Designing Last-Mile Trust: Behavioral Nudges and Community Agents for First-Time Digital Finance Users

Abstract

Last-mile adoption of digital finance often stalls at the first transaction because novices face uncertainty, cognitive load, and thin interpersonal trust. This study investigates whether light-touch behavioral nudges, combined with locally credible community agents, convert intention into verified first use. A community-based field survey (n=250), measured how much participants were exposed to nudge prompts, their perceptions of the agent's credibility, their confidence in using digital financial services, their perceived level of risk, their level of digital literacy and whether they successfully made a first transaction. The descriptive statistics for simple, transparent methods are used to present data and to analyze the relationship between these variables as well as the potential interaction between variables using logistic regression. The logistic regression was done using heteroskedastic-robust standard errors to ensure accuracy in the estimation of both the main and interaction effects. Additionally, predicted probabilities were estimated based on representative values for each variable. The results from the study indicated that nudge prompts had a significant increase on the probability of making a first time transaction digitally; furthermore, the perceived trustworthiness of the agent moderated the effect of the nudge prompts on the propensity to make a first time digital transaction, providing evidence for a mechanism where salient prompts can be most effective when there is a reduction in perceived risk through an agent being physically nearby or otherwise available. In addition to the positive moderation effect of agent trustworthiness, the study also found that the propensity to adopt digital financial services is positively related to both trust in digital financial services and digital literacy while negatively related to perceived risk of using digital financial services