深圳地区购买公司软件客户价值评估数据

采集销售记录表中深圳地区的数据，通过客户在2022年1月1日距离2025年3月31日间隔的最近一次消费时间天数（R）、客户在2022年1月1日至2025年3月31日之间消费频次(F)和客户在2022年1月1日至2025年3月31日之间消费金额(M)， 采用 RFM 模型对客户进行价值评级，实现精准化运营，通过对深圳地区客户价值管理，满足不同价值客户的个性化需求。对A级客户，每个月进行一次回访维护，对B级客户，每个季度进行一次回访维护，对C级客户每半年进行一次回访维护，对D级客户每年进行一次回访维护。另外可以为本地区客户群体高度重叠企业提供不同价值类型的客户个性化服务的数据支持。数据处理：对从销售记录表中采集到的数据进行脱敏、降噪、清洗、聚集、分析。2、数据加工：运用RFM模型结合客户在2022年1月1日距离2025年3月31日间隔的最近一次消费时间天数（R）、客户在2022年1月1日至2025年3月31日之间消费频次(F)和客户在2022年1月1日至2025年3月31日之间消费金额(M)的得分排名对客户进行一个综合排名，最终得出一个RFM总评分。a.提取出最近一次消费时间距离当前分析时间的天数（R）、客户在2022年1月1日至2025年3月31日之间消费频次(F)和客户在2022年1月1日至2025年3月31日之间消费金额(M)进行分类，最近一次消费时间间隔最短的客户排在最上面。按照从1-5评分，前20%的客户获得5分，接下来的20%用户获得4分,再下来20%的客户为3分，再下来20% 的客户为2分，最后20% 的客户为1分。 b.根据客户在2022年1月1日距离2025年3月31日消费频次(F)从高到底依次对用户进行分类，前20%的客户在用户活动频率的分数为5，以此类推。 C, 根据客户在2022年1月1日距离2025年3月31日消费金额(M)，前20%的客户在消费金额的分数为5，以此类推。消费金额最少的20%客户则分数为1。 RFM得分=0.3*（R得分）+0.3*（F得分）+0.4*（M得分） 评分大于等于4分的为A级客户，大于等于3小于4的为B级客户，大于等于2小于3的为C 级客户，低于2的为D 级客户

This dataset is developed by collecting Shenzhen-region data from sales records. Three metrics are defined for customer value assessment using the RFM model:
1. R (Recency): The number of days between the customer's most recent purchase within the period from January 1, 2022 to March 31, 2025 and the analysis cutoff date March 31, 2025;
2. F (Frequency): The total number of purchases made by the customer during the period from January 1, 2022 to March 31, 2025;
3. M (Monetary): The total consumption amount of the customer during the period from January 1, 2022 to March 31, 2025.

The goal of this work is to enable precise customer operation management by grading customer value via the RFM model, so as to address the personalized demands of customers with different value tiers in the Shenzhen area. Targeted retention maintenance is implemented as follows: Grade A customers will receive monthly follow-up and maintenance; Grade B customers will be serviced once per quarter; Grade C customers once every six months; Grade D customers once per year. Additionally, this dataset can provide data support for enterprises with highly overlapping local customer groups to deliver personalized services tailored to different customer value types.

Data preprocessing steps include desensitization, denoising, cleaning, aggregation and analysis applied to the collected sales record data.

Data scoring and processing: A comprehensive customer ranking is generated based on the score rankings of the three RFM metrics, and a final total RFM score is calculated. The specific rules are as follows:
a. Scoring for metric R: Sort customers in ascending order of the days between their most recent purchase date and March 31, 2025 (customers with shorter intervals rank higher). Assign scores from 1 to 5 using a 20% percentile grouping rule: the top 20% of customers receive 5 points, the next 20% get 4 points, the subsequent 20% get 3 points, the following 20% get 2 points, and the last 20% get 1 point.
b. Scoring for metric F: Sort customers in descending order of their total purchase frequency during the 2022-01-01 to 2025-03-31 period. Assign scores from 1 to 5 following the same 20% percentile rule, with the top 20% getting 5 points and the bottom 20% getting 1 point.
c. Scoring for metric M: Sort customers in descending order of their total consumption amount during the 2022-01-01 to 2025-03-31 period. Assign scores from 1 to 5 following the same 20% percentile rule, with the top 20% getting 5 points and the bottom 20% getting 1 point.

The total RFM score is calculated as: Total RFM Score = 0.3 * (R Score) + 0.3 * (F Score) + 0.4 * (M Score). Customer value grading is conducted based on the total score:
- Grade A: Total score ≥ 4
- Grade B: 3 ≤ Total score < 4
- Grade C: 2 ≤ Total score < 3
- Grade D: Total score < 2

搜集汇总

背景与挑战

背景概述

该数据集包含深圳地区720条客户价值评估数据，采用RFM模型对客户进行分级，适用于精准化运营和客户价值管理。