People use ride shares for many reasons. Sometimes the reason is a random occurrence, like visiting a friend, whereas other times the reason is highly routine, like going to work every day in a ride share or going to an afternoon yoga class twice a week. The authors show that this is in an important distinction: If a customer uses a service as part of their routine, they tend to be a higher-value and overall better customer than if they use the service randomly. Another way to think about this is to imagine two customers: one who calls a ride five times a week at random and another who calls a ride five times a week in a very predictable way. The authors show that, in the long run, your company is better off acquiring more customers like the latter. While these results may sound intuitive, identifying routines in simple transactional data is anything but trivial. The researchers develop a model for achieving this goal by leveraging a form of machine learning called Bayesian nonparametrics to understand customers’ routines. Although the focus is on ridesharing, the model is general, and the authors suspect many companies can benefit from a deeper understanding of their customers’ routines.
What You Need to Know
- In ride sharing data, researchers observe that customers with routines are more valuable than customers without. They hypothesize that this finding holds beyond the context of ride sharing.
- Customers with routines also appear to respond differently to pricing and service disruptions than customers without routines. This is especially true when the disruption happens during a routine.
- How routine your customers are can be a new KPI in understanding your customer base.
Routines shape many aspects of day-to-day consumption. While prior work has established the importance of habits in consumer behavior, little work has been done to understand the implications of routines—which we define as repeated behaviors with recurring, temporal structures—for customer management. One reason for this dearth is the difficulty of measuring routines from transaction data, particularly when routines vary substantially across customers. We propose a new approach for doing so, which we apply in the context of ridesharing. We model customer-level routines with Bayesian nonparametric Gaussian processes (GPs), leveraging a novel kernel that allows for flexible yet precise estimation of routines. These GPs are nested in inhomogeneous Poisson processes of usage, allowing us to estimate customers’ routines, and decompose their usage into routine and non-routine parts. We show the value of detecting routines for customer relationship management (CRM) in the context of ridesharing, where we find that routines are associated with higher future usage and activity rates, and more resilience to service failures. Moreover, we show how these outcomes vary by the types of routines customers have, and by whether trips are part of the customer’s routine, suggesting a role for routines in segmentation and targeting.
Ryan Dew, Eva Ascarza, Oded Netzer, and Nachum Sicherman, “Detecting Routines: Applications to Ridesharing CRM,” Journal of Marketing Research, forthcoming. doi:10.1177/00222437231189185