From "bronze" to "King", there are three secrets of enterprise digitalization

Digital transformation involves the management of enterprises 、 operating 、 Decision making and many other aspects , If only the past people 、 Financial management action , Solidified into the system through process specification , It can only be regarded as “ Bronze grade ”. This paper starts from the fundamental of digitalization “ big data ” and “ Artificial intelligence ” Starting from the development of the two technologies , Sort out the underlying logic of digital development . meanwhile , Give enterprise digital transformation “ Three steps ” programme .

Now? , Until now, , And the boss asked me ：“ Lao Yang , What is the digital transformation of enterprises ？”

I said, ：“ Boss , Your question is very profound , Let me try to answer , The so-called enterprise digital transformation is to put data throughout the whole enterprise operation process , Help enterprises achieve all business digitalization , All data business .”

Careful students will say ：“ Lao Yang , Aren't you right nonsense , What kind of answer is this , Is there any difference between this and not saying ？ What is all business digitalization ？”

This is really not clear in one sentence or two , Listen to me .

Click on the picture to see 《 New programmers 003》

from “ bronze ” To “ The king ”, Enterprise digitalization also has stages

Let me first sort out the stage of enterprise digitalization . In fact, all enterprises have been walking on the road of digitalization , Even though thorns are everywhere , Never stop . Enterprise digitalization can be divided into the following four stages .

• The first step of enterprise digitalization —— bronze . Generally, enterprise digitalization includes personnel management system and financial management system , This is to put some people before 、 Financial management action , Solidified into the system through process specification . But is that enough ？ Obviously not enough . This is just the first step of enterprise digitalization , This is just to make the basic management actions of the enterprise recordable 、 Traceability 、 Measurable , Most of it is “ bronze ” Level , It is still ten blocks away from the real enterprise digitalization .

• The second step of enterprise digitalization —— Silver . But some enterprises find that only people 、 Information management of finance is not enough , You also need to be “ those , on whom one 's livelihood depends ” Our customers and users are also managed , therefore “ Customer management system ” and “ User management system ” Also on the stage . Careful enterprises will find ： These systems exist alone , Can't be connected in series . It's time to let “ Office automation ” It's coming out . But is it over ？ Obviously not . This is only the second step of enterprise digitalization , This is just to manage the things that the enterprise depends on in the system , Most of it is “ Silver ” Level bar , Five blocks away from enterprise digitalization .

• The third step of enterprise digitalization —— diamond . The above two steps can at best belong to the scope of enterprise informatization , On the whole, it is more process oriented , The main purpose is to reduce costs and increase efficiency in the process . After reducing costs and increasing efficiency to a certain extent , Naturally, there will be a need to generate income , It's human nature . What the boss pursues “ Higher 、 faster 、 stronger ” In fact, it is a new growth point , When an enterprise's main business is on track , The boss will spare no effort to think about new growth points , Without exception . Because most bosses are familiar with “ Life springs from sorrow and calamity. ” Road . Think and think , There are no more than two growth points ： First, a new way to expand the main business ; Second, open up new businesses . By what means ？ In this data 、 In the era of technological explosion , There is really no other way except digitalization , So enterprise digitalization also appears . among , The representative of the new path to expand the main business is Suning , Suning moves offline retail directly to online ; The representative of opening up new business is Wanda , Wanda is not transforming its own shopping center , But to pay 、 Receipt 、 Software system of marketing, etc . There is no distinction between the two , Only the difference between the amount of investment and the size of the risk （ Look back , Indeed, Suning is relatively stable ）. But is this the complete body of enterprise digitalization ？ Obviously not necessarily . This is only the third step of enterprise digitalization , But it is an epoch-making step . This is the No. 1 position of the enterprise, which has identified digitalization from the perspective of ideology , Laid the foundation for success or failure of digitalization , It's probably diamond grade , leave “ The king ” It's only one throw away .

• The fourth step of enterprise digitalization —— The king . New path or new business growing from digitalization , From the production link , To the producer , To the production method , To factors of production , Earth shaking changes have taken place in the production of goods . These are beyond the reach of traditional enterprise management . It needs to be managed from the enterprise 、 Business operation 、 Make fundamental digital changes in many aspects such as enterprise decision-making , This requires a handful of people to pass “ Add, subtract, multiply and divide ” and “ Statistical analysis ” Wait for basic mathematical knowledge to continue , It also requires that most must pass “ machine learning ” and “ Deep learning ” Algorithm can be analyzed to support the conclusion , So that the enterprise can manage 、 Enterprise operation and enterprise decision-making are more digitalized . Of course, the premise of all this is transmission 、 Storage 、 Calculate the force 、 Great progress has been made in basic technologies such as algorithms . Without these technological advances , machine learning 、 Deep learning is empty talk . This has truly entered the king stage of enterprise digitalization . To be honest , Few enterprises can reach this level , It is “ This enterprise should only exist in the sky , How many times can the world hear ” The rhythm of .
  Careful students will say ：“ Why , Lao Yang , You said it was one hundred sixty-eight , What is the definition of enterprise digitalization ？”
  Um. , Children can teach , I didn't miss it .

You make it clear “ Digitization ” and “ Enterprise digitalization ” What's the difference ？

Find out what is “ Enterprise digitalization ”, First of all, understand what is “ Digitization ”.

“ Digitization ” Through computer technology , Connect all kinds of things happening in the real world with the expression of virtual numbers , Through data and algorithms, we can deduce the deep-seated laws of the real world —— All kinds of laws that cannot be recognized by common sense and logic .

What is that? “ Enterprise digitalization ”？

Digitalization of enterprises is to manage enterprises 、 Experience in operations and decision making 、 The method is expressed in numbers , Then reconstruct the business model of the enterprise through data and Algorithm / Service mode , So that the whole process of enterprise operation can be described 、 Measurable 、 Traceability 、 Predictable , Realize the transformational growth of enterprises , Form a new core competitiveness .

Enterprise digitalization is a huge system engineering , It is the data throughout the whole enterprise management , Focus on customers and assets , Based on production links and producers , Grow into digital operation through digital management , And achieve the process of digital decision . Finally, all businesses are digitalized , The result of all data commercialization （ See the picture 1）.

chart 1 Enterprise digitalization logic

Digital management ： The easiest part to ignore

Digital management is the most neglected part , First, the practice of digital management is “ Internal skill ”, What others can't see is often not motivated to do ; Second, digital management is a long-term process , It is difficult to see success in the short term performance . But according to the past data , The key to the success or failure of an enterprise often lies in digital management ,“ Cultivate one's health and make one's family better ” After that “ Governing the country and the world ”, I haven't figured out what else to talk about ？

Digital operations ： The most effective part

Digital operation is the most effective part ： For one thing, operation accounts for 80% of the daily work of the whole enterprise , For example, sales report 、 Sales forecast 、 cost analysis 、 Conversion rate analysis and other links , These works can be realized by digital system , And it is more comprehensive than manpower 、 Scientific and accurate ; Second, as long as we invest in digital operation , A report , A curve can be mapped to the production link , And verified , The effect will be very good .

The effect is reflected in many fast things , Nature will be the place where resources gather . therefore , One time marketing management 、 Commodity management 、 Inventory management 、 Warehouse management 、 Platforms such as supply chain management have sprung up . They are all working on digital operations , I did get very “ The explosion ” The achievement of .

Digital decision making ： The hard part

Digital decision-making is the hardest part , For one thing, many big decisions are the sudden inspiration of managers , There is no logic , So it's hard to digitize or formulate it ; Second, most decisions depend on many influencing factors , However, it is difficult to collect the data of these influencing factors , And the conclusion derived from data and Algorithm , And most of them are unexplainable , And unexplained conclusions are difficult to make decisions , It is more difficult to persuade the team to implement .

however , People gradually realize that data itself is an asset , In addition to being able to guide the development of existing businesses , Data can also provide enterprises with more innovations , Even the change of business model . So you will find , Although data is only an auxiliary means in decision-making , But digital decision-making is imperative . There must be a word ： The future is bright , But the road is definitely tortuous .

from “ optional ” To “ Will options ”

Through the ages , The emergence and development of something must be at the same time 、 The right place 、 and . The large-scale development of the Internet is due to the popularity of personal computers ; The large-scale development of mobile Internet is due to the popularity of smart phones ; The large-scale development of cloud computing is due to chips 、 Memory 、 machine 、 The popularity of network and other hardware ; The large-scale development of big data is due to Computing 、 The popularity of resources such as storage ; The large-scale development of artificial intelligence is accompanied by computing power 、 data 、 The popularity of algorithms .

What about enterprise digital transformation ？ Of course, there is no exception , It also has its own “ days 、 The right place 、 and ”.

Enterprise digitalization “ days ” To put it bluntly, it is the development of big data and artificial intelligence , Understand their development process , That's why I understand “ days ”.

See figure 2 Time axis ,2003 year 、2004 year 、2006 year , Google output GFS、MapReduce、BigTable Three papers , Known as the big data Troika , It has also become the cornerstone of big data .Hadoop It was inspired by the troika .Hadoop It has the following three characteristics .

chart 2 History of big data

• Hadoop reference GFS Build up HDFS, It is a machine that runs on ordinary machines 、 Distributed file system for large-scale storage and access , It's the cornerstone of big data storage . Make big data feasible , Controllable in hardware cost , It can be realized in software technology .
• Hadoop reference MapReduce Build up Hadoop
  MapReduce, It is a way of big data distributed computing , Decompose the computing task of big data to multiple ordinary machines , Then merge to get the calculation result . It is the cornerstone of big data computing , Make big data computing feasible , Controllable in hardware cost , It can be realized in software technology .
• Hadoop reference BigTable Build up HBase, It is a large table for large-scale storage and computing at the bottom , After all, tables are more in line with human needs , You can think of it as NoSQL The cornerstone of the database .
  2006 year Hadoop from Nutch Separate out into Apache Top open source projects . From then on , Technologies related to big data have sprung up ：2008 Data warehouse in Hive;2010 Column database for HBase;2012 Resource manager in Yarn;2013 The streaming computing framework of Spark、Storm;2014 Real time computing framework Flink. These things have made great progress in the big data industry .

Big data is changing with each passing day , Artificial intelligence over there is not to be outdone .

as everyone knows , AI comes from 1956 Dartmouth Conference , It has been developed for more than 50 years , It can be said that it has experienced three ups and downs （ See the picture 3）.

chart 3 The history of artificial intelligence
The first stage , from 20 century 50 S to 20 century 60 The S is the first climax , It is mainly the theorem proof dominated by logic . However , Due to the lack of computing power , And at that time, AI itself did not have the ability to learn ,20 century 70 The s ushered in the first trough of artificial intelligence , Various pressures and funding problems also follow , The prospect of artificial intelligence is also immediately cast a shadow .

The second stage , Fortunately, there is always a small number of people who do not play cards according to common sense continue to insist on research , Probably dormant 10 year , Finally in the 1980 year , Carnegie · The first expert system of Mellon University XCON The birth of .XCON How much cost the system can save the enterprise every year has always been a mystery （ The highest is $40 million , The lowest is millions of dollars ）,XCON Expert system has experienced nearly 10 The golden age of , It is also the second climax of artificial intelligence . However , With the disillusionment of the fifth generation computer , AI has entered the second winter .

The third stage , Experienced two highs and two lows , People's cognition of artificial intelligence also returns to rationality and objectivity , At the same time, the storage and computing power of big data have also been greatly improved , Artificial intelligence technology also has a breakthrough development . So , stay 1997 year , Finally, there is one “ decent ” AI products have come out ——IBM Of “ Deep blue ”. Its victory over Kasparov, the world chess champion, is an important milestone , Experienced two highs, two lows, and two hibernations , AI has finally entered a stage of steady development .

today , It is no exaggeration to say , A programmer who does not understand AI is by no means a good programmer . Why? ？ Here are some facts ：2006 Years later, the deep learning led by neural network has made a great breakthrough ;2016 In, the accuracy rate of Google machine translation reached 87%;2016—2017 year , Google's AlphaGO Amazing performance of ; The global market scale of artificial intelligence reaches 2.43 Trillions of dollars , And nearly every year 30% The growth rate of is increasing ; The major technology enterprises are inextricably linked with AI . All these facts show that AI has basically “ Make it through ” 了 , In the future, we can either do artificial intelligence , Or by AI “ do ”.

Big data and artificial intelligence have gone through 30 Years of precipitation and accumulation , The basic theory and technology have entered a mature stage . On the whole , The big data and AI industries have also entered a period of rapid development . With big data 、 The development of artificial intelligence , Data volume and data dimension that were completely unrecognizable before , Now you can understand it in minutes . So the enterprise digital transformation has really changed from impossible to possible , Enterprise digitalization can also stop simply generating data reports and statistical analysis .

So-called “ The right place ”, To some extent, it can be considered that enterprise digitalization and “ big data 、 Artificial intelligence ” Equate . according to an uncompleted statistic , from 2015 Year to now , The state has promulgated no less than 20 A big data and AI policy （ See the picture 4）.2015 year 8 The month was promulgated. 《 Action plan for big data development 》,2017 year 1 The month was promulgated. 《 Big data industry development plan 2016—2020》,2018 year 4 The month was promulgated. 《 Scientific data management methods 》,2020 year 2 The month was promulgated. 《 Guide to classification of industrial data 》,2020 year 5 The month was promulgated. 《 Guidance on the development of industrial big data 》, The support of big data and artificial intelligence at the national level has been very obvious .

chart 4 Big data national policies
It is the same in foreign countries ,Yahoo、IBM、EMC、 Microsoft has invested a lot of resources to research and use big data and artificial intelligence , Also produced many Apache Top open source projects . The domestic BAT Started relatively late , among B（ Baidu ） More romantic , We follow the idea of "technology first, scene later" , It has gathered the world's top big data 、 AI talent , Basically, it has formed its own big data artificial intelligence ecosystem .A（ Alibaba ） More practical , It mainly applies big data and artificial intelligence to e-commerce 、 Logistics and other retail services enable business . meanwhile , It's on NASA plan .T（ tencent ） large , The main focus is on talent reserve 、 Calculate the force 、 Algorithm . Of course, there are some who try to counter attack “ Promising youth ”, Such as iFLYTEK speech recognition , Shang Tang and Kuang Shi of computer vision , And the squirrel education of intelligent adaptation education . Indeed, it can be called a hundred flowers blooming , I have to sigh here , To catch up with big data AI , There must be “ Two brushes ”.

So-called “ and ”, In the context of this outbreak , Major enterprises have shown their digital capabilities , Also made great contributions , But at the same time, many problems have been found , Such as data collection 、 Data processing 、 Data analysis 、 Data application and other aspects . It proves that the current digital transformation of enterprises is far from enough , We need to break and stand , Whether it is the decision-making level of the enterprise 、 The management and executive levels are also aware of the urgency and importance of digital transformation .

stay “ days 、 The right place 、 and ” In the background of , Enterprise digital transformation is no longer an enterprise “ optional ” 了 , It is “ Will options ”. Through digital transformation , Enterprises in management 、 In operation 、 The decision-making will say goodbye “ pat head ” The day of , Using data to manage enterprises can ensure that enterprises remain invincible in the competition .



Before talking about the implementation path, I need to emphasize again , Enterprise digital transformation is absolutely “ Position one project ”. It's no exaggeration to say , All enterprise digital transformation that is not the responsibility of the first position is “ A paper tiger ”.

The implementation path of enterprise digital transformation is no more than three steps ： Data connection and data access 、 Data processing 、 Data visualization .

Data connection and data access

Why should we take data connectivity and data access as the first step of enterprise digital transformation ？ All enterprises have come from informatization , Informatization is usually a system provided by different suppliers , And these systems will inevitably form One by one “ The chimney ”, The data in these chimneys are limited to a certain dimension , There is no way to conduct multidimensional data analysis without getting through , Not to mention more advanced digital operations .

for instance , Passenger flow of a commercial real estate developer 、 Traffic 、 members 、 The store 、 Data such as commodities come from different suppliers . Now I want to know who likes to go to which shopping malls to buy which goods , This requirement is basically an impossible task for any supplier , At this time, it is necessary to get through the data . But don't think that data access is simply to put the data of various suppliers into a database , This is too elementary . The key to getting through the data is , With the only ID To identify data , Only this one ID Exactly , Only then can we know how many people have entered the field , And what stores and goods everyone likes .

You may ask ： This is only using enterprise data , In most cases, the data on the enterprise side is not enough , Scenario fusion must be carried out with three party data , In order to have more labels for deeper analysis and conclusion output , So what do we do ？ Never use only ID Go to the vast ocean of data “ Salvage ” Well ？

This actually belongs to the category of data access . The premise of data access is to ensure privacy and security . This depends on privacy 、 Privacy computing technologies such as federated learning , To complete the joint analysis of multi-party data 、 Training 、 modeling 、 forecast , So as to realize the circulation of data value , Reach the data “ Available but not available ”. Federal learning is a separate topic , I won't go into details here , A logic diagram is attached for reference （ See the picture 5）.

chart 5 Federation learning logic diagram

Data processing

Data processing is the core step of enterprise digital transformation , The data magnitude of each enterprise （Volume） More and more big , Format and content （Variety） More and more diverse , And enterprises in the depth of data mining 、 On the dimension of analysis 、 The requirements of computing speed are increasing . To mine under such a huge amount of data / Analyze the value , Relying on traditional data analysis methods is basically impossible . This requires that enterprises must be in data collection 、 data storage 、 Data calculation 、 data mining / All levels of data processing such as analysis are stable 、 Efficient technical tools . And the output of these technical tools requires the accumulation of a team of 100 people for three to five years . For most enterprises , It is impossible to cultivate a team of 100 people , Accumulate for another three to five years , Basically “ Lily is cold ”. What do I do ？ This requires big data companies specializing in data platforms and tools to provide technical capabilities in this regard . With MobTech For example , These four levels have formed a complete product matrix and experienced 9 year （ See the picture 6）.

chart 6 MobTech Data analysis matrix Data visualization

Data visualization is the facade of enterprise digital transformation , Everyone can understand this . No matter how valuable the analysis results , All need to be determined by the curve 、 Statements, etc. are displayed at a glance . Of course, rich and powerful enterprises can choose Tableau Commercial software , Small and beautiful enterprises can choose Superset And other open source solutions , Are both good .

To sum up , This paper mainly expounds the definition of enterprise digital transformation 、 Range 、 The need for , And the implementation path of enterprise digital transformation , I believe it can help you have a comprehensive understanding of enterprise digital transformation , Come on together ！

author ： Champion Yang MobTech Broad partners / Chief data Officer . Have 15 More than years of R & D technology management experience , It is a recognized technical expert in the industry . Once served Wanda network technology 、 Alibaba 、 Suning Tesco and many other large well-known Internet companies . For big data and AI Have unique views on digital R & D management , And establish a digital R & D management system for multiple teams , Once published 《 Data enabling ：IT Team technology management practice 》 And other related books .

​ This article from the 《 New programmers · Cloud native and comprehensive digital practices 》. stay 《 New programmers 003》 in , We focus on “ Developers in the cloud native era ” And “ Full digital transformation ” Two themes . Ali 、 Bytes to beat 、 NetEase 、 Well quickly 、 The enablers of cloud native technology of Internet giants such as Amazon , From the technical definition 、 Technology application 、 Practice case sharing, etc , Fully analyze the cloud native data with the hard core output of directly hitting the kernel , Help developers quickly find technology paradigms suitable for their own development in the cloud native era .
meanwhile , We will also talk to Microsoft 、 Intel 、 Huawei 、 Schneider 、 Siemens and other first companies to start digital transformation report , Fresh cases shared by more than a dozen technical experts , A glimpse of Finance 、 The new retail 、 Digital transformation achievements in industrial Internet of things and other fields , Help developers who pay more attention to digital transformation learn from the experience of pioneers .
Read more information and technology related articles , Welcome to scan below QR code subscribe 《 New programmers 003》 Paper book + e-book .
​

Annual subscription （001-004 period ） Enjoy more member benefits

Click picture subscription （001-004 period ）