Looking back at 2020, the retail industry "walking at the crossroads of the transformation period" has been hit to varying degrees under the dual pressure of e-commerce shocks and the new crown pneumonia epidemic, and the entire market has experienced a temporary decline. At the same time, the outbreak of the epidemic has also accelerated the structural changes in the retail industry. Online reorganization, digital ecology, and system construction have accelerated, and the continuous innovation of various methods has also accelerated the transformation of retail models to a certain extent. For innovative companies, this is a rare opportunity.
In the post-epidemic era, what new ways are there in the retail industry? Facing the next retail Red Sea market, how will companies maintain their leading positions? Let's take a look at how this clothing company does it.
1. It is an inevitable choice to change from "product center" to "customer center"
In the post-epidemic era, the retail industry has begun a new top-down transformation. In the future society, retail companies must be digital companies as the main form of existence. The management of digital user assets and user operations will become the core competitiveness of retail companies. Therefore, the traditional product-driven retail concept needs to undergo a radical change. This clothing company changed its marketing focus from "product center" to "customer center", and customer relationship management has become the core issue of the company at this stage.
However, the current customer relationship management level still has the following shortcomings:
- 1. The effect of customer segmentation is not good, and different value groups cannot be distinguished reasonably.
- 2. It is impossible to customize personalized service plans for customers, and it is difficult to improve service quality, which leads to difficulties in recovering from loss of customers.
- 3. Existing marketing resources cannot accurately match high-value customers, which seriously hinders the improvement of corporate profits.
In fact, this clothing company also did customer group classification and precision marketing at the beginning, using the method of "Excel tool + VBA language to write RMF model". This method lacks a complete data integration system, each system is independent of each other, data is not connected, and data islands are formed, and the value of data is difficult to be fully utilized. At the same time, as the amount of data grows faster and faster, the existing system can no longer respond quickly.
2. "Smartbi Mining+LRFMC model" solution
Aiming at the current status of the company's customer relationship management, Smartbi proposed a new solution: Smartbi Mining+LRFMC model.
Smartbi Mining has the following advantages:
1. Possessing professional algorithm capabilities, with 50+ mining components built-in;
2. The product is simple and easy to use, business personnel can also participate in data mining, and the whole process of analysis is visualized;
3. The product is distributed computing, which can meet the data of large enterprises Mining requirements;
4. It can be seamlessly integrated with the BI platform, and the mining results can be analyzed and displayed through Smartbi's rich and diverse visualization methods.
At the same time, based on the company’s original RMF model, indicator L and indicator C are added, collectively referred to as the LRMFC model:
indicator L: number of days to become a member
indicator R: time interval of the most recent consumption
indicator F: consumption frequency
indicator M: cumulative consumption Amount
Index C: The
LRMFC model of the average discount of purchased goods is improved on the original model. The company's acceptance will be relatively high. It will meet the needs and the difficulty of verifying the model will be much less. And it can make the grouping of customers more fine-grained, and more effectively classify customer value.
Three, the implementation process of the solution
1. Mining the modeling process
As shown in the figure, the mining and modeling process of the entire project needs to go through five stages: data source, data extraction, data exploration and preprocessing, modeling and application, results and feedback.
(1) Data source
Integrate the source data that users need to collect from CRM customer relationship management, IPOS store management system, SAP and other business systems into the data source database to open up the data of each system.
(2) Data extraction
The data in the data source database is extracted and divided into historical data and incremental data.
(3) Data exploration and preprocessing:
First, explore and analyze the data in the target database. First, analyze the missing values and abnormalities of the data in the target database. Through operations such as discarding, filling, replacing, and deduplication, it is possible to remove abnormalities, correct errors, and To make up for the missing purpose. In addition, perform data cleaning, attribute specification, and data transformation on the data to achieve data availability.
(4) Modeling and application
Decompose customer characteristics based on the LRFMC model, then use the K-means algorithm to group customers, and finally use the grouping results to verify.
(5) Results and feedback
The model is optimized according to the application results of the model, and so on until the model is adjusted to the optimal state.
2.
Put the customer label on the members into 6 categories: low-value members, high-value members, important development members, important retention members, important retention members, and ultra-high consumption members.
Analyze the characteristics of different customer value groups, and implement label management of customer value groups. For the retail industry, effective customer segmentation is an important means of in-depth analysis of customer needs and responding to changes in customer needs. Companies can use customer tags to target different types of customers and combine consumption characteristics formed by past historical data to carry out refined marketing actions.
3. Perform BI visualization display
Data mining results for BI visualization display, effectively carry out data integration, and assist marketing decision analysis
(1) Membership value grouping dashboard.
For example, from the column chart of "Returns and Exchanges of Different Groups of Members", it can be seen that low-value return orders are higher than exchange orders, and the return orders of important development members are basically the same as exchange orders. It is also very high, indicating that low-value members and important development members have more stringent requirements for product quality. You can analyze the reasons for the return order, find several important reasons for customer returns, and develop corresponding strategies for these reasons to improve Product service and quality.
(2) Membership value statistics board
4. Advantages of the solution
1. Liberate manpower : liberate personnel from repetitive and complex Excel analysis, improve work efficiency, and invest manpower to a higher level.
2. Assist marketing decision-making : delineate high-value customer groups through models for Customers provide marketing decision support in the double festival marketing activities.
3. Smoother communication, direct marketing strategy to the first-line business : open data to store sales personnel, reduce the communication cost between each store and the headquarters, improve the speed and efficiency of marketing execution; and stores can also verify and feedback marketing strategies well The correctness of the marketing strategy facilitates timely changes to the marketing strategy.
4. The marketing response rate is improved : Compared with the customer’s original timely and effective marketing strategy using the RFM model, it focuses on the core issues and greatly improves the marketing execution rate.