Data-driven marketing: how machine learning will improve decision-making for marketers

Email marketing is an effective channel in marketing strategies not only as a tool to increase brand visibility and brand awareness, but also as an excellent tool to help promote and sell. It will continue to be a critical channel for content marketers. Even with the advent of social media and networking platforms, email marketing still remains the most preferred channel for generating leads, informing and influencing customers.

In this paper, we present our experiences using a learning model on predicting the "click" and "conversion" of email-marketing. We present a comparative study on the most popular machine learning methods applied to the challenging problem of email marketing personalization. Subject and sender lines have a strong influence on click rates of the emails, as the customers often open and click emails based on the subject and the sender. We propose a method to aid the marketers by predicting subject-line click rates by learning from history of subject lines. In the first step of our experiences, all models were applied and evaluated by cross-validation. In the second step, the improvement of the performance offered by the boosting has been studied. In order to determine the most efficient parameter combinations we performed a series of simulations for each method and for a wide range of parameters. Our results demonstrate that it is possible to predict the rate for a targeted marketing email to be clicked or not.