Personalised E-Newsletter Marketing
In the Digital Era, companies are looking for an innovative way to engage with customers. The new technologies give us the opportunity to interpret data and factors such as customer history, gender and age, in order to define business strategies and predict the future probabilities of customer behaviours. Customer segmentation and predictive analysis have become relevant elements to be taken into account in the asset management of a company, regardless of the reference market.
Business application: Personalised E-Newsletter Marketing
The main business goals are to extract valuable information to define marketing strategy and to segment customers, predict customer behaviours and recommend personalised E-newsletter contents using machine learning techniques. These approaches are fundamental to improve the E-newsletter offer strategy and better understand the needs, wants and expectations of readers, plus identifying inactive users, evaluating the performance of the marketing campaigns, and finally reducing and saving the firm costs.
The methodology and models applied are the regression and the cluster analysis:
- The regression analysis is the primary machine learning technique used for predictive analytics. Regression is used to find out which of the variables used are most significant in explaining, for example, the e-newsletter reading frequency
- The cluster analysis tries to segment customers in homogenous groups based on several variables, from demographic variables to total readings. It relies on the discriminant analysis to check if groups are statistically meaningful and if variables significantly discriminate between the groups. Common cluster models include, for example, behavioural analysis and category-based analysis.
By: Giovanna de Vincenzo