Interactions of banks with their customers are increasingly shifting to web and mobile channels. Being at risk of losing the role of the customer agent to FinTechs and digital challenger banks, incumbent banks are seeking ways to exploit technologies such as mobile phones as channels to generate insights and to sell products that their customers need. For many banks the potential of mobile banking to individualize products and personalize services from information collected by the mobile device' sensory components are largely untapped. In this paper, we draw on design science research in exploiting spatio-temporal information to build an algorithmic model to target customers with credit offers. While past research aimed to solve similar problems mainly through customer segmentation, our approach demonstrates the benefits of having a transparent and interpretable decision model for each individual customer. Our artifact enables the development of digital products and services, without large-scale, often unavailable data.