Recommended to look at this article, may be able to help you ~ ~ ~ ~

Hualian super data center director Fu Lihu once told a story: Beijing Hualian as a domestic large commercial supermarkets, every day from the country's stores, transaction data of millions of data, the annual only user purchases of the data accumulated on more than 2TB for data analysis, the application of the demand is very strong. The demand for data analysis is very strong. For this reason, Hualian introduced SAP's BW system for data analysis in 2008, and then introduced SAP's BO product in 2012 to do more advanced data analysis for business guidance.

But what makes Fu Lihu helpless is that, when using SAP's BO to query a hundred million rows of reports, it takes about 20 minutes, and the system will crash if 4 people are online at the same time....... Using expensive foreign software can't solve the problem, and Fu Lihu started to look for a solution in China, which led to the relationship between Hualian and Hazhong BDP.

The story of Hualian is not an isolated one. Recently, the popular "fast fashion" retail start-up Mingchuang Youpin and Hazardous BDP reached a cooperation because the use of SAP's BI system, data aggregation, extraction and presentation time is counted in hours, which is very inefficient. For example, it takes 6-8 hours to export a report, and there are frequent interruptions during the data export process, which is a huge inconvenience for data analysts to analyze in real time. ......

Business Intelligence (BI) is a concept that was first developed by SAP in 1996. This concept was first proposed by Gartner in 1996, with SAP, Oracle and a number of overseas software giants across the ocean to the domestic, was once considered to be following the ERP, enterprise management software field of the new growth of the blue ocean.

However, the harsh reality is that the software giants advocate that the traditional BI implementation failure rate has been high. According to incomplete statistics, the actual application of business intelligence in the enterprise failure rate of 70%, eye-popping.

Traditional BI is dead is not alarmist. The high implementation failure rate, behind the reflection of the traditional BI multiple dilemmas.

First, the technical difficulties. Hualian supermarket and the case of the famous brand, in fact, reflects the traditional BI ETL, data warehouse, OLAP and other technologies, are on the verge of elimination, because it can not solve the massive amount of data (including structured and unstructured) processing problems.

An engineer spat on the Internet: "The original BI miners, extract some samples on a single machine to run a R on the joy, but now it can not, for 50 million users to engage in a three-degree circle of interactions to try?"

Computational performance in the era of "small data" has made traditional BI difficult in the Internet era. Therefore, only updated methods can bring new opportunities. Basically, all the functions of the traditional BI can be replaced by the corresponding big data components, and big data technology has a cost advantage, the elimination of technology is the general trend. Second is the business dilemma. As we all know, whether it is a high-profile large enterprises, or China's 20 million small and medium-sized enterprises, the procurement of SAP, Oracle's software services for the enterprise is an expensive IT costs, the task of informatization of Chinese enterprises can not count on them to complete. If technology can not be universal, technology will always be a game for the few. In addition to the high cost, the traditional software according to the project cycle running the delivery method can not adapt to the rapidly changing needs of enterprises. In the implementation of traditional BI, there is often a phase of the project seems to work well, but the subsequent new needs of the enterprise, new projects become distant, or rotten.

Fortunately, cloud computing has emerged. The concept of Software as a Service (SaaS) has completely subverted the traditional software business - pay-as-you-need, online access to resources, rapid iteration constitutes a new standard cognition of software services for enterprises in the Internet era.

Traditional BI manufacturers shouted how many years of "to help enterprises make informed business decisions", now in addition to a bunch of reporting systems, some decision trees and other statistical algorithms, what is left? Traditional enterprises have introduced so many BI consulting, write so many reports, how much value has really happened? At its root, in the traditional BI vendors, the target audience is only the boss, decision-making and implementation of the disconnect, can not sink to the front line, and ultimately degenerate into a face-saving project, can not produce real value. The failure of traditional BI, technology-led drive business results in the hollowing out of technology. The development of such reports for the purpose of presenting, not on the value of the positioning, was eliminated by history is inevitable.

The enterprise's big data to play a value, the target audience should be aimed at those who are really in the front line of the business to do operations, do analysis, look at the data of the people - why xxx APP registered members of today's decline in the activity of the xxx goods why the morning sold more than the afternoon? Why xxx channel advertising is ineffective for a week? ...... These real business scenarios that are played out every moment, it is impossible to wait for the boss to answer them all. And to really do the staff in the head of the idea can get real-time results, it requires data analysis tools as much as possible to reduce the technical threshold, significantly improve the technical performance, simple drag and drop to show the beautiful data charts, it is best to take into account the PC terminal and the mobile terminal, only the business sector to use the data analysis, the value of the data in order to maximize the play.

Data-driven is not only the boss, data should be dissolved into the blood of every ordinary employee of the enterprise, data-driven will not be reduced to an empty talk.