DEVELOPMENT OF ALGORITHMS FOR BUILDING A MODEL OF CONSUMER BEHAVIOR IN THE PHARMACEUTICAL MARKET

The pharmaceutical market is constantly changing towards the development of communication types and effective management of customer needs due to the intensive development of modern technologies. The purpose of the study is to develop algorithms for building predictive models of consumer behavior in the pharmaceutical market. The methods used in this work were: transaction time series analysis, churn prediction, clustering, and data visualization. The materials of the study were the transaction histories of consumers when purchasing goods of the pharmacy assortment in 1121 retail pharmacy organizations located in 8 federal districts of Russia, belonging to the pharmacy chains “Ozerki”, “Doctor Stoletov”, “Samson-Pharma”, “Superapteca”, “Mosapteca” (Big Data). The research period is 2018-2023. Criteria for selecting consumers by transaction histories for pharmacy assortment goods: minimum of 5 transactions, interval between purchases - on average ≤ 150 days, disinclination to discontinue purchases (churn). The developed algorithms include: clustering of consumers of pharmacy assortment goods (based on the criteria: number of consumers, average amount of check, total amount of revenue, amount of revenue per customer), analysis of consumer potential, development of targeted offers that best meet the individual needs and preferences of consumers. Scientifically substantiated algorithms for building a model of consumer behavior allow pharmacy organizations to make management decisions aimed at increasing consumer loyalty, increasing sales of pharmacy products.