Social-Sciences- and Humanities-Based Profiling of Energy Consumers Towards Increasing Demand Response Engagement

Abstract

This paper investigates the effectiveness of demand response (DR) programs across various European residential contexts by examining the propensity of households to participate in energy management strategies. Utilizing a comprehensive, literature-driven questionnaire, this research collected 284 data entries from six European countries, including Denmark, Italy, Greece, Spain, Austria, and Romania. Through a multidimensional segmentation methodology, residential users were categorized based on their responses, revealing varied potential for adaptive DR programs. Key findings show a strong positive correlation between energy literacy and DR willingness—suggesting that informed consumers are more likely to participate in flexibility programs. Notable barriers included technological concerns, financial limitations, and a lack of awareness. Motivational factors ranged from financial incentives to environmental and social considerations. Segment-specific insights enabled the identification of tailored outreach strategies, recommending different engagement pathways for high-potential versus low-readiness groups. The results emphasize the importance of tailored DR strategies informed by distinct consumer profiles. Policy recommendations underscore localized, personified approaches to enhancing DR participation and supporting a sustainable energy transition.

1. Introduction

1.1. Background and Literature Review

In recent years, power system resilience has become increasingly dependent on the effective alignment of supply and demand [1]. The current global initiative to develop a more intelligent, environmentally friendly, and dependable electrical grid is a primary focus in the endeavor to address climate change and establish a sustainable energy landscape [2]. Renewable energy sources (RESs), such as solar and wind, play a crucial role in the process of decarbonizing the energy industry [3]. Nevertheless, the unpredictable and sporadic characteristics of renewable energy sources present considerable obstacles to the efficient management of grid operations [4]. Consequently, the successful integration of these sources necessitates the development of inventive strategies to successfully balance the supply and demand of electricity [5]. This challenge escalates as electric vehicle adoption advances the integration of the electrical grid with transportation [6] and as the adoption of electric heating raises residential electrification [7,8].

To address the complexities posed by variable RESs and evolving consumption patterns, the implementation of demand response (DR) programs has emerged as a compelling strategy [9,10]. DR entails providing incentives to shift or shed the electricity demand in wholesale and ancillary power markets to help balance the grid. DR programs are principally classified into two types: price-based programs (also known as implicit DR), which employ price signals and tariffs to incentivize consumers to modify their electricity usage patterns, and incentive-based programs (also known as explicit DR), which provide direct financial compensation to consumers who adjust their usage as part of a demand-side response program [11,12]. Such mechanisms seek to modulate the electricity consumption of flexible loads at specific time intervals, with the objective of curtailing non-renewable and expensive electricity generation during peak demand periods [13]. Flexible load refers to time-varying electricity demand that can be altered in response to DR signals, thereby contributing to peak cutting and valley filling [14].

By actively engaging electricity consumers in this process, their flexible load potential is harnessed to reduce the mismatch between energy production and consumption [15]. DR programs advance energy efficiency, grid stability, and carbon emission reductions. These programs empower consumers to actively participate in grid stabilization by adjusting their energy consumption patterns based on real-time grid conditions. The integration of DR improves grid infrastructure use while limiting the need for costly investments in grid capacity expansions, rendering it a cost-effective and environmentally friendly solution [16]. Additionally, the deployment of DR encourages energy efficiency and conservation, fostering a more sustainable and resilient energy ecosystem by promoting responsible and strategic energy consumption [17].

Attention has gradually shifted towards unlocking the vast potential of residential consumers as a promising source of flexibility in decarbonized energy systems [18]. Households enrolled in DR programs, such as time-of-use pricing programs, experience electricity price variability throughout the day, with higher rates typically occurring during periods of peak grid demand [19]. However, the efficient exploitation of flexible loads largely depends on residents’ energy consumption behavior, which varies from one individual to another [20]. Alongside the growing interest in energy management technology, smart home solutions, and Internet of Things (IoT) ecosystems [21], significant progress has been made in personalized DR programs for residential users through the application of advanced clustering algorithms [22,23,24,25,26]. Nonetheless, a notable scarcity of real-world demonstrations and validations of these algorithms persists, often due to a lack of data [27,28]. Furthermore, frequent neglect of critical personal aspects leaves a fundamental gap in understanding electricity usage patterns. To overcome these challenges, employing questionnaires to gather essential data represents a viable approach to accurately identifying and grouping residential users. Its advantages include cost effectiveness and reliance on rudimentary technology, as well as the provision of insights for operant DR programs. Importantly, most existing DR segmentation studies have been single-country, focused on a narrow subset of constructs (e.g., price responsiveness or socio-demographics alone), or stopped short of translating findings into operational personas that can guide program design. In settings where smart-meter traces are unavailable or sparse, there remains a methodological gap for transparent, theory-guided segmentation based on harmonized survey instruments applied across countries.

The questionnaire shall examine several elements that could significantly impact users’ propensity to participate in DR programs [29]. A comprehensive understanding of pertinent complex factors is crucial in advancing the development of efficient and tailored DR programs that can successfully engage and accommodate various consumer groups [30]. Given the enormous variations in traits and behaviors among households, it is crucial to acknowledge these drivers in order to develop strategies that effectively cater to the individual needs and preferences of consumers [31]. Research conducted in various global settings has used structured surveys to explore these factors (

Table 1. A comparative summary of survey studies focused on DR programs, detailing the number of respondents and geographic locations. The proposed survey in this study extends previous efforts by thoroughly examining all key examined dimensions—social demographics, energy literacy, DR willingness, drivers, and barriers—across six countries.

	Social	Energy	DR			No.	Country
	Demographics	Literacy	Willingness	Drivers	Barriers	Respondents
[32]	✓	✓	✓			300	Xi’an, China
[33]	✓	✓				186	California, USA
[34]	✓	✓				1721	Netherlands
[35]	✓		✓			2729	China
[36]	✓		✓	✓		146	Mayotte
[38]	✓	✓				1468	Finland
[39]	✓					300	Cyprus
[37]	✓			✓	✓	164	Germany
Proposed	✓	✓	✓	✓	✓	284	Greece, Romania, Spain
Survey							Italy, Austria, Denmark

), with studies such as those by Liu et al. [32] and Zanocco et al. [33] examining aspects ranging from basic energy behaviors to responsiveness to DR programs under varying pricing scenarios. These studies underscore the disparities in energy literacy and the ensuing challenges in aligning DR program incentives with actual user behaviors and preferences. Further investigations reveal varied levels of engagement and awareness among participants, influenced by demographic and regional features [34,35,36]. For instance, Brounen et al. [34] identified a significant correlation between energy awareness and environmental ideologies among Dutch households, while Wang et al. [35] demonstrated the impact of household composition and power-saving habits on DR responsiveness in China. Moreover, the integration of multi-method approaches coupled with specific questionnaire categories suggested in the related literature [37,38,39] points to a nuanced approach to understanding and stimulating consumer participation in DR programs.

Research indicates that the adoption of sustainable practices, such as DR programs, is strongly driven by a combination of environmental awareness and personal values [40]. Studies highlight that intrinsic factors, such as environmental knowledge, jointly with external factors, such as social norms and financial limitations, shape pro-environmental behaviors [41]. Environmental beliefs often influence energy-saving behaviors profoundly, revealing that individuals with firm environmental convictions are more likely to participate in energy initiatives, including DR programs [42]. In conjunction, positive environmental attitudes propel the adoption of energy-saving practices by low-income households [43]. These behavioral regularities map onto knowledge and intention constructs, such as energy literacy and DR familiarity, that are consistently implicated in technology adoption and pro-environmental action while remaining measurable and actionable for program design.

At the same time, cost savings represent a complex priority for consumers considering energy-efficient behaviors. Notably, the impact of energy-saving advertisements emphasizing financial benefits fluctuates with political orientations, with individuals holding pro-environmental views reacting more favorably [44]. The significant role of economic considerations is further evidenced by the uplift in energy conservation spurred by incentives beyond direct pricing, such as financial savings [45]. Cost reduction has been shown to frequently boost participation in DR programs, while financial constraints often determine the adoption of energy-saving behaviors [46,47]. This interplay between financial and environmental priorities reflects the complexity of public attitudes toward energy conservation and environmental responsibility. Accordingly, the developed questionnaire also captures financial, environmental, technological, and social drivers, enabling us to link segment profiles to differentiated outreach strategies and incentive designs.

In parallel, advancements in smart technologies form a vital part of the energy transition and DR that is inclusive of regular homeowners [48,49]. Public skepticism about the usability, reliability, and privacy of such systems is often linked to broader concerns about their credibility in promoting substantial energy savings and may thus hinder their adoption [50]. Therefore, understanding the interests of residents investing in these technologies is crucial to increasing DR program uptake. Social norms and peer influence have also been shown to significantly affect individual energy-saving practices, prompting individuals to align their energy conservation practices with those prevalent in their community [51]. Scholars argue that community trust and cohesion are critical in driving participation in sustainability efforts [52]. Collective attitudes are an important factor, as individuals are often motivated by the behaviors and attitudes of their social groups, such that social groups’ attitudes and behaviors can either encourage or discourage energy-saving actions [53]. Hence, understanding social connections can amplify the effectiveness of energy conservation strategies such as DR programs.

1.2. Contribution and Positioning

Drawing on the aforementioned studies, the present research addresses the lack of comprehensive, cross-country analyses that integrate socio-demographic, behavioral, and attitudinal data to inform targeted DR strategies in Europe. Existing research often focuses on isolated factors—such as energy literacy, demographic characteristics, or specific drivers—without fully capturing the complex interplay of these dimensions across diverse residential contexts. This study introduces a meticulously designed questionnaire that captures a broad range of participants across varying geographical and residential settings. The sampling strategy did not aim to obtain large, statistically representative datasets within a single context but rather to assemble insights from contrasting residential user groups across Europe. Accordingly, the dataset includes 20 responses from social housing tenants in Denmark, 17 from a cohousing community for seniors in Italy, 69 from residential customers of an energy provider in Greece, 17 from inhabitants of a social housing project in a Spanish district, 28 from occupants of an Austrian multi-unit residential complex, and 112 from a Romanian university campus. Although the number of participants varies across sites, the diversity of contexts reflects the heterogeneity of European residential settings and provides a valuable basis for cross-comparison. Following that, this wide demographic coverage facilitates sound segmentation on users based on their willingness to participate in DR programs, ultimately providing nuanced insights for potential DR engagement strategies. The thoughtful design and comprehensive analysis of the questionnaire responses shed light on residential user profiles, with the aim of equipping decision-makers throughout the value chain to enhance DR acceptance and foster a user-centered energy transition in the residential sector.

This study advances the literature on residential DR and consumer behavior in four ways. First, it provides a cross-country segmentation across six diverse European residential contexts using a single, harmonized survey instrument, enabling like-for-like comparisons. Second, it integrates persona-based design with DR outreach implications, translating empirical segments into actionable user personas and program-facing recommendations. Third, it quantifies the literacy–willingness linkage across contexts, showing that higher energy literacy is consistently associated with greater willingness to participate in DR (see Section 3). Fourth, it demonstrates a pragmatic, theory-guided segmentation framework suited to low-data pilot settings in which smart-meter clustering is not feasible, yielding interpretable and operationally useful segments for utilities and aggregators. Together, these contributions complement data-driven clustering approaches and provide a generalizable, practitioner-oriented pathway from survey evidence to targeted DR engagement strategies.

2. Methods

2.1. Survey Design and Implementation

The development of the questionnaire was rooted in an extensive review of the pertinent literature on similar survey-based methodologies examining attitudes and decision-making in DR participation. The survey was meticulously structured to investigate six principal dimensions: socio-demographic characteristics, energy literacy, familiarity with DR mechanisms, willingness to engage in DR activities, primary motivators for participation, and obstacles that hinder such engagement. Participants were recruited either in person or through online means by pilot-site partners via voluntary invitation. The collected samples reflect local contexts rather than probability samples.

In more detail, the questionnaire captured comprehensive socio-demographic information about the respondents in the following key dimensions: gender, age, educational attainment, employment status, household member composition, household income, proficiency with digital technology, and expenditure patterns. To assess energy literacy, participants were asked to use a five-point Likert scale to rate their knowledge with respect to the most energy-intensive appliances, the energy sources powering their homes, available energy efficiency enhancements, and typical energy expenditures per month, as well as the concept of peak load. Regarding DR aspects, participants were offered a five-point Likert scale to evaluate their familiarity with DR, as well as their willingness to engage in DR programs in terms of their load-shift flexibility in energy-intensive household activities, namely heating, cooling, laundry, water heating, cooking, and electric vehicle charging. The main drivers of participation were explored by assessing attitudes toward energy consumption through statements such as “I think about the environment when I consume energy,” “I care about the energy bills I pay,” “I am interested in investing in smart home technologies to increase my home’s energy efficiency,” and “I would like my friends/neighbors to know that I am actively participating in DR programs.”. These statements were designed to gauge environmental, economic, technological, and social motivations towards DR participation. Lastly, the questionnaire appraised potential barriers to DR engagement, such as insufficient awareness, inadequate financial incentives, technological limitations, and the impact of DR on daily routines. This comprehensive approach was intended to elucidate the complex interplay of factors influencing residential users’ participation in DR programs.

The reliability and internal consistency of the questionnaire have been evaluated with the assistance of Cronbach’s alpha. This metric is expressed as a number between 0 and 1, describing the extent to which all the items in a test measure the same concept, and hence, it is connected to the inter-relatedness of the items within the test [54]. Its score for the examined dataset amounts to 0.84, suggesting solid internal consistency and reliability and thus justifying the questions’ alignment. This calculation refers to the total answers across all countries since a size of about 200 participants guarantees a dependable measurement. Despite the excessively low population for the individual countries, Cronbach’s alpha was also calculated for them as an additional investigation step; albeit frail, the results produced high values for each country’s Cronbach’s alpha (over 0.8). As regards the questionnaire’s validity, a factor analysis resulted in a Kaiser–Meyer–Olkin (KMO) value of 0.84, suggesting suitability for this type of analysis, and four factors with eigenvalues greater than 1.0, collectively explaining 68.1% of the total variance. The aforementioned findings provide strong evidence for the construct validity and reliability of the questionnaire as a measurement instrument.

2.2. Segmentation Methodology

The flowchart regarding the segmentation methodology is depicted in Figure 1. The initial step was to collect and consolidate all responses per pilot by grouping them into the identified principal dimensions.

The aforementioned preparation phase led to the core segmentation procedure. The primary axis used for this segmentation was DR willingness, which was calculated based on the participants’ responses on their readiness to adjust their energy usage during peak periods for energy-intensive household activities, including space heating, space cooling, laundry, water heating, cooking, and EV charging. This metric specifically considered ratings of high willingness (4 or 5 on a 5-point Likert scale), ensuring the focus was on participants most likely to engage in DR programs actively. This choice is justified both as an explicit indication of user stance and as a threshold for numerically balanced and coherent participant groups.

After calculating the DR willingness per participant, all other metrics were examined against high and low levels of this metric, respectively. A significant connection was detected between DR willingness and energy literacy. A low overall DR willingness is linked to lower scoring in the energy literacy questions. On the flip side, a high overall DR willingness is linked to high energy literacy levels. This is an indication that participants can be split into DR willingness groups presenting internal uniformity at least in the energy literacy aspect.

A principal objective in this process is to create clear and balanced user groups: clear in terms of internal cohesion concerning behavior of the users in each group that is distinct compared to all other users and balanced in terms of the population proportionality among groups. For this purpose, groups formed up to this step that did not meet this objective were filtered further based on DR familiarity.

The selection of the segmentation dimensions was verified through the examination of the variance values across all questionnaire dimensions. Indeed, DR willingness is by far the item with the highest variance, with DR familiarity ranking second. As for the use of energy literacy as a determinant dimension, it has been proven to be a pillar of demand-side management [55]. The variance in each metric is available in Appendix A.

Once the groups were finalized, their scoring across the factors used for the segmentation was compounded with their scoring across all other responses. The same process was conducted at a pilot level. The outcome was the identification of a persona at the segment and pilot levels, respectively. A persona reflects the dominant features of the user subset it refers to and substantially delineates their make-up. Each group’s persona was compared to the overall pilot profile to assess alignment.

This dive into a nuanced breakdown of the population’s composition per pilot and per pilot segment was harnessed to produce insights. Remarkably, the blend of each persona’s features, priorities, attitude, and perceived barriers was synthesized to evaluate its prospects for joining a DR program and inferring suggestions for outreach strategies.

3. Results

This section presents the segmentation analysis of the questionnaire data collected from residential participants across the six European pilot sites. The objective is to delineate distinct groups with a view to delineating a persona per pilot group and juxtapose it with the respective pilot persona. Eventually, the persona’s constitution is utilized to evaluate its prospects of joining a DR scheme and to offer insights for outreach plans that resonate with user groups. The segmentation was primarily conducted along two key dimensions: participants’ willingness to shift their electricity load in response to DR programs (referred to as “Overall DR Willingness”) and their familiarity with DR concepts (referred to as “DR Familiarity”). These dimensions were selected due to their demonstrated ability to differentiate participant groups with internal cohesion and distinct external characteristics. Figure 2 illustrates the allocation of each pilot population into groups characterized by low or high scores along the two segmentation axes.

Table 2. A summary of participant characteristics across six European pilot sites. Categories include socio-demographic data, energy literacy and digital expertise levels, and drivers of and barriers to DR participation. Values such as “N/A” indicate missing data; “No dominant” denotes the absence of a clear majority response in the respective category; Key distinguishing features of each participant group are highlighted. Scores for energy literacy, DR familiarity, and overall DR willingness are averages based on the participants’ responses using Likert scales (1–5). The barriers reported summarize multiple-choice responses highlighting the concerns most cited by participants.

Category	Danish	Italian	Greek	Spanish	Austrian	Romanian
Number of Participants	20	17	69	17	28	112
Gender	17 female, 3 male	12 female, 5 male	16 female, 53 male	3 female, 14 male	12 female, 16 male	39 female, 73 male
Age	No dominant	65+	30–49	30+	No dominant	18–29
Education	Primary–high school	N/A	Higher+	No dominant	No dominant	Higher
Digital Expertise	Mid	Low–mid	High	Mid	Mid–high	Mid–high
Occupation	Employed	Retired	Employed	Employed	Employed	Student
Adults	0–2	0	1–2	1–2	1–2	2–4
Seniors	0–1	1–2	0	0	0	0
Children	0–3	0	0–2	0–2	0	0–1
Income	Low	Mid	High	N/A	Mid–high	Mid–high
Spending Behavior	Very conservative	Conservative	Conservative	Liberal	No dominant	Mid
Energy Literacy Average	3.6	2.4	3.6	3.4	3.7	3.1
DR Familiarity Average	3	2.8	3	2.6	3.5	2.5
Overall DR Willingness	2.0	3	2.5	2	2	2
Environmental Drivers	High	High	High	High	High	Mid
Financial Drivers	Low	Low	Very high	N/A	Mid	High
Technical Drivers	High	Low	Very high	Mid–high	High	High
Social Drivers	Mid	High	Mid–low	Mid	Low–mid	Mid
DR Barriers	- Lack of awareness	- Tech challenges	- Limited fin rewards - Lack of awareness - Daily impact	- Lack of awareness	- Tech challenges - Limited fin. rewards	- Daily impact - Lack of awareness - Limited fin. rewards

summarizes the descriptive statistics derived from the questionnaire responses. For succinctness, the values are aggregated at the pilot country level. This setup facilitates a comparison across the different pilots and primarily traces the connections of the pilot-level analysis to the subsequent pilot segment elaboration and the corresponding outreach insights. The values that represent distinctive features of each pilot are presented in purple font. Appendix B present a breakdown of this analysis at the pilot and segment levels. The elements highlighted with a blue background indicate values that distinguish a segment from its peers in the same pilot.

The Danish Pilot. The Danish pilot (20 participants) included immigrant dwellers of a social housing facility. These participants are mostly female with a primary or high school education, low income, and very conservative spending behavior. This population was segmented into two groups that were starkly distinctive in terms of the DR willingness measure. Segment A (eight participants) features participants with higher income levels compared to those of the other pilot members; exhibits no particular spending patterns; and demonstrates considerable DR willingness. Strategic outreach for this group should focus on emphasizing ecological and technological DR benefits, leveraging their quite high energy literacy and openness to energy efficiency technologies. Segment B (12 participants) encompasses exclusively female members with a very low income and conservative spending habits; shows a firmer financial stance; and is apprehensive about the impact of DR programs on their daily habits. Despite their low DR willingness, the prospects are potentially good if financial benefits are highlighted and literacy is reinforced.

The Italian Pilot. The pilot in Italy (17 participants) comprises tenants of a housing complex for elderly people with limited digital expertise, though they are very environmentally conscious. Since their energy expenses were incorporated into their fees, financial factors were not relevant to their interests. This population was discernibly divided into two segments on the basis of DR willingness. Segment A (nine participants) accumulates chiefly male participants with higher energy literacy that are tech-savvy in comparison to the other group, thus indicating more positive DR prospects. The outreach for them should capitalize on their familiarity with DR and high energy literacy, as well as the alignment of their social and technical merits to address their lack of awareness and concerns about the impact of DR on their daily lives. Conversely, Segment B (eight participants) includes female participants with minimal energy literacy and lower social drivers of DR, resulting in low DR prospects. Outreach efforts should focus on leveraging their notable interest in benefiting the environment.

The Greek Pilot. The Greek pilot (69 participants) identified four segments of residential users, mostly middle-aged with high incomes but conservative spending behavior and also substantially differing attitudes across environmental, financial, and technological dimensions. In terms of DR willingness, Segments A and B both display high DR prospects, though they differ in their DR familiarity, together with multiple other contrasts. The willing and familiar Segment A (22 participants) is dominated by middle-aged people with postgraduate degrees and liberal spending habits. Moreover, they register the strongest stance across all attitude drivers. Outreach should focus on eco-financial incentives and technological advancements outweighing the impact of DR on their daily habits. The willing but unfamiliar Segment B (19 participants) encompasses mainly male participants with lower education and conservative spending behavior. Building familiarity with the benefits of DR while stressing the financial, technological, and environmental benefits are key to a successful outreach strategy. On the other hand, Segments C and D gather the unwilling participants that stand apart in terms of their DR familiarity. Segment C (11 participants) involves unwilling and familiar individuals who are liberal spenders but have weaker attitudes, suggesting high prospects if technological barriers are addressed and environmental benefits are underscored. Finally, Segment D (17 participants) includes primarily unfamiliar and unwilling male participants, where mid-level DR prospects can be enhanced by raising DR awareness based on their financial interests and digital expertise.

The Spanish Pilot. The Spanish pilot (17 participants) relates to inhabitants of social housing. Its population has been divided into two groups based on DR familiarity. It features the familiar Segment A (11 participants) with well-educated male participants with better digital expertise than the rest of the pilot members, showing high DR prospects merely by raising DR awareness. Furthermore, their more prominent attitude in comparison to the total pilot population indicates that stressing environmental, technological, and social benefits would be advantageous. At the same time, Segment B (six participants) includes unfamiliar female participants with lower digital expertise and limited attitude scores, implying low DR prospects. The approach to this segment should focus on assisting with literacy and using their positive tech attitudes to improve their engagement.

The Austrian Pilot. The Austrian pilot (28 participants) is composed of two segments of willing individuals split by familiarity, in addition to a third segment gathering all unwilling responders. The willing and familiar Segment A (10 participants) promises high DR prospects, consisting mainly of well-educated male participants with liberal spending habits and pronounced interest across all attitude dimensions. The outreach should focus on mitigating technological challenges and highlighting the entire spectrum of DR benefits—environmental, financial, technological. The willing and unfamiliar Segment B (eight participants) encompasses female participants with a secondary education and mid–high incomes but conservative spending habits, showing positive DR prospects. Outreach should emphasize environmental and technological benefits. At the same time, the unwilling Segment C (10 participants) is marked by low DR prospects, consisting of participants with the weakest attitudes across the pilot population and stating a lack of awareness as their main impediment. However, the significant diversity observed among this segment’s members does not allow for targeted strategy insights.

The Romanian Pilot. Four segments were created for the tenants of a university residency hall (112 participants), who registered higher income and digital expertise, mid-level spending behavior, and high scores in their financial and technological attitudes. The willing and familiar Segment A (30 participants) includes men who spend liberally with higher digital expertise and incomes, showing significant DR prospects. The suggested strategy is to boost awareness and mitigate concerns about DR’s impact on their daily routines, tapping into their strong attitudes, especially towards financial and technological benefits. The willing and unfamiliar Segment B (31 participants), potentially with high DR prospects, involves many women who match the general pilot characteristics and requires raising awareness about financial, technological, and foremost social aspects. Finally, unwilling Segments C (23 participants) and D (28 participants), with familiar and unfamiliar profiles, respectively, suggest tailored outreach to leverage their attitudes and expertise for better engagement in DR programs. Awareness must be raised in Segment C, especially regarding environmental factors. Segment D needs to be presented with the financial benefits while employing their higher digital savviness.

4. Discussion

The goal of the segmentation process is to identify a persona for each pilot and then for each pilot segment based on the analysis of the questionnaire responses. A persona serves to provide technology developers with a hypothetical user based on data from the target population. Delineating a typical user allows technology developers to create solutions tailored to actual user preferences, skills, and motivations, as opposed to relying on personal judgment and potential bias. The profiles extracted from this analysis are defined across dimensions that form an integral part of the segmentation analysis.

This cross-European survey has shed light on the diverse factors influencing residential engagement in DR programs. By examining energy consumption behaviors and flexibility in usage patterns for energy-intensive household activities and identifying both barriers and enablers, this study highlights the complex interplay of factors shaping DR readiness among European households. The results underscore the importance of considering regional, socio-demographic, and attitudinal variations when designing DR strategies. Notably, the segmentation revealed substantial heterogeneity in DR willingness and familiarity across the six pilot sites. These differences emphasize the need for custom-built DR initiatives that are sensitive to the distinct profiles and preferences of various consumer groups. Tailored approaches are essential to elevate DR programs’ effectiveness, as they ensure that engagement strategies resonate with users’ specific attributes and motivations. The customer segmentation analysis performed as part of this study played a pivotal role in providing actionable insights. The DR willingness and DR familiarity dimensions utilized to generate the segments offered a fundamental distinction within the population. The ensuing exploration of the residual factors added substantial pieces to the persona puzzle, spurring the translation of its features into actionable insights.

This nuanced understanding of segment responsiveness is pivotal for providing insights for policymakers, energy stakeholders, and technology developers. Personalized communication strategies that address the specific motivations and barriers for each consumer segment are essential. For instance, segments demonstrating significant adoption potential may benefit from advanced DR programs that capitalize on their already favorable attitude towards energy flexibility. Conversely, segments with lower energy literacy might require diligently targeted educational campaigns that simplify DR concepts and highlight tangible benefits or enhanced incentives to improve their engagement level. Financial incentives remain a powerful motivator but should be complemented by clear communication about non-monetary benefits, such as environmental impact and social recognition.

From a utility or aggregator perspective, the segments identified here imply distinct communication levers. For high-willingness–high-familiarity users, concise, event-specific messages that emphasize automation options, the expected event duration, opt-out/override controls, and immediate performance feedback (e.g., bill/app summaries) are likely to sustain engagement. For high-willingness–low-familiarity users, plain-language onboarding with visual examples of the expected savings in EUR and CO₂ avoided, together with bill protection trials, can reduce perceived risks and accelerate adoption. Low-willingness–high-familiarity users may respond to framing that highlights environmental/community benefits and low-friction participation (short events, flexible overrides). Low-willingness–low-familiarity users could benefit from awareness campaigns via trusted local channels (bill inserts, SMS/app nudges, community partners), explicit data-privacy assurances, and simple calls-to-action.

The present findings should be interpreted in light of several sampling constraints. The current study used samples recruited through pilot partners. In this regard, the data are not probability-based but reflect site-specific contexts (e.g., customer mix, local practices). In addition, the sizes varied markedly across pilots (e.g., Italy and Spain n = 17 vs. Romania n = 112), which limits the statistical power and the stability of the segment estimates for small-n sites. Third, because several pilots are small, the results for those sites are exploratory and hypothesis-generating rather than definitive; cross-pilot contrasts should be read as contextual patterns rather than population-level differences. Fourth, the measures are self-reported Likert items, which may introduce common-method bias. Consequently, the segmentation and personas are intended for analytic generalization to similar contexts and for program design insights, not for estimating prevalence or making causal claims. Future studies should employ stratified/probability sampling with balanced sample sizes and validate the segments against behavioral outcomes (e.g., smart-meter responses during events).

In addition, given the observed regional and demographic variations, policy frameworks should be flexible and adaptive, allowing for localized DR programs that resonate with specific community needs. Integrating insights from behavioral and social sciences into DR program design stands to elevate their efficacy.

5. Conclusions

This study delivered a harmonized, cross-country segmentation of residential consumers across six European pilots, showing that DR readiness is shaped by the interplay of willingness, familiarity, energy literacy, and attitudinal drivers, alongside perceived barriers. The resulting personas translate empirical patterns into operational guidance for utilities, aggregators, and policymakers seeking to tailor DR enrolment, messaging, and incentives. At the same time, these findings are bounded by the limitations of the data’s representativeness and generalizability, which are addressed by sampling via pilot partners. Even though variation in the sample sizes across countries and small pilots restrict inferential certainty and preclude prevalence estimation, the reported differences should be best viewed as contextual signals to inform program design rather than as statistically generalizable conclusions.

Future work should adopt probability-based or stratified sampling to enhance the representativeness and balance across pilots, increase the sample sizes at smaller sites to stabilize the segment estimates, and validate the proposed segments against observed behavioral outcomes—such as participation during DR events measured using smart-meter data. As larger and more balanced samples become available, researchers should evaluate alternative modeling approaches, including latent class analysis and multilevel models, to assess the robustness and potential hierarchical structure of the segments. In parallel, segment-specific communication and incentive strategies ought to be tested through rigorous field experimentation to identify the most effective levers for scalable DR deployment. Collectively, these steps will strengthen external validity and sharpen the translation from segmentation insights to practice.