Creating, maintaining and retaining relationships with a company’s customer base is one of the
crucial business and marketing tasks in heavily competitive markets such as telecommunications
or subscription-based models. The researchers therefore take this thesis as an opportunity to
thoroughly explore the topic of Customer Relationship Management, more specifically Customer
Churn and Retention Strategies. The topic is explored via systematic literature review from
different angles counting business context, degree of CRM individualization, segmentation,
selection of key customers, enabling process, employee engagement and performance
assessment. Additionally, the contemporary tool of Deep Learning is explored and utilized in
order to predict customer churn and yield supplementary marketing insights. As a result, a
conceptual framework compositing the existing literature on the topic of CRM Retention
strategies is created and later synthesized with the outcome of the Deep Learning Customer
Churn Prediction model. This approach allows the researchers to conclude to what extent can the
latter enhance the former. The results indicate that the utilization of Deep Learning in predicting
customer churn represents an effective way to cope with customer defection. Not only is this
approach capable of accurately predicting which customers are going to churn, it can also
provide understanding of churn predictors which in combination with churn reasons can yield
valuable marketing insights. These results can be supplemented by additional information such
as customer lifetime value or proneness to marketing interventions. However, such a model
ought not to standalone in the process of the whole CRM Retention strategies development as
multitude of other factors such as business context or retention goals come into picture.