当前位置:网站首页>Compete for the key battle of stock users and help enterprises build a perfect labeling system - 01 live review
Compete for the key battle of stock users and help enterprises build a perfect labeling system - 01 live review
2022-07-27 10:52:00 【Several stacks of dtinsight】
Digital reading guide :
Last Wednesday was the new data series of kangaroo cloud stack ——「 Number wisdom enables reality Zhan liujiao 」 The first live broadcast of . Kangaroo cloud stack Senior Product Expert Wood Research It brings you 「 The key battle for stock users , Help enterprises build a labeling system 」 The share of , We will finish the essence of live broadcast. , Let's review the content again , Quickly start to build a labeling system .
You can see
Digital marketing is the vanguard of digital transformation
Methodology of label system construction
The construction case of label system in a fund
Live courseware & Video playback :
Add a live broadcast assistant stack at the end of the text , Get courseware for free + Course playback
speech / Wood Research
Arrangement / Hua Xia

Digital marketing is the vanguard of digital transformation
《⼗ The fourth five year plan for the development of digital economy 》 The emphasis on , want ⼤⼒ advance Digital transformation , Forming data-driven intelligent decision-making can ⼒, Improve the overall operation efficiency of the enterprise . Do a good job in digital transformation , Enterprises can produce 、 study 、 for 、 pin 、⽤ And so on , and “ pin ” It happens to be the first key element , Enterprise transformation often starts from the marketing scene , So we say Digital marketing is the vanguard of enterprise digital transformation .
In the process of digital marketing transformation , Because the digital construction process of each enterprise is different , There are often many challenges , Such as :
How to get through enterprise data ?
How to build customer portraits ?
marketing ROI How to improve ?
How to iterate the marketing strategy ?

To solve the above problems , In the business ⽣ Production and business should ⽤ Before , Let the industry digital marketing , We need to get into ⾏:
Data integration ⼀ Storage
OneID Data connection
OneModel Data model construction
Integration of data assets ⼀ management
Lay these foundations , Then we can start to build our own personalized Label system , Make labels the cornerstone of enterprise digital marketing , Help consumers draw more accurately ; Realize the use of data to guide marketing , Instead of just vague experience ; Let the active data flow back to the data warehouse , As part of the tag data source , formation Marketing business closed loop .
![]()

Methodology of label system construction
So how to help enterprises build a perfect labeling system ? We concluded that “ Three goals + Five steps ” Methodology of :

Three objectives of label system construction
● Goal one Solve business problems
To determine what specific business problems we are building a labeling system to solve , What effect do you want to achieve , In terms of time, we should treat it by stages , Small step run , Promote and apply as soon as possible .
● Goal two Business personnel are easy to use 、 Easy to use
The construction goal of the label system is to be flexible and extensible , So that business people can easily derive labels 、 High production efficiency . At the same time, there are many ways to apply labels , To adapt to different business requirements .
● Goal three Sustainable renewal iteration
Label system as an important data asset , We need to continuously improve it , Form a good interaction between business application and data development department , Help produce more valuable data .
Five steps for the construction of label system
● First step Clarify the construction objectives
The technical department and the business department jointly determine the label construction objectives , Starting from the business scenario , Make construction planning from top to bottom , Achieve “ Think before you use ”, avoid “ Use first and think later ”, Ensure the application of labels , do 「 Business command gun 」.

● The second step Label system design
The next step after clarifying the construction objectives is the design of label system .
Before that , Let's start with some labels Important concepts Make some introductions :

The overall process of label system design can be divided into 4 Big steps +2 Big stage :

In the process, we need :
01 Planning entity objects
Determine the object of the label system , Sort out the relationship between labels , Design label system , Create label categories .

02 Probing data
Warehouse building specifications according to standard data , Sort out the data fields of customers 、 Business process 、 Data sheet 、 The amount of data in the table 、 Data distribution, etc , Based on this data , Which labels can be processed .
![]()
03 Design label category
According to the actual business situation , Design label category , Generally established 2-3 layer , It is suggested that a sub category should not exceed 20 A label , No less than 3 A label .
04 Design label content
Include label definitions 、 Processing method judgment 、 Processing range judgment, etc .

● The third step : Label processing and updating
Label processing and updating includes various types of label processing 、 Label test and label online and update steps , Realize the marketing demand at the technical level .

● Step four Label lifecycle management
Label lifecycle management refers to including label evaluation 、 Several management processes responsible for the actual use status of labels, such as label governance and label iteration , Help realize the organic combination of label construction and marketing , Don't waste any action index .

● Step five Label application and reflow
Label application and reflow refers to the label circle group 、 Several specific landing result operations of portrait insight and external service , It is also an essential part of label construction .


The construction case of label system in a fund
In the above, we introduced the methodology of label system construction , How to put theory into practice , Kangaroo cloud stack smart tag products give the answer .
Smart tag management and analysis platform (DataTag), Extract by label 、 Label management 、 Group subdivision 、 Full portrait , Build a business value oriented labeling system and diversified groups , Label data assets , Data tag value , Applied to enterprise intelligent operation and marketing .

Next, we take the case of a fund customer , Let's introduce the practical application of label system construction .

In an all-round digital platform economy era , The digital transformation of enterprises has become an inevitable trend from inside to outside , In the digital wave , The businesses of fund clients are developing rapidly 、 The amount of customer data has increased dramatically , Company to customer 、 product 、 channel 、 Data analysis of anti money laundering and other aspects 、 Operation puts forward higher requirements , In this context , Build a complete label platform , Improve marketing efficiency , No doubt the best choice .
● The customer is in pain
1、 Customer data is distributed in various systems , There is no unified analysis platform
2、 All kinds of data analysis and extraction mostly adopt semi-automatic working mode , No scientific labeling system
3、 Customer activities Long development cycle , Inaccurate operation , Lack of timely tracking of marketing effect 、 Operation feedback is not timely
● Construction plan
1、 docking Star Ring TDH engine
2、 Provide Label implementation services , Sort out the label system for customers , Finished label development , Establish customer labeling system , Sales support 、 service 、 Compliance and other requirements from different perspectives
3、 Convenient label development 、 operation 、 Display and output, etc Life cycle management ; Support individual and group portraits of customers 、 Circle selection and comparison of customer base , Support rapid promotion 、 Smart marketing 、 Precise service

● Construction process
1、 Business level : Based on the actual business status and business needs of customers , Tailor made label system for business
2、 Data level : Return the consumer data in each business system to the data center , According to the label design , Conduct data exploration 、 cleaning 、 modeling 、 Development and other work, and implement online

● Business effect
1、 Label system : Build customers 、 fund 、 fund manager 、 channel 4 A label system
2、 Tag number :600+ label
3、 Processing efficiency : Label processing efficiency from 1 Week shortened to 2 God , Improve operational efficiency to 2~3 Time / Zhou
4、 Tag application : Application in fund marketing 、 Anti money laundering risk control 、 User churn, recall and other directions

边栏推荐
- Delete in MySQL: the difference between delete, drop and truncate
- No Identifier specified for entity的解决办法
- Learning C language together: structure (2)
- Tcp/ip protocol
- File upload vulnerability related
- OpenAtom OpenHarmony分论坛,今天14:00见!附大事记精彩发布
- 基于Spark封装的二次开发工程edata-base,介绍
- MySQL deadlock, pessimistic lock, optimistic lock
- Tdengine helps Siemens' lightweight digital solution simicas simplify data processing process
- ASP.NET Core依赖注入之旅:1.理论概念
猜你喜欢

JVM -- Analysis of bytecode

TensorFlow笔记——基本函数及概念

phpstudy中Apache无法启动

Custom page 01 of JSP custom tag

antd table中排序th阻止悬停变色+table悬停行变色+table表头变色

Use__ slots__ And__ dict__ To save space (it's simply a qualitative leap, and leetcode's personal test is effective)

img src为空或者src不存在,图片出现白色边框

让人深思:句法真的重要吗?邱锡鹏组提出一种基于Aspect的情感分析的强大基线...

Beijing publicized the spot check of 8 batches of children's shoes, and qierte was listed as unqualified

MySQL index, transaction and storage engine
随机推荐
DNS principle and resolution process
Set up Samba service
How to turn off the application of computer self startup
[Linux] install MySQL
MySQL log management, backup and recovery
A brief introduction to R language pipeline symbols (% >%) and placeholders (.)
Share machine learning notes (PDF version) + practical projects (dataset + code)
Metaspolit
Codeforces Round #807 (Div 2.) AB
jvm--字节码浅析
MySQL日志管理、备份与恢复
The largest square of leetcode (violence solving and dynamic programming solving)
ctf (hardrce)
php生成文字图片水印
img src为空或者src不存在,图片出现白色边框
[QNX Hypervisor 2.2用户手册]9.9 logger
Samba server
Mail server
Tcp/ip protocol
It is thought-provoking: is syntax really important? Qiu Xipeng group proposed a powerful baseline for aspect based emotional analysis