Enhancing behavioral nudges with large language model-based iterative personalization: A field experiment on electricity and hot-water conservation

2026-04-04

AI Summary

Traditional behavioral nudges are limited in dynamic environments due to high cognitive load. This study proposes an iterative, large language model (LLM)-based personalized nudging system and evaluates its efficacy through a three-arm randomized controlled trial involving 233 university students in real dormitory settings, comparing LLM-personalized nudges against image-enhanced standard nudges and text-only standard nudges on electricity and hot water conservation. The findings provide the first empirical evidence that LLM-driven iterative personalized nudges significantly enhance energy-saving behaviors, yielding an average reduction of 0.56 kWh per room per day and an 18.3 percentage point increase in adjusted electricity savings, with effects emerging within the first two intervention rounds and persisting thereafter. A consistent but weaker effect was observed for hot water savings, though it attenuated over time. The study further identifies behavioral friction as a critical boundary condition for this approach.

Abstract

Nudging is widely used to promote behavioral change, but its effectiveness is often limited when recipients must repeatedly translate feedback into workable next steps under changing circumstances. Large language models (LLMs) may help reduce part of this cognitive work by generating personalized guidance and updating it iteratively across intervention rounds. We developed an LLM agent for iterative personalization and tested it in a three-arm randomized experiment among 233 university residents in China, using daily electricity and shower hot-water conservation as objectively measured cases differing in friction. LLM-personalized nudges (T2) produced the largest conservation effects, while image-enhanced conventional nudges (T1) and text-based conventional nudges (C) showed similar outcomes (omnibus p = 0.009). Relative to C, T2 reduced electricity consumption by 0.56 kWh per room-day (p = 0.014), corresponding to an 18.3 percentage-point higher adjusted saving rate. This advantage emerged within the first two intervention rounds, alongside iterative updating of personalized guidance, and persisted thereafter. Hot-water outcomes followed the same direction but were smaller, less precisely estimated, and attenuated over time, consistent with stronger friction in this domain. LLM-personalized nudges emphasized prospective and context-specific guidance and were associated with higher participant engagement. This study provides field evidence that LLM-based iterative personalization can enhance behavioral nudging, with behavioral friction as a potential boundary condition. Larger trials and extension to more behaviors are warranted.

Problem

• behavioral nudging
• personalization
• behavioral friction
• sustainability
• large language models

Innovation

• large language models
• iterative personalization
• behavioral nudges
• field experiment
• behavioral friction