Email Campaign Evaluation Based on User and Mail Server Response

The goal of an email service provider company is to send out a large number emails help its clients realise successful marketing activities. Thousands sent every minute need be analysed in real time reduce spam or phishing. paper describes method that uses real-time tracking key campaign metrics such as the opens count, clicks hard bounces etc., identify campaigns should stopped because they can classified point this solution we do not analyse content. Nevertheless, proposed neural networks are highly effective—the F1-score above 0.95 for any used sample. Furthermore, approach allows us use same model regardless language email. was developed and verified collaboration with Freshmail, leading services Poland. Validation on data provided by confirmed high effectiveness.