2023 Vol.4 Mar.No1

• Title: Analysis of the sales of Tmall Vitamins
• Name: Wang Siqi
• Company: Tianjin University of Commerce 300134
• Abstract:

  COVID-19 has accelerated the process of digital transformation of traditional industries enabled by digital technology. Based on the sales of Tmall Vitamins in 2020-2021, this paper analyzes the sales situation and advantages from the dimensions of stores, drugs and brands, and forecasts the total sales in the next three months.

   

  The data in this paper is from the Saikr platform. The data shows the product information and sales of Tmall Vitamins, with a total of 75110 rows and 10 columns. First of all, according to the analysis of store data, there are 26 stores in total, and the store with the highest proportion of sales is Alibaba Health Pharmacy, accounting for 44.99%.

   

  Because the drug id is unique for each drug, a total of 9697 drugs can be obtained by analyzing the drug id. After that, we can get the top 10 drugs with the highest sales volume, and then draw the monthly curve of the total sales volume of these 10 drugs.

   

  This paper uses the processed data to calculate 456 brands. The top ten brands in sales volume can be obtained by sorting in descending order of classification and summary. The sales volume of elevit is the highest, accounting for about 2.3% of the total sales volume. The reason why these ten brands sell well may be that the drugs of these brands are sold in popular stores and the price of the products is reasonable by using the importance ranking method of RandomForest characteristics. 

   

  Then this paper firstly extracts the monthly total sales data to get the time series of sales. Use the time series decomposition prediction model to predict the total sales in the next three months. The prediction results are more accurate, with an accuracy of 80%, and the operation speed is faster. The forecast results for the next three months are: 139708976.9 yuan, 97113714.98 yuan, and 103019100.4 yuan.

   

  Finally, this paper gives some e-commerce business strategies from the following dimensions: price, discount, time, promotion strategy, creating public praise, and selecting corresponding businesses for sales.

  
  

• Keyword: RandomForest characteristics importance ranking; time series decomposition prediction; model classification summary
• DOI: 10.12250/jpcarme2023040126
• Citation form: Wang Siqi,Analysis of the sales of Tmall Vitamins,2023,Vol.4,6-10.

[1]Lu Ju. Optimization Research of YBB Pharmaceutical E-commerce O2O Platform[D]. Guangxi University,2022.

[2]Huang Jing. Research on China's pharmaceutical e-commerce model[D]. Shanghai Jiaotong University,2014.

[3]Ding Haifeng,Li Liqing. Time series prediction and analysis of total health expenditure in the Yangtze River Delta region of China based on ARIMA model[J]. Chinese medical management science, 2022,12(02):4-10.

[4]Yang Lu. Prediction of high turnover of listed companies based on data mining[D]. Nanjing Normal University,2021.