Purpose – Over the last decade, technological and social trends have significantly influenced the relationship between customers and insurers. New buying patterns, price comparison platforms and the usage of different interaction channels driving single-product purchases and impacting lapses have influenced insurers’ customer portfolios and development. The purpose of this paper is to study the features driving the customer relationship along three areas, namely, customer acquisition, development and retention. Design/methodology/approach – After defining 14 related hypotheses, the authors use econometric analyses to quantitatively support these hypotheses in the three areas of interest. The authors build on a large-scale longitudinal data set from a Swiss insurance company covering the period from 2005 to 2014 and including 2,757,000 customer-years. The data comprise information on private customers, their contract history, including coverage and losses and the channels used for buying insurance. This analysis focuses on the two most common non-life insurance products, namely, household/liability and car insurance. Findings – The authors provide descriptive statistics and results from econometric analyses to determine the significant features and patterns affecting customer development and retention. Among the main results, the authors underline the significant influence on cross-selling given by the customer’s age and the interaction channel. Customers from rural regions are more loyal and likely to conduct cross-buying when compared to their peers from urban regions. Car insurance holders are more likely to lapse than household/liability insurance clients. Finally, while newly acquired customers tend to buy only a single product, the authors show the importance of cross-selling for retaining customers. In fact, customer retention is positively influenced by the number of products hold. Research limitations/implications – This work is relevant for academics and practitioners alike, adding a quantitative basis to the understanding of managing customer relationships and for the development of further prospective models. Further work could investigate or add products, extend the study to other companies and focus on customer development with time. Originality/value – This study explores a large-scale longitudinal data set. The analyses of customer acquisition, development and retention can support insurers to construct their own models for customer relationship management.