In marketing, dealing with massive amounts of data can be overwhelming for a human to handle alone. This is where machine learning comes to the rescue. See in this article how we conducted a special booksales research for Institute of Territorial Marketing and Branding.

‍

Data Science Project with ITMB

In collaboration with the Institute of Territorial Marketing and Branding (ITMB), we conducted a data science project and performed a statistical analysis based on materials provided by the institute. The aim of ITMB's marketing research was to increase bookstore sales. To achieve this, the target audience needed to be identified and understood.

‍

Clustering Customers

The institute had already conducted a social survey and compiled a massive database of different bookstore customers. Our task was to identify the target audience within this dataset. We clustered the customer database and social survey respondents to divide all buyers into groups with similar attributes, while those from different subsets had significant differences. We assigned attributes from the social survey and database to all the clusters we obtained, such as city, gender, average bill, and shopping channel used, both offline and online.

‍

K-means Algorithm and Attribute Selection

Python was the primary tool used for developing the data science project, and we used the K-means algorithm for data clustering. The attributes identified as most important were used to obtain high-quality results. Some clusters had similar attributes, and we identified additional differences between them by dividing them into subgroups, thereby increasing their number.

‍

Results

Our collaboration with the Institute of Territorial Marketing and Branding (ITMB) resulted in a successful data science project that identified the target audience for increasing bookstore sales. By using machine learning and statistical analysis, we were able to provide valuable insights and recommendations for improving bookstore sales.

‍