Promoting sustainable behavior: addressing user clusters through targeted incentives

The research addresses the urgent need for climate change action and the significant contribution of household emissions. The study focuses on understanding the drivers of sustainable behavior, acknowledging that diverse individual factors influence sustainable choices. Mitigating demand through individual behavior change presents significant potential for reducing greenhouse gas emissions in sectors like transport and food (Creutzig et al., 2021). However, behavior change is complex and influenced by various factors, including technology adaptation, infrastructure use, and socio-cultural adaptations (IPCC). Several theoretical frameworks, such as the COM-B model (Michie et al., 2014), the theory of planned behavior (Ajzen, 1985, 1991), and the protection-motivation theory (Rogers, 1995), attempt to explain these influences, highlighting psychological (attitudes, values, self-efficacy) and social components (demographics, social norms). Cognitive mechanisms, like the attitude-behavior gap and inaccurate perceptions of environmental impact, also play a role. The study argues for personalized targeting, recognizing that different individuals respond differently to incentives and strategies. Personas, established in user-centered research, offer a tool for understanding user attitudes and behaviors, allowing for the development of targeted interventions. The research investigates the effectiveness of an individualized targeting approach by identifying clusters of sustainable intentions and analyzing the differences among them, aiming to inform the design of personalized interventions.

The literature review examines existing frameworks for understanding sustainable behavior, such as the COM-B model (Capability, Opportunity, Motivation), the Theory of Planned Behavior (attitude, subjective norm, perceived behavioral control), and Protection Motivation Theory (self-efficacy, rewards, costs). It explores various influencing factors including psychological elements (attitudes, values, self-efficacy, personality traits), social factors (demographics, social norms), and cognitive mechanisms (attitude-behavior gap, inaccurate environmental impact perceptions). The review highlights the potential for individualized interventions based on the diversity of these factors. The existing literature on individual factors influencing sustainable behavior is extensively reviewed, but formal analyses of different user clusters are scarce. The authors utilize the established use of personas in user-centered research to address this gap, drawing upon examples of persona development in research and practice (Brickey et al., 2012; Pruitt & Grudin, 2003; Tu et al., 2010; Czioska et al., 2021; Gatterer & Tewes, 2023; Kroth, 2019; Brincken, 2022; Tautscher et al., 2020).

The study employs a mixed-methods approach combining qualitative and quantitative methods. Part 1 involves semi-structured interviews (n=10) to explore individual differences in sustainability factors, enablers, and impediments related to washing habits. Participants were recruited through an agency and represent a small sample of the German population. The interviews covered washing habits, attitudes towards sustainability, and the role of sustainability in purchasing decisions. Data analysis used reflexive thematic analysis (RTA) and the KJ method (affinity mapping) to identify themes and create five qualitative personas. Part 2 uses a quantitative online survey (n=342) to validate the personas through data-driven clustering. Participants were recruited through an agency, aiming to represent the age demographics of Germany's adult population. The survey employed validated scales covering sustainable attitudes and values, knowledge and information, social factors, price, and perceived consumer effectiveness. Data preprocessing involved exclusion criteria based on consent, age, completion rate, and engagement time. Hierarchical Density-Based Spatial Clustering of Applications with Noise (HDBSCAN) was used for clustering, and a decision tree was employed to enhance interpretability. The Calinski-Harabasz index and silhouette scores were used to evaluate clustering results. A post-hoc Tukey's HSD test examined significant differences between clusters.

Part 1 identified three main themes: prerequisites for sustainability (understanding, knowledge, self-efficacy, attitudes), enabling factors (social factors, tangibility), and barriers (stress, time, price, convenience). Five qualitative personas emerged: Activist, Sustainability-interested, Hedonistic, Indifferent, and Dismissive. Part 2, using HDBSCAN clustering on survey data, identified five clusters mirroring the personas: Socially Sustainable (strong belief in climate change, high engagement, social motivation), Responsible Savers (environmental concern, knowledge of sustainable products), Unconcerned Spenders (low belief in climate change, hedonistic, low engagement), Comfort-Oriented (low engagement, prioritizes comfort and money), and Skeptical Consumers (skeptical about sustainable labels, low engagement). The decision tree revealed key factors driving cluster assignments (Figure 2). Table 2 provides descriptive statistics of the clusters across various scales. Table 3 shows significant differences between clusters regarding economic benefit, care for sustainability, social status, and skepticism. The qualitative personas were mapped onto the quantitative clusters, showing considerable overlap but also differences due to the limited scope of the quantitative analysis (e.g., exclusion of stress and time as barriers). Costs and convenience emerged as significant obstacles across all clusters.

The study's findings address the research question by demonstrating the existence of distinct user clusters with differing motivations and barriers regarding sustainable behavior. The mixed-methods approach strengthens the validity of the results. The five identified personas and clusters provide a practical framework for targeting specific user groups with tailored interventions. The significance of the results lies in providing user-centered insights to inform product development, communication strategies, and personalized interventions that aim to overcome barriers and motivate sustainable practices. The limitations of self-reported intentions and the potential attitude-behavior gap are acknowledged. The need for future research to validate these findings through behavioral observations and experiments is emphasized.

The study contributes by integrating individual needs into easily usable personas, validated by data-driven clusters. This framework bridges the gap between theory and practice, enabling systematic targeting of user groups in product development and communication. Future research should focus on validating the effectiveness of personalized interventions, expanding to larger samples and different cultural contexts, and regularly reviewing segmentation variables. While individual action is crucial, a concerted effort by governments, the private sector, and civil society is needed for a sustainable transition.

The study's limitations include potential sample biases (online access panel, exclusion criteria, German context), respondent fatigue in the survey, the limited number of factors validated quantitatively, and the reliance on self-reported intentions rather than direct behavioral observation. The potential for selection bias due to data exclusions is acknowledged, particularly concerning the exclusion of participants who completed the survey too quickly. The study focuses primarily on self-reported intentions, which might not fully reflect actual behavior, emphasizing the need for future research to assess actual behavior and reduce the potential attitude-behavior gap.