From the application status of big data, see where the business operation decisions go

What is big data?

What is big data? In most people's understanding, companies use data to optimize their processes, products, and decisions to make operations more effective. But I think this does not cover the full scope of big data.

In fact, big data is a comprehensive concept that encompasses both technical and commercial aspects. On the one hand, it is technical. On the technical level, in fact, from the moment the computer was born, it was accompanied by the generation of data. But for massive data storage and efficient data calculation, a very powerful hardware system is needed to support it, and the cost of millions of dollars of hardware and the maintenance cost of tens of thousands of dollars per month are not affordable for every enterprise. In the face of high costs, the use of data analysis has become a kind of “unexpected” wealth asset of the enterprise.

But in recent years, technology has continued to evolve, and the emergence of distributed storage and computing systems like Hadoop has greatly improved the efficiency of data storage and computing, making it possible to apply massive amounts of data to business. At the same time, with the rapid development of the Internet industry, the concept of big data has begun to attract more and more people's attention.

The other is commercial. For business conduct, the most important thing is to enable companies to get more revenue through the use of data. The traditional data is business-oriented. For each business line, there will be data collection and accumulation. I believe many companies have done a good job in this area. It can be said that the private level is enough. If even the private level is not done well, then I think it is necessary to establish data construction first. After all, there is data to carry out the following related content. What makes business more valuable and even disruptive is the diversity of data. This diversity means that multiple data can be mutually verified, and the data can be generated with possible reliable commercial value through the association and interaction between them.

For companies, a scientific and effective data indicator system can increase the choice of business decisions and the foreseeable range. Often, data can be grouped into four categories by whether it has a positive effect or whether it can be foreseen.

Those with positive and predictable data are often focused on operational metrics, and those with adverse effects and predictable data are often circumvented as risks, and how data metrics are divided is determined by the specific business situation. But in addition to predictable data, there is a large amount of unforeseen data. For example, the double eleven, Taobao's goal is to sell 10 billion a day, but the result is 19.1 billion, then 9.1 billion is an unforeseen surprise. For us, we need to turn unforeseen into foreseeable, that is, turn surprises into predictable fixed income, let it play more value, turn unexpected tragedies into predictable, and most likely Avoid it.

If the above is an explanation of the definition of big data, then the following is the relationship between business and data.

On the one hand, it is internal data of the enterprise, including structural data of companies such as financial data, operational data, and market data, as well as analytical data and mobile application data of the website; on the other hand, external data of the enterprise, including social external data such as Baidu and 360. And third-party data such as TalkingData to complement its business data dimensions.

For example, if a user sees a service or product but does not generate consumption, this set of data may not appear in the business data of the enterprise, and absolutely only appears in the analysis data of the website. In other words, if you want to know the future opportunities of the company, at least it is possible to turn those users who have seen but not used into customers. If you can convert 20%, how much will your market increase?

For another example, if you can't explain the reason for the increase in market share, then this is an “surprise”; but when you understand the data, when you understand the cause of the “surprise” and make corresponding adjustments, it will be unpredictable. The possibilities are getting less and less, making it positively stable, and the company will be smarter (Data Smart).

Compared to the exploration of unforeseen “surprise”, the vigilance of unknown accident risks is even more important.

In today's competitive data, the news from the media is now no longer just to see what the opponent did, but more to feel the crisis from those "small voices." For example, 10% of users of Company A used to go to Company B to see Company A and then to Company A. Now the ratio has become 40%, which means that Company B's influence is bigger than before, and it is dangerous for Company A.

Therefore, not only the analysis and use of internal data, we also need to supplement our data analysis dimensions with some unstructured external data. Social data such as Weibo index and Baidu index is a large unstructured data. These social data are not only used to evaluate the company's reputation, but also help the company make some decisions. If you simply stay in your own data, you are often prone to blindness, and use one-sided data to erroneously depict the user's overall picture. Just like when the seller of Taobao leaves the support of Taobao data, it can only be called data analysis, and it must not be called big data analysis.

The so-called big data is a huge amount of data that needs cross-view, cross-media, and cross-industry. It can also be understood as a huge innovation in the collection method of data compared to the traditional collection methods in the past. When the scale and richness of the data reached a certain level, everyone began to propose the concept of big data.