big data has a complete set of its own technology system, including statistics, programming, JAVA, database, Hadoop, Spark, NoSQL, machine learning, Natural Language Processing, algorithm, data visualization and so on.

Hadoop is the only technology and programming language needed.

Many.

Moreover, the big data tools on the market are different from each other, and the technologies needed to use open source software (such as Hadoop, Spark) or SAP (SAP HANA) are different.

High technology requirements, and fewer talents with comprehensive data technology, this has become the biggest problem that restricts big data entrepreneurship.

Threshold five: money

Actually, I don't want to write money, but I have to write money.

Big data industry is not short of capital. As long as you have no problem with the business mode of your startup project, and have strong technical capability and teamwork, whether you are in China or in the United States, there is no problem with a A round, and capital attention is very hot.

But before you get financing, you need a lot of money to start your own business.

Talent, hardware

And technical costs are high.

To understand this, if a few good friends can spend 50 or 3 months to make a APP project, if you want to start a business in the big data industry, please prepare 600-800 yuan to play again.

Threshold six: business model

What is the most profitable industry in China? I think it's e-commerce and online games.

E-commerce and online games are also the fastest realizable industries in the Internet.

And big data, its liquidity is not as simple as online games and e-commerce.

In many enterprises I visited, they have money, data, talents and technology, but they do not know what data they can carry in their hands.

That is to say, big data is not the most clear and direct business mode at present.

Big data is only combined with business scenarios to produce value.

Big data is like crude oil. You know where it is, you can exploit it, but you need to smelt it, and after vacuum distillation, hydrorefining, solvent refining, solvent dewaxing and other refining process, the product is pported to the gas stations after finished oil, so that the vehicle can be fully powered to produce the ultimate power.

Big data also requires a complex process to achieve business value.

Well, you might ask, is big data paction a business model? Personally, I think it depends on what the paction is. The original unstructured data, the data cleaning behind it requires too many processes, and data storage is also a great cost, and the paction price is too high.

I believe that whether it's business users or personal users, we prefer to buy the big data data source that can be used.

You said Jingdong and Tencent completed the first big data paction. I think it's a joke. Did the big data of Jingdong and Tencent come together earlier? I can use WeChat to do shopping directly, and data is interworking. Why trade?

So, the most difficult part of big data business is business mode thinking. If you haven't found a channel to allow big data to cash in, do not rush to start a team business.

It is not enough to have idea in big data industry, and running the whole business mode is the key.