The latest marketing research shows exemplary promise in using Machine Learning, True Uplift Modeling, and Big Data to boost efficiency and effectiveness of marketing campaigns. Initiated by CEO Karl-Heinz Hopbach, marketintelligence (MI) Germany, and Prof. Dr. Detlef Schoder, University of Cologne (UoC), the aim was to beat traditional, long-lasting direct marketing approaches significantly. MI brings in more than 15 years of experience in data curation, analysis, campaign design, and execution, whereas the team of Professor Schoder leverages MI’s campaign skills with scientific machine learning methods. They teamed-up with Bofrost, one of Europe’s largest direct sellers of frozen food and ice cream, and successfully applied theory into practice. Following the CRISP-DM industry standard for data mining projects, the joint team from MI and UoC has since conducted several campaigns exceeding more than 500,000 recipients. As a result, new customer sales increased by up to 66% compared to randomly selected recipients (control group). And when compared to non-recipients (new customers from households that did not receive a mailing) this number was even up to 131%, in other words, more than doubled. While deploying machine learning methods, the team proved it outperforms standard approaches significantly, while being more cost effective.

Further steps will include an even more fine-tuned and proprietary True Uplift Approach incorporating contextual data, such as, spatial data. See the full report below.