A new design of the database marketing system
Database marketing and marketing theory as a combination of information technology, system architecture requires a reasonable basis for implementation. Database marketing system for the problem, different domestic and international expert opinion. Some scholars believe that: from the database marketing and customer relationship management (CRM) to analyze the relationship between the point of view, database marketing, though the basis for CRM. But it lacks the CRM sales automation, customer service and business intelligence (data mining, etc.) content. Its database of historical data is often out of touch with the company, so only as a preliminary analysis of type of promotions.
Given the current marketing importance in the enterprise, I believe that this system is a full marketing should be the guiding ideology, covering CRM sales, customer service and business intelligence content of a complete system. Based on the following two main reasons: First, database marketing is the CRM data warehousing, data mining prerequisite. Integrate the two parts can enhance the marketing function of the system. Unified management of marketing work, which have achieved technical possibilities; 2, sales, customer service, although both differ from marketing, but marketing can be said to play their role as the core. This is embodied in: sales data to provide to the marketing department, marketing department to guide through the decision-making step into a smooth sales, added value for customers; customer service to marketing materials to be an important reference, based on customer consumption characteristics of the specific details of the corresponding service, achieving customer satisfaction.
In short, the study of database marketing system should be driven by marketing. Sales, customer service and business intelligence for the accessibility of enterprise-level marketing system, in a sense, it is a full marketing ideas at the core of CRM. The study of this new database marketing system shown in Figure 1. Of the database marketing system is such a system: sales, customer service department through customer contact centers and sales channels to collect customer data, and through the collation of the database to the marketing sector clients. Marketing departments to observe customer data, statistics and analysis, when necessary, part of the data by streamlining the conversion into the data warehouse processing, such as in-depth analysis of data mining, will draw valuable knowledge patterns through visualization that to the marketing department . Marketing and production departments of these findings, the financial sector, information integration, to develop the next phase of marketing strategy and send to each sector as the reference for its activities.
2 Implementation of the new database marketing system
2.1 The sources of customer data: customer contact and contact
Corporate marketing staff without the full real-time customer data and market information. Will be difficult to carry out. Therefore, the contact and liaison with customers has become particularly important. Database marketing, customer contact and liaison has two parts: First, sales of products and services through distribution channels and customer contacts in a timely manner under the customer's sales record: First, the customer through the customer service department to communicate with the customer contact center and communication. Through multi-media, multi-channel integration and intelligent search methods to help clients find the best seats (customer service points) and provide quality services. One of the customer contact center is a set of phone, email, fax, network, communications and a series of digital and non-digital channels of customer interaction platform. In this platform, customers can choose their own way at any time with the company to communicate, and enterprises to understand the views and needs of customers, with the fastest speed to help customers solve real problems, at the same time, the center also try to collect the relevant information. Achieve customer information. Zero-loss "for the future provide the basis for marketing and sales. The advantage is: For enterprises, access to efficient and easy access to all records and information management, improve customer satisfaction; For customers, convenience and choice personalized, real-time interaction is strong, there is intimacy, which have good impression on the enterprise. integration of multiple channels of customer contact center routing services component and the corresponding server to client contact information-based server and database through the appropriate implementation of business logic, find the best seats on behalf of, and make customer information update and records. the whole process for computer processing, customer call efficiency is very high. A typical customer contact center shown in Figure 2.
Customer Contact Centre for customer contact and information-gathering is a good way. But can it with the traditional methods (such as visits, informal contacts, etc.) combined with each other and collaboration, business will be more beneficial is beneficial.
2.2 The handling of customer data
According to various sources of customer data collected there are many problems: repetition, incomplete, not standardized, inconsistencies, etc., they all follow-up analysis and excavation of the great impact. And therefore the roughness of these massive data required in advance of the order. This order includes two parts: First, the end-customer data (sales and customer data) for the screening and selection recorders were unified marketing database, and to consolidate and update the data regularly, can be called data preprocessing part: 2 is a marketing database for further data mining of part of the data, through specific clean-up, integration, transformation and reduction. Into an acceptable form of data warehousing, data mining for the next sufficient time to prepare, can be described as part of the data reprocessing. As the complexity of data processing technology, and is not for specific analysis.
2.3 Consumption of key data: Data mining
Customer data in accordance with the requirements for cleaning and finishing, also need further excavation and analysis, to explore and discover the market value of in-depth information. Data Mining, from a database perspective, it is found hidden in large data sets and their interesting data model derived data into useful information and knowledge of the process, the information and knowledge can be widely used in various applications, such as business management, market analysis and so on. For corporate marketing. It can help the company's massive customer data from the records found in consumer psychology and behavior related to customer characteristics of commercial value data model to assist companies through their development of more accurate and effective marketing decisions.
The basic steps of data mining include:
(1) data cleaning and integration: elimination of noise or inconsistent data, the various data sources together, and the results deposited in the data warehouse. One of the data warehouse is a subject-oriented, integrated, time-varying and non-volatile data sets organized, using multi-dimensional data model design.
(2) data selection and transformation: from the data warehouse to retrieve the relevant data and analysis tasks, and to explore the unification into a suitable form (such as through summary or aggregation operations).
(3) Data Mining: Using intelligent methods to extract data from huge data models.
(4) model evaluation and knowledge representation: According to some interestingness measure, identify the real value of the model and the results obtained through visualization and knowledge representation technologies available to users (such as business managers and other decision makers).
Out the importance of data mining in particular, customers in the business use of a wide range of marketing, mainly characterized and distinguished association analysis, classification and prediction, and cluster analysis and so on. To facilitate understanding, only one of a kind of model described algorithm.
2.4 Clustering Data Mining Case Study: K-means algorithm theory and application of
K-means algorithm process is as follows:
(1) input contains N objects (customer) database and the number of clusters K (the number of customers to points) in the N objects in K were randomly selected as the average of all clusters or centers;
(2) the remaining (N-K) objects to each cluster according to distance of the center assigned to the nearest cluster from each;
(3) to re-calculate the average value of each cluster;
(4) repeat the two steps. Until the average of each cluster does not change. Each class of stabilizing. The results will come under the classification of each specific target group of clusters (with the same consumption characteristics of the customer base).
Cluster analysis applied to marketing, such as large shopping supermarket type recognition for member customers in order to provide better quality service can be a member of this supermarket customer database after Suoxu of consumer data, through the numerical treatment of multi-dimensional data model: customer identification number (total consumption, the number of purchases, income level, ... ...). Which total consumption, the number of purchases, income levels are multidimensional variables directly related with the type of customer data. According to membership (N) the number and types of clients (K) of information such as cluster analysis, the corresponding results obtained clusters, the 2-D simulation results shown in Figure 4.
Clustering result is large-scale supermarket shopping by customers of all members of different dimensions into three distinct target cluster (class), A, B, C, each type of customers have their own consumer preferences and different consumer characteristics. Supermarket can compare and analyze characteristics of different kinds of customers, and conduct targeted consumer marketing.
Data mining is the database marketing system of the core component of analytical CRM applications is also the core technology, plays a very crucial role. However, the complexity of theory and technology, data mining and many need to be perfected. It is worth noting yes, not a database of all types of information are need dredging, and some textual highly of the information (eg, high Wenzifenxi, images, etc) and some non-numerical information may Zhiyaotongguo Some Rengong order, the same way will also help marketing efforts, and decision-making.
2.5 to develop the appropriate marketing decisions
Daily sales data, customer contact information obtained during the data mining model for enterprise knowledge discovery products and services for different marketing activities to provide a good reference for facts, and then production sector, financial sector and other basic data (such as inventory and cost data) were taken into account, the marketing department to develop a corresponding result can be more scientific and rational marketing decisions and strategies, such as the 4P marketing strategy, product strategy only one can be optimized as follows:
(1) according to some product specific sales data and average profit margins of consumer understanding customer preferences. Thus the existing product line or product companies to assess the project, and the corresponding portfolio strategy using the right products (such as product portfolio expansion or contraction. For product line extensions) to better meet market demand;
(2) through the historical sales data mining products to find their life cycle stage in which, according to the products in import, growth, maturity, decline characteristic of different stages of product development strategies accordingly;
(3) new products, in accordance with customer evaluation of a product or service or product satisfaction improved and updated. For these new products because of consumer differentiation of the specific consumer groups. Can for their consumption characteristics of new products, repositioning or developing products to meet their tastes and so on.