Researchers from the University of Cologne, LMU Munich, and Paderborn University found that retirement savings reminders work better when tailored to individual traits rather than sent as a uniform message to all customers.
The team conducted two field experiments in 2021 and 2022 with approximately 41,000 customers of a European private pension provider, mostly in Switzerland. Customers received app reminders about potential tax-deductible savings, randomly assigned as either loss-framed messages (emphasizing what they would miss by not saving) or gain-framed messages (emphasizing what they would gain by saving).
Using causal machine learning to analyze the results, the researchers identified who responded to each message type. Loss-framed messages worked better for people whose jobs involve processing large amounts of information. Gain-framed messages resonated more with older people (ages 50 and older) and those who had already made additional contributions.
On average, both message types produced nearly identical effects. However, substantial differences emerged when messages matched individual characteristics. Among customers most responsive to loss messages, the proportion making additional contributions was 9.4 percentage points higher with the loss message. Among those most responsive to gain messages, additional contributions were 8.4 percentage points lower with the loss message.
If each person received their optimal message type rather than a uniform loss message, the researchers estimate that 51% more customers would contribute extra (6.86% instead of 4.53%). The average savings amount would increase by 18% (2,442 euros instead of 2,062 euros per saver).
Dr. Markus Weinmann, an economist from the University of Cologne, stated, "If everyone received the reminder that resonated with them, around 50% more people would save extra for their retirement. On average, they would pay in almost a fifth more than they would if there were a uniform message for everyone."
The research appears in Scientific Reports.