Measuring marketing effectiveness is hard. Campaigns are reported on different KPIs because they target users in different stages of the consumer journey or they lack measurability. For a marketing lead, responsible for both internal and external specialists, understanding the returns on marketing investments is a nightmare. We solve this problem through data science models that generate continuous recommendations for improved performance. Our model can calculate the returns of marketing investments across different channels and campaigns, and even across offline and online, by learning from millions of rows of marketing data.