Improving Email Marketing Campaign Success Rate Using Personalization

Abstract

Email marketing provides one of the best methods for direct communication with consumers. However, the success rate of an e-mail marketing campaign is often low because of its generic content and inadequate segmentation of customers. This paper aims to showcase the application of a two-step personalization process to improve effective open and click rates for email marketing campaigns. Consumer behaviour is monitored over a period of time in terms of email opens and click pattern. This behaviour is stream-lined into keywords sorted in order of user preference. The keywords are updated at regular intervals to account for behavioural changes in user preference. While sending the email, keywords relevant to the campaign are picked individually for each user. These keywords are used to form attractive subject lines using probabilistic language models such as noisy channel model (Mark and Charniak, Proceeedings of the 42nd annual meeting on association for computational linguistics, 2004) and hidden Markov model (Lawrence and Juang, Fundamentals of speech recognition, 1993).