Sustainable Web-Design and Digital Marketing Potentials

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

It is my honor to read this paper. Your study makes a timely and useful contribution by demonstrating that technically sustainable choices, especially faster load times, and transparent communication (seals, green hosting disclosure, and an information page) measurably improve brand perception, perceived quality, and purchase intention in a DCE.

To tighten the paper for publication, please document your literature search protocol and slightly temper the novelty claim; anchor each DCE attribute in the cited literature with a compact mapping table; and clarify the experimental design (attribute naming, base levels, fast/slow anchors, and the exact choice‑task structure).

Regarding the results, please reconcile the Design coefficient interpretation with Table 2, clearly restate the coding, and include marginal effects or predicted probabilities alongside a standard fit metric. Given the shared respondents across outcomes, specify a panel setup or cluster the standard errors. Finally, adjust the introductory CO₂ extrapolation (or remove it) and pass the manuscript through for minor language and consistency fixes, including the optimization algorithm used in Apollo.

With these targeted edits, the paper will communicate its contributions cleanly and be ready for publication.

Good luck!

Author Response

It is my honor to read this paper. Your study makes a timely and useful contribution by demonstrating that technically sustainable choices, especially faster load times, and transparent communication (seals, green hosting disclosure, and an information page) measurably improve brand perception, perceived quality, and purchase intention in a DCE.

 

Comment 1

To tighten the paper for publication, please document your literature search protocol and slightly temper the novelty claim; anchor each DCE attribute in the cited literature with a compact mapping table; and clarify the experimental design (attribute naming, base levels, fast/slow anchors, and the exact choice‑task structure).

Response 1

While the literature search is already addressed in the introduction, following another reviewer’s advice the current literature review is expanded. In this context the search strategy is expanded as well.

The adjusted literature review has an impact on the novelty claim of the article which consequently has been adjusted as well.

While the literature on DCEs in web design remains slim the attributes and levels have been related to other studies on the topic at least.

Table 1 already names all attributes and levels and the preceding text defines the base levels. Without any time constraints to participants’ decisions, all choices cards can be interpreted as slow anchors. This aspect is mentioned in the text. The choice structure is illustrated by inclusion of an exemplary choice card into the article.

 

Comment 2:

Regarding the results, please reconcile the Design coefficient interpretation with Table 2, clearly restate the coding, and include marginal effects or predicted probabilities alongside a standard fit metric. Given the shared respondents across outcomes, specify a panel setup or cluster the standard errors.

Response 2:

Thank you for pointing out the inconsistency. It has been fixed in the text.

In the addition to the reported coefficients, based on the predicted probabilities shares are reported as well. As a fit metric McFadden’s rho² is reported. Even though not fully comparable to the R² measure it is the best choice for the evaluation of DCE outcomes.

The Apollo package in R automatically, realizes a panel setup, so that a special treatment of the standard errors is not required.

 

Comment 3:

Finally, adjust the introductory CO₂ extrapolation (or remove it) and pass the manuscript through for minor language and consistency fixes, including the optimization algorithm used in Apollo.

Response 3:

The CO₂ extrapolation has been put into context and backed up by another source.

The whole article has been language checked again.

The final part of the comment is not understandable, how is the optimization algorithm used in the Apollo package to be checked? Does it refer to the referenced source, which is correct or the application of the Apollo package?

 

With these targeted edits, the paper will communicate its contributions cleanly and be ready for publication.

Good luck!

Reviewer 2 Report

Comments and Suggestions for Authors

Thank you for providing me with the opportunity to review this paper. This manuscript addresses a timely and relevant topic at the intersection of digital sustainability, UX design and marketing communication. A mixed-methods approach combining expert interviews with a discrete choice experiment (DCE) is methodologically appropriate. However, significant limitations in sample size, sampling bias, and methodological transparency undermine the robustness of these conclusions. The theoretical contribution is modest, and the writing requires substantial refinement.

1. The current introduction is overly redundant and fails to highlight the importance, innovation, and contribution of the research. In the introductory section, the authors need to establish the real problem in society that brings us to this point of research. Once it is established as a real problem in society that needs to be addressed, the authors need to highlight why this is important and what others have done to address this problem. Then, they need to discuss the research gap and what is unknown. Then, establish your motivation for addressing this gap. In the introduction, there is also a need to write a few lines about the findings and a summary of the contributions.
2. The literature review is insufficient, with fragmented and superficial coverage of existing research. It fails to adequately synthesize and critique current theoretical and empirical studies on sustainable web design, thereby undermining the demonstration of the unique positioning and breakthroughs of this study within the academic field. This limitation ultimately compromises the scholarly value of this study.
3. The absence of a clear theoretical framework is another limitation of this study. Although the article touches on sustainable web design, it fails to establish systematic theoretical guidance or draw substantially from existing theories in environmental psychology, user experience, or sustainable development. This results in insufficient theoretical depth and a lack of academic persuasiveness. Consequently, the research questions and hypotheses remain relatively superficial, lacking demonstrated scholarly innovation and rigor.
4. The connection between theory and empiricism is stillunderdeveloped. Although the study combines interviews and choice experiments, it fails to demonstrate how these two methods complement and validate each other. It is recommended that the methodology section explicitly states how the qualitative findings informed the design of the quantitative variables and how the quantitative results, in turn, contributed to theory building.
5. The criteria for expert selection are clearly defined; however, the design of the interview questions, coding analysis process, and measures taken to ensure reliability and validity are not elaborated in detail. It is recommended to supplement the methodology with aninterview protocol, coding framework, and mechanisms such as double codingor cross-verification to enhance the scientific rigor and replicability of qualitative analysis.
6. The issue of sample representativeness is significantin this study. The authors acknowledge that the sample lacks representativeness and is skewed toward individuals with higher environmental awareness, potentially leading to overly optimistic results. Enhancing sample diversity by including participants from varied cultural backgrounds, age groups, and internet usage habits, or conducting stratified analyses to improve external validity is recommended.
7. The design of the choice experiment requires detailed elaborationin this section. Althoughthe paper mentions the use of a discrete choice experiment, it fails to specify the attribute design, experimental protocol, and randomization procedures. It is recommended that specific design details be provided to enhance methodological transparency.
8. The academic contribution and practical value of thisstudy are insufficient. The paper needs to clearly articulate its theoretical contribution, whether it enriches sustainable design theory, expands user behavior research, or proposes a new measurement index system. It is recommended that the theoretical innovations be systematically summarized in the conclusion section. Furthermore, the practical implications are underdeveloped, with limited concrete guidance provided toweb designers and digital marketers based on the research findings.
9. The Discussion section should be contextualized against the existing literature, explicitly highlighting both similarities and differences from prior studies. Crucially, it must underscore the novel contributions of this study.
10. The formatting of the references is inconsistent, with some entries incomplete and requiring further standardization. Additionally, the current body of literature cited is insufficient. The overall quality of the references also needs to be enhanced.
11. The linguistic expression requires further refinement. Certain passages in the paper exhibit a somewhat mechanical style, and it is recommended to enhance the fluency and logical coherence of the academic language to improve the overall readability. Several sentences are excessively long and would benefit from being broken downinto shorter sentences. Additionally, the text contains substantial repetitive content, monotonous sentence structures, and imprecise use of technical terminology, which collectively undermine the professionalism and academic impact of the paper.
12. I suggest that the author strictly refer to the journal's relevant writing guidelines to format the paper before submission.

Author Response

Thank you for providing me with the opportunity to review this paper. This manuscript addresses a timely and relevant topic at the intersection of digital sustainability, UX design and marketing communication. A mixed-methods approach combining expert interviews with a discrete choice experiment (DCE) is methodologically appropriate. However, significant limitations in sample size, sampling bias, and methodological transparency undermine the robustness of these conclusions. The theoretical contribution is modest, and the writing requires substantial refinement.

 

Comment 1:

The current introduction is overly redundant and fails to highlight the importance, innovation, and contribution of the research. In the introductory section, the authors need to establish the real problem in society that brings us to this point of research. Once it is established as a real problem in society that needs to be addressed, the authors need to highlight why this is important and what others have done to address this problem. Then, they need to discuss the research gap and what is unknown. Then, establish your motivation for addressing this gap. In the introduction, there is also a need to write a few lines about the findings and a summary of the contributions.

Response 1:

The first part of this comment cannot be followed. The introduction already illustrates the underlying problem and formulates the corresponding need for action. This part has been reworked to make this point clearer to the reader.

The discussion of potential solutions has been expanded. The research gap has already been stated and deduced from literature. By expanding upon the literature research strategy, this foundation has been expanded upon.

 

Comment 2:

The literature review is insufficient, with fragmented and superficial coverage of existing research. It fails to adequately synthesize and critique current theoretical and empirical studies on sustainable web design, thereby undermining the demonstration of the unique positioning and breakthroughs of this study within the academic field. This limitation ultimately compromises the scholarly value of this study.

Response 2:

The literature search strategy is already detailed and while it has been expanded the scope of article directly relatable remains limited in regard to studies with the same methodological approach. This has been detailed in the expanded introduction. The background on sustainable web design in general has been expanded drawing more strongly on the existing literature.

 

Comment 3:

The absence of a clear theoretical framework is another limitation of this study. Although the article touches on sustainable web design, it fails to establish systematic theoretical guidance or draw substantially from existing theories in environmental psychology, user experience, or sustainable development. This results in insufficient theoretical depth and a lack of academic persuasiveness. Consequently, the research questions and hypotheses remain relatively superficial, lacking demonstrated scholarly innovation and rigor.

Response 3:

The attributes and levels of the DCE are more strongly anchored in the existing literature. Additionally, the theoretical foundation underlying the use of DCEs has been stressed more strongly.

 

Comment 4:

The connection between theory and empiricism is stillunderdeveloped. Although the study combines interviews and choice experiments, it fails to demonstrate how these two methods complement and validate each other. It is recommended that the methodology section explicitly states how the qualitative findings informed the design of the quantitative variables and how the quantitative results, in turn, contributed to theory building.

Response 4:

The methodology section, in particular, the part of the overall research design has been reworked to illustrate more clearly the link between the theoretical foundation, and the two parts of the empirical study.

 

Comment 5:

The criteria for expert selection are clearly defined; however, the design of the interview questions, coding analysis process, and measures taken to ensure reliability and validity are not elaborated in detail. It is recommended to supplement the methodology with aninterview protocol, coding framework, and mechanisms such as double codingor cross-verification to enhance the scientific rigor and replicability of qualitative analysis.

Response 5:

Additional information about the process of analysing the interviews has been added to the article, including a more detailed deduction of the interview questions as well as a discussion of the process to assure objectivity of the analysis. If “interview protocol” refers to the transcripts, they have been made available to the journal editor but due to confidentiality agreements are not to be fully included in the article.

 

Comment 6:

The issue of sample representativeness is significantin this study. The authors acknowledge that the sample lacks representativeness and is skewed toward individuals with higher environmental awareness, potentially leading to overly optimistic results. Enhancing sample diversity by including participants from varied cultural backgrounds, age groups, and internet usage habits, or conducting stratified analyses to improve external validity is recommended.

Response 6:

This issue is addressed by controlling for these critical issues in the context of model I’, II’ and III’. This is stressed more strongly in the text. Additionally, the recommendations are added to the section on limitations and outlook.

 

Comment 7:

The design of the choice experiment requires detailed elaborationin this section. Althoughthe paper mentions the use of a discrete choice experiment, it fails to specify the attribute design, experimental protocol, and randomization procedures. It is recommended that specific design details be provided to enhance methodological transparency.

Response 7:

While Table 1 and the preceding text illustrate the design of attributes and levels, the later text addresses the realization of the experiment as such, including the randomization. To make the process clearer to the reader, an exemplary choice card is included in the text.

 

Comment 8:

The academic contribution and practical value of thisstudy are insufficient. The paper needs to clearly articulate its theoretical contribution, whether it enriches sustainable design theory, expands user behavior research, or proposes a new measurement index system. It is recommended that the theoretical innovations be systematically summarized in the conclusion section. Furthermore, the practical implications are underdeveloped, with limited concrete guidance provided toweb designers and digital marketers based on the research findings.

Response 8:

The introduction already states the contributions of the study. It in particular states that the article targets practical issues of web design and consumer perceptions, i.e., preferences, and does not aim to expand design theory.

We take this comment, however, as a motivation to add an additional section to the conclusions detailing the theoretical contributions of the article.

Additionally, the practical recommendations are revised.

 

Comment 9:

The Discussion section should be contextualized against the existing literature, explicitly highlighting both similarities and differences from prior studies. Crucially, it must underscore the novel contributions of this study.

Response 9:

The discussion section has been more strongly linked to the literature referenced before and references used to motivate the implemented attributes and levels.

 

Comment 10:

The formatting of the references is inconsistent, with some entries incomplete and requiring further standardization. Additionally, the current body of literature cited is insufficient. The overall quality of the references also needs to be enhanced.

Response 10:

The formatting of references follows the standard of the journal and reflects all information available for the sources.

The body of literature has been expanded as far as possible while still addressing the underlying core research objective.

 

Comment 11:

The linguistic expression requires further refinement. Certain passages in the paper exhibit a somewhat mechanical style, and it is recommended to enhance the fluency and logical coherence of the academic language to improve the overall readability. Several sentences are excessively long and would benefit from being broken downinto shorter sentences. Additionally, the text contains substantial repetitive content, monotonous sentence structures, and imprecise use of technical terminology, which collectively undermine the professionalism and academic impact of the paper.

Response 11:

The text has been reworked and in part to make it more accessible to readers.

 

Comment 12:

I suggest that the author strictly refer to the journal's relevant writing guidelines to format the paper before submission.

Response 12:

All requirements regarding citations, writing style and naming of sections have been observed, so this comment cannot be followed.

Reviewer 3 Report

Comments and Suggestions for Authors

Following a thorough analysis of the manuscript titled “Sustainable Web Design and Digital Marketing Potentials”, several critical shortcomings have been identified that raise serious concerns regarding its scientific validity and suitability for publication. These include methodological weaknesses, a lack of scientific rigor, limited originality, and unsupported claims—each of which compromises the article’s academic integrity. A detailed evaluation of the manuscript is provided in the attached file.

Comments for author File: Comments.pdf

Author Response

Comment 1:

Scientific review of the article:

"Sustainable Web-Design and Digital Marketing Potentialsu by Perret et al. (2025).

The conceptual framework remains loosely structured. The authors summarize relevant prior studies, but no explicit theoretical model is formulated linking sustainability attributes (seals, hosting, loading speed, etc.) to behavioral outcomes via measurable mediators such as trust, attitude toward the website, or perceived authenticity. Incorporating models like the Theory of Planned Behavior (Ajzen, 1991 [DOI: 10.1016/0749-5978(91)90020-T]) or the Elaboration Likelihood Model could have strengthened the causal logic.

Furthermore, construct operationalization is largely implicit: variables such as "positive company perception" or "purchase likelihood" are not validated scales but single self-report items, reducing psychometric reliability.

Response 1:

While we agree that the theoretical link is rather tenuous and have worked on it accordingly by placing the study within the context of rational choice and random utility theory both proposed the theory of planned behaviour or the elaboration likelihood theory are unsuitable in the context of DCEs. Where DCEs aim to elicit latent preference structures both proposed theoretical approach aim at eliciting factors that may lead to a behaviour changes.

 

Comment 2:

Methodological Evaluation

Qualitative Phase

The expert interviews are well-structured, yet the sample size (n = 4) is too small to achieve thematic saturation (Guest et al., 2020). All experts are male, producing gender bias. The data analysis lacks methodological transparency-no coding framework, inter-coder reliability, or qualitative analysis software is specified.

Response 2:

The analysis of the expert interviews has been expanded to reflect on most the aspects mentioned in the comment. Additionally, the research design has been expanded to detail the link between the two empirical parts of the study and that the main objective of the qualitative part lies in the provision of an additional validation of the quantitative part of the study.

 

Comment 3:

Quantitative Phase

The discrete choice experiment (DCE) is appropriately chosen to elicit preference structures (Ryan et al., 2007), and the orthogonal fractional factorial design is sound. Yet the experimental realism is limited: stimuli are presented as textual descriptions, not functional website prototypes. Consequently, ecological validity is questionable.

Response 3:

We agree with this assessment, which is why it has been addressed in the context of the limitations.

 

Comment 4:

The sample size (n = 107) is insufficient for stable mixed logit estimation with five attributes. Power analysis is not reported. The demographic structure (71 % with academic degrees, median age� 29 years) introduces sampling bias toward digitally literate, sustainability-oriented respondents, acknowledged by the authors but still critical.

Response 4:

We agree that the sample is rather small, as acknowledged, however, the sampling error still lies below 10%. This argument has been added to the text. While mixed logit requires a larger sample size, in the case of the analysis its main purpose is as a robustness check for the MNL estimation.

The potential bias in the sample is in part accounted for by controlling for these characteristics in models I’, II’ and III’.

 

Comment 5:

The statistical approach-multinomial and mixed logit using the Apollo package (Hess & Palma, 2019)- is technically correct. However, the interpretation occasionally overstates significance: small coefficients (e.g., � � 0.13) are treated as substantively meaningful without marginal effects or predicted probabilities.

Response 5:

Predicted probabilities or rather the respective ratio has been added to the analysis and is discussed in the article as well.

 

Comment 6:

Literature Integration

The review is broad but not systematically derived from a structured literature search (e.g., PRISMA). Many references are grey sources (blogs, websites), which weakens scientific grounding. Empirical sustainability studies from HCI (Human-Computer Interaction) and digital ecology are underrepresented (see: Preist et al., 2016; Knowles et al., 2018). Including such work would contextualize design-energy relationships beyond marketing perspectives.

Response 6:

Following as well the recommendations by another reviewer, the literature review has been expanded to focus more on sustainable digital services and infrastructure in general, including the two sources provided.

 

Comment 7:

Discussion and Interpretation

The discussion accurately synthesizes the quantitative results but lacks critical depth. For instance, the finding that "minimalistic design exerts no significant influence" could be interpreted through visual cognition or aesthetics theories, but this is not explored.

Response 7:

The discussion section has been reworked to engage more strongly with the expanded theoretical foundation and related articles and theories.

 

Comment 8:

The authors equate "fast loading time" with "sustainable design," yet speed improvements may also stem from non-sustainable optimization (e.g., CON caching using fossil-powered servers). Moreover, causality is implicitly assumed: the results demonstrate association, not behavioral change. Experimental manipulation with real websites or field A/B testing would be required to infer causal effects on purchase behavior.

Response 8:

Since the text might be easily misunderstood it has been adjusted. Fast loading times are no equal to sustainable web design but a consequence thereof. As stated above, analysing behaviour change is not the intention of the study. Thus, the text has been adjusted to make it clearer to the reader the main focus lies on eliciting consumer preference patterns.

 

Comment 9:

Ethical and Reproducibility Standards

The study reports ethics approval (Code K-2025-JP-14) and compliance with the Declaration of Helsinki, which meets formal standards. However, data availability is limited to "on request." For reproducibility, the dataset and analysis scripts (R code for Apollo) should be deposited in an open repository (e.g., Zenodo, OSF) following FAIR principles (Wilkinson et al., 2016).

Response 9:

The journal offers the option of on demand provision of the implemented data, which consequently has been used. It also does not become clear in how far an on demand provision of data does in any way limit reproducibility.

 

Comment 10:

Insufficient Scientific Contribution and Novelty

The authors claim their work is the "first investigation that adopts a consumer-centric perspective" on sustainable web design (p. 4).

This assertion is neither rigorously supported by a comprehensive literature review nor justified empirically.

Response 10:

Considering the expansion of the literature review that novelty claim has been adjusted.

 

Comment 11:

The authors cite a single relevant peer-reviewed study (Mat Nawi et al. [17]) as prior work on the topic. However, this claim of originality is weak given that the literature base on sustainable UX, energy-efficient web design, and green IT is far more extensive. E.g.:

Calero et al. (2014) addressed energy-efficient software engineering, including websites;

Ardito et al. (2021) analyzed sustainability in UX design and software interfaces.

Response 11:

As stated above the literature review has been reworked to consider a broader range of relevant keywords in the search.

 

Comment 12:

Severe Methodological Weaknesses in Sampling

The quantitative sample (n = 107) is not representative, as explicitly acknowledged by the authors themselves (p. 12). Furthermore, participants were not randomly selected, and self-selection bias is likely due to the online format and topic sensitivity. The data is not generalizable, invalidating any population-level inferences. This contradicts the authors' claims about digital marketing strategy applicability and undermines external validity.

Response 12:

While a self-selection bias does exist, it assure at the same time that the participants have a suitable understanding of the topic to assure a higher validity of their answers. In how far the results are representative, i.e., for which sub-group of the population has been addressed and by controlling for multiple characteristics the biases inherent in the sample are at least in part alleviated.

 

Comment 13:

Overreliance on Subjective Measures

The core metrics-perception of the company, product quality, and purchase likelihood-are entirely self-reported and hypothetical. These are collected in a controlled artificial setting via a discrete choice experiment (DCE), lacking ecological validity.

No behavioral data (e.g., clicks, bounce rates, actual purchases).

No pretest or validation of the DCE design.

The DCE attribute levels (e.g., "minimalist" vs "sophisticated" design) are vague and unstandardized, raising questions about measurement consistency and reproducibility.

Response 13:

Even though, DCEs in most cases are working with artificial settings, literature has shown that results are well suited to elicit preference patterns.

A pretest has been carried out and information about it has been added to the article.

 

Comment 14:

Misapplication of Advanced Quantitative Techniques

The authors implement multinomial logit and mixed logit models, including interaction terms with tertiary covariates (age, gender, sustainability type). However:

There is no statistical justification or reporting of model diagnostics (e.g., McFadden's R2, goodness-of-fit, residual analysis).

Reported results (Table 2) lack effect size interpretation,

confidence intervals, and clear causal reasoning.

The coherence between models (I, II, III) is claimed using Cramer's V, but this test is insufficient for validating cross-model robustness.

Hence, the use of econometric tools is superficial and technically flawed, serving more as a decorative add-on than a reliable inferential base.

Response 14:

McFadden’s rho² has been added as well as predicted probabilities. There statistics are discussed as well.

Cramer’s V and a Chi Squared test are solely used to test for potential biases resulting from the fact that the three choices determining the three models are surveyed at the same time.

 

Comment 15:

Overinterpretation of Results and Unsupported Claims

The paper draws strong conclusions about the marketing potential of sustainable web design (e.g., enhanced brand perception, increased conversion rates) without:

Triangulating findings with real-world A/B testing or behavioral tracking.

Supporting generalizations with statistical power analysis or effect size magnitudes.

Controlling for confounding factors (e.g., prior brand attitudes, website familiarity, pricing).

Statements such as "the most significant impact on user behavior is not directly evident in the practices themselves" (p. 8)    or "communication measures significantly increase purchase decisions" (p. 11) are speculative and not causally proven.

Response 15:

The respective parts are reformulated to make it clearer to the reader that they do not report on causal consequences even though discrete choice experiments are a more sophisticated form of A/B testing.

 

Comment 16:

Ethical and Scientific Transparency Issues

The data is not publicly available, with vague mention of it being accessible upon "reasonable request" (p. 13), violating open science principles.

The method of expert recruitment is anecdotal, with only four male experts selected via personal networks, resulting in gender and disciplinary bias.

There is no evidence of peer validation of the coding used in expert interview analysis.

Response 16:

These aspects are already addressed above.

 

Comment 17:

No Theoretical Framework

The entire study lacks anchoring in an established theoretical model. There is no mention of:

TAM (Technology Acceptance Model)

Theory of Planned Behavior

UTAUT (Unified Theory of Acceptance and Use of Technology)

Value-Belief-Norm theory for environmental behavior

As a result, hypotheses are not derived deductively, and the study lacks theoretical robustness.

Response 17:

All of these models are targeting different outcomes than the implemented DCE. However, as stated above two fitting theoretical models are used to provide a theoretical foundation for the study.

Reviewer 4 Report

Comments and Suggestions for Authors

1. Abstract and Keywords

   • The abstract is informative but overly descriptive. Please condense and focus more on the core objectives, methodology, key findings, and practical implications

   • The keywords should be refined to include more specific terms such as “consumer perception,” “digital sustainability communication,” and “discrete choice modeling.”

2. Introduction

   • The introduction provides a strong background but lacks a clear research gap and theoretical framing. Please articulate the novelty of this study more explicitly by highlighting what distinguishes it from prior work in sustainable digital design and marketing.

   • Strengthen the link between web design sustainability and its implications for digital marketing and consumer behavior.

3. Literature Review

   • The literature discussion reads as an extended background rather than a critical review. Reorganize it to present thematic sub-sections (e.g., sustainable UX design, green hosting and consumer trust, digital greenwashing).

   • Provide a concise conceptual framework or diagram summarizing the theoretical linkages among sustainability practices, user perception, and purchase behavior.

   • Several sources are descriptive; include more peer-reviewed academic literature instead of web sources where possible.

4. Research Design and Methodology

   • The mixed-method approach is appropriate, but details need clarification. Please specify:

     • Sampling criteria and justification for selecting only four experts.

     • How the discrete choice experiment (DCE) attributes and levels were validated.

     • Any reliability or validity checks conducted for the instrument.

   • Include a short subsection on ethical considerations beyond mentioning approval—describe how consent and confidentiality were maintained.

   • The description of the DCE is technical but lacks visualization. Add a figure or table summarizing the experimental design and flow.

5. Results and Analysis

   • While the statistical results are comprehensive, the explanation is largely textual. Add visual elements—such as bar charts or marginal effect plots—to enhance readability.

   • Provide deeper interpretation of findings rather than restating coefficients. Explain why certain variables (e.g., minimalist design) had negative or insignificant effects.

   • Include a subsection summarizing key insights from expert interviews to better connect the qualitative and quantitative phases.

6. Discussion

   • Strengthen the discussion by linking results to theory and past studies. Currently, it mostly reiterates findings. Discuss how results contribute to knowledge about sustainable marketing and digital user behavior.

   • Address managerial and policy implications—how can firms implement sustainable web design without falling into greenwashing traps?

7. Limitations and Future Work

   • The limitations section is well written, but future research directions should be expanded. Suggest potential comparative or longitudinal studies, cross-cultural validation, or AI-driven sustainable web analytics.

8. Language and Formatting

   • The manuscript is well-structured but occasionally verbose. Simplify long sentences and ensure consistent tense and terminology.

   • Check formatting per Sustainability journal guidelines, including reference style, citation consistency, and figure/table numbering.

   • Ensure uniformity in statistical notation (e.g., p < 0.05) and model naming (Models I, II, III).

9. References

   • The reference list includes a mix of academic and web-based sources. Replace non-scholarly URLs with peer-reviewed or official sources where possible.

   • Ensure all references are formatted according to MDPI guidelines

The manuscript addresses a novel and timely topic—sustainable web design and its digital marketing implications—but requires substantial revision to enhance scholarly rigor, clarity, and integration of theory with results. Strengthening methodological transparency, refining discussions, and improving presentation will significantly improve its quality and publication readiness.

Author Response

Comment 1:

Abstract and Keywords

The abstract is informative but overly descriptive. Please condense and focus more on the core objectives, methodology, key findings, and practical implications

Response 1:

The abstract has been rewritten to make it more concise.

 

Comment 2:

The keywords should be refined to include more specific terms such as “consumer perception,” “digital sustainability communication,” and “discrete choice modeling.”

Response 2:

The selection of keywords has been slightly adjusted.

 

Comment 3:

Introduction

The introduction provides a strong background but lacks a clear research gap and theoretical framing. Please articulate the novelty of this study more explicitly by highlighting what distinguishes it from prior work in sustainable digital design and marketing.

Response 3:

While it currently already has a description of the research gap. This has been strengthened by relating the study more strongly to underlying theories. Additionally, the literary background has been expanded.

 

Comment 4:

Strengthen the link between web design sustainability and its implications for digital marketing and consumer behavior.

Response 4:

As mentioned in the previous comment, the literary basis of the study has been expanded, including a broader frame of reference regarding sustainability in digital and software design.

 

Comment 5:

Literature Review

The literature discussion reads as an extended background rather than a critical review. Reorganize it to present thematic sub-sections (e.g., sustainable UX design, green hosting and consumer trust, digital greenwashing).

Response 5:

In the context of the expansion of the literature review it has been more clearly structured target the two issues, i.e., sustainable web design as such and the use of choice experiments in this context. The other mentioned aspects while of a fringe interest to the article are not the main focus and are thus skipped.

 

Comment 6:

Provide a concise conceptual framework or diagram summarizing the theoretical linkages among sustainability practices, user perception, and purchase behavior.

Response 6:

In the context of reworking the theoretical foundation of the article, the underlying framework has been detailed more as well.

 

Comment 7:

Several sources are descriptive; include more peer-reviewed academic literature instead of web sources where possible.

Response 7:

While reworking the literature review this aspect has been addressed as well.

 

Comment 8:

Research Design and Methodology

The mixed-method approach is appropriate, but details need clarification. Please specify:

Sampling criteria and justification for selecting only four experts.

How the discrete choice experiment (DCE) attributes and levels were validated.

Any reliability or validity checks conducted for the instrument.

Response 8

The overall motivation of the qualitative part of the study is detailed more and its role in the context of the mixed-method design is made clearer to the reader.

The DCEs attributes and levels were more closely linked to literature.

Information about a preliminary pretest have been added.

 

Comment 9:

Include a short subsection on ethical considerations beyond mentioning approval—describe how consent and confidentiality were maintained.

Response 9:

The respective parts in the text have been expanded.

 

Comment 10:

The description of the DCE is technical but lacks visualization. Add a figure or table summarizing the experimental design and flow.

Response 10:

An exemplary choice card has been included to illustrate the choice situation the participants were faced with.

 

Comment 11:

Results and Analysis

While the statistical results are comprehensive, the explanation is largely textual. Add visual elements—such as bar charts or marginal effect plots—to enhance readability.

Response 11:

In addition to Table 2 predicted probabilities, i.e., the respective ratios, are reported. The results are also reported graphically via a heatmap in a new Figure 2.

 

Comment 12:

Provide deeper interpretation of findings rather than restating coefficients. Explain why certain variables (e.g., minimalist design) had negative or insignificant effects.

Response 12:

The discussion of results has been enhanced, linking the results more strongly to theory.

 

Comment 13:

Include a subsection summarizing key insights from expert interviews to better connect the qualitative and quantitative phases.

Response 13:

A new sub-section has been added to the conclusion summarizing the theoretical contributions.

 

Comment 14:

Discussion

Strengthen the discussion by linking results to theory and past studies. Currently, it mostly reiterates findings. Discuss how results contribute to knowledge about sustainable marketing and digital user behavior.

Response 14:

As mentioned in the context of comment 12, the discussion of results has been enhanced and is more strongly linked to theory.

 

Comment 15:

Address managerial and policy implications—how can firms implement sustainable web design without falling into greenwashing traps?

Response 15:

This has already been addressed in the previous section 4.1. The recommendations have been adjusted to be more on point and more actionable.

 

Comment 16:

Limitations and Future Work

The limitations section is well written, but future research directions should be expanded. Suggest potential comparative or longitudinal studies, cross-cultural validation, or AI-driven sustainable web analytics.

Response 16:

This is a good point which is picked up in the respective section.

 

Comment 17:

Language and Formatting

The manuscript is well-structured but occasionally verbose. Simplify long sentences and ensure consistent tense and terminology.

Response 17:

Language has been slightly reworked throughout the article.

 

Comment 18:

Check formatting per Sustainability journal guidelines, including reference style, citation consistency, and figure/table numbering.

Response 18:

To our knowledge the article uses the official guidelines and citation/referencing style of sustainability.

 

Comment 19:

Ensure uniformity in statistical notation (e.g., p < 0.05) and model naming (Models I, II, III).

Response 19:

We have rechecked the article and cannot find any inconsistencies.

 

Comment 20:

References

The reference list includes a mix of academic and web-based sources. Replace non-scholarly URLs with peer-reviewed or official sources where possible.

Response 20:

As stated above the literature basis has been reworking and enhanced.

 

Comment 21:

Ensure all references are formatted according to MDPI guidelines

Response 21:

Regarding the officially available guidelines, all references are already following the style recommended for Sustainability.

 

The manuscript addresses a novel and timely topic—sustainable web design and its digital marketing implications—but requires substantial revision to enhance scholarly rigor, clarity, and integration of theory with results. Strengthening methodological transparency, refining discussions, and improving presentation will significantly improve its quality and publication readiness.

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

I have no more comments. Good luck to you.

Author Response

Comment:

I have no more comments. Good luck to you.

Response:

Thank you.

Reviewer 3 Report

Comments and Suggestions for Authors

Thank you for the opportunity to evaluate the revised manuscript. While the authors have made appreciable efforts—particularly in expanding the literature review, clarifying certain methodological steps, and improving model reporting—the core scientific issues identified in the first round remain only partially addressed. In several cases, the revisions are cosmetic or discursive rather than substantive, and the manuscript still exhibits major conceptual, methodological, and interpretive weaknesses that undermine its scientific contribution. 

The attached file comprises the key unresolved issues that preclude acceptance of the manuscript in its current form. 

Comments for author File: Comments.pdf

Author Response

Thank you for the opportunity to evaluate the revised manuscript. While the authors have made appreciable efforts-particularly in expanding the literature review, clarifying certain methodological steps, and improving model reporting-the core scientific issues identified in the first round remain only partially addressed. In several cases, the revisions are cosmetic or discursive rather than substantive, and the manuscript still exhibits major conceptual, methodological, and interpretive weaknesses that undermine its scientific contribution.

Below I summarize the key unresolved issues that preclude acceptance in its current form.

 

Comment 1:

Absence of a Coherent Theoretical Model

Despite additions referencing random utility theory and hedonic consumer choice, the paper still lacks a clear conceptual framework linking:

• web design attributes
• psychological mediators (e.g., perceived sustainability, trust, usability, aesthetics)
• outcomes (company impression, perceived quality, purchase likelihood).

No theoretical diagram, no hypotheses, and no formally derived relational logic are provided.

The authors defend the exclusion of behavioral theories (TPB, TAM, UTAUT, S-0-R) as "unsuitable," but this is incorrect: even within a DCE context, preference formation can and should be grounded in established behavioral and environmental psychology frameworks. The current manuscript remains theoretically underdeveloped.

Response 1:

The consumer choice theory by Lancaster is itself a behavioral theory, as is the random utility theory, positing that consumers base their choices on the attributes of goods and services. Consequently, it is these attributes and their levels that drive consumer choice and thus consumer behavior. The article therefore already has the theoretical foundation asked for.

The previous argument was that the objective of the cited theories, i.e., in particular, TAM-based theories, lies in explaining in a fixed framework the consumers adoption and use of technologies and services. Thus, their focus lies in user beliefs and perceptions of a product as drivers of the usage intention. This focus conflicts with DCEs whose primary objective lies in the elucidation of preference structures and thus a dissemination in how far each of the different attributes impact the final choice.

While current development have integrated both perspective in the context of ICLV models, the present article adopted an exploratory perspective and has thus adopted only a DCE perspective with the objective of elucidating the preference structures. Consequently, the stated mediators are not part of a DCE-based analysis and thus not considered in this context. The article has been appended to more strongly pronounce its exploratory nature.

Finally, since DCEs as compared to fixed models are mostly used as tools of exploratory data analysis, as is the case in this article as well. Consequently, hypotheses are not stated but instead open research questions have been deduced and used in the context of the study.

While we do not yet have integrated a figurative display how the studied attributes and the three outcomes can be linked with each other, the last revision already introduced a discussion how the results can be interpreted in the context of an SOR model. We additionally added a graphical display of this part of the analysis.

 

 

Comment 2:

Persistent Measurement Weaknesses

Major issues in construct operationalization remain uncorrected:

• All outcome variables remain single-item self-reports, without any psychometric validation.
• No evidence of reliability, discriminant validity, or conceptual justification for the chosen
• Attribute definitions such as "minimalistic" "sophisticated" design remain vague and unverifiable; no manipulation checks were conducted.
• The authors rely on subjective interpretations of textual stimuli in a way that cannot ensure consistent meaning across respondents.
• These weaknesses directly undermine the validity and reproducibility of the study's core measurements.

Response 2:

The first two comment seems to disregard the standard design of DCEs since their inception. The outcomes are the participants’ choices between two different website versions each. As such they are not self-reported answers by the participants and while multi-item constructs are theoretically possible in ICLV models they are not part of the standard DCEs design. As participants, before taking part in the choice experiment, were presented with additional explanations of the stimuli, these description have been added to the article to increase reproducibility of the study.

 

Comment 3:

Serious Sampling Problems Not Adequately Resolved The sampling strategy remains fundamentally problematic:

• The final sample (n = 107) is a convenience, self-selected, highly educated, sustainability-oriented group, incompatible with
• The authors attempt to justify adequacy by invoking a "sampling error below 10%" based on Paniotto's formula; however, this formula assumes a probability sample, which they do not have. The argument is therefore methodologically invalid.
• Mixed logit models with 5 attributes and 107 respondents remain statistically underpowered; the authors acknowledge this but still present these models as "robustness checks" without reporting convergence diagnostics.

The revisions do not sufficiently mitigate the fundamental external validity issues of the dataset.

Response 3:

While we agree with the assessment of the sample, we did argue and strengthened the argument that the implemented sample might not be representative of the average internet user, but the sample represents rather well those internet users that react the strongest on the provided stimuli.

Paniotto’s formula has been developed against the background of probabilistic samples, but it also known to be rather conservative regarding the sample size. We have additionally considered Yamane’s formula that is used for convenience samples as well. It returns results comparable to those of Paniotto.

Even though the point of convergence diagnostics remains rather vague we have conduct as well a likelihood ratio test for each of the models and reported the results.

 

Comment 4:

Limited Realism and Ecological Validity of the Experimental Design

Although the authors inserted additional explanation and limitations text, the core issue remains:

• Stimuli consist only of simplified textual descriptions of websites, not functioning prototypes or screenshots.
• Critical real-world cues (interaction design, content structure, aesthetics, trust signals) are absent.
• The attributes are not anchored in a controlled experimental visual environment, which is standard practice in UX-related DCEs.

Thus, ecological validity is still low, and claims about digital marketing implications remain overstated relative to the artificial nature of the stimuli.

Response 4:

We agree with the assessment of the limitations which are addressed even more strongly in the limitations section. We also added a respective cave-at to the interpretation of the results in a digital marketing context. However, this reduction of the stimuli and, as mentioned before, an explanation thereof channel the participants’ attention to the main aspects and ensure that they are not biased in their choices by tertiary design elements as might be the case with actual websites. Also attributes such as green hosting labels, in realistic environments often placed in a website’s footer, as less likely to be overlooked.

 

Comment 5:

Incomplete Statistical Interpretation and Model Evaluation

While p2  and prediction percentages were added, several issues persist:

• No marginal effects, odds ratios, or confidence intervals are reported.
• Interpretation of log-coefficients continues to overstate substantive implicat
• No diagnostic tests (e.g., IIA tests, convergence checks, residual analysis) are inclu
• Mixed logit results are presented without the transparency necessary for assessing model stability under small-sample conditions.

Overall, the quantitative analysis remains insufficiently rigorous for the claims made.

Response 5:

While the reported standard errors would fulfil the same function as confidence intervals, those have been added to the results. Considering that the core variables are binary marginal effects might be misleading, however, the results of likelihood ratio tests have been reported as well. As mixed logit results are considered to account for IIA, i.e., are considered to be IIA-free a comparision of the MNL results with the mixed logit results already indicates potential issues with IIA. A respective note has been added to the text.

The robust standard errors used for the determination of significance levels in both the mnl as well as the mixed-logit estimation are as per their implementation in the Apollo package clustered standard errors and thus particular suited for small sample sizes. A respective note has been added to the text.

 

Comment 6:

Overinterpretation and Residual Causal Language

Despite some improved wording, the manuscript still implies:

• that certain design choices "increase purchase decisions" or "foster behavior" rather than merely shift stated preferences.
• that DCEs are "a more sophisticated form of A/B testing," which is conceptually incorrect and misleading for

Without behavioral data, mediation testing, or field validation, these causal-leaning interpretations remain scientifically unjustified.

Response 6:

Regarding the first part of the comment: Parts of the text have been further adjusted, even though the term “foster behavor” does no longer appear in the text.

The comment regarding A/B tests is correct in so far as they operate differently but generate comparable results. We, however, agree with the assessment that this phrasing might be confusing for casual readers and deleted it.

 

Comment 7:

Data Transparency Remains Insufficient

The authors continue to provide data "on reasonable request", without depositing:

• the dataset,
• the choice design files, or
• the R scripts (Apollo code).

This falls short of FAIR principles and current reproducibility standards. No revision was made in this regard.

Response 7:

We do not follow this argument, as the chosen option is well covered with the journal policy regarding data sharing.

We suspect that the second point refers to the choice design, i.e., the implemented choice cards. In this context we have appended an overview of the choice cards and the R scripts in the appendix to the article.

 

Comment 8:

Qualitative Component Still Limited by Sampling and Lack of Saturation

Although methodological transparency improved (coding procedure, use of MaxQDA), the qualitative study still suffers from:

• an extremely small n = 4,
• all-male expert sample,
• recruitment through personal networks,
• no thematic

These limitations are not adequately acknowledged nor treated as significant constraints on the qualitative phase.

Response 8:

On the one hand the limitations to the expert as they exist have been reported in the article. As detailed the main criterion for the expert selection has been their expert knowledge about the topic and not their gender. While both male and female experts have been contacted only male experts responded to the invitation for an interview, thus any perceived inequality in the sample composition is purely incidental or due to a self-selection from the expert side.

At no point in the article is it stated that experts were contacted through personal networks. While social media networks, i.e., in particular, LinkedIn, have been used to find suitable candidates; thus, the search was not restricted to the personal network of the authors.

We agree and acknowledge the issue regarding thematic saturation, i.e., the focus solely on practitioning web designers. If the study would integrate the interviews as a part of equal standing it would have be expanded regarding its size and scope. As stated in the article, the objective of the interviews, however, remains adding an additional justification from a practitioners’ perspective to the theoretical justification of chosen attributes and levels. As such the focus solely on active web designer is considered more than fitting, as they are ideally qualified to offer exactly this perspective.

 

Comment 9:

Contribution and Novelty Claims Remain Overstated

Even with an expanded literature review, the paper still does not convincingly demonstrate novelty.

Much of what is presented (preferences for fast loading, positive reactions to eco-labels, visual minimalism effects) is already established in HCI, UX, and sustainable computing literature.

The study adds little beyond re-confirming known insights with a small, biased convenience sample.

Response 9:

While we agree with this comment that most of the attributes have been studied before in the literature, as is illustrated by the realized review of the literature. The literature as of yet misses a comprehensive and comparative approach focusing on the perceptions and preferences of consumers. This is what the article delivers and what currently has not been researched.

 

Recommendation

While the authors have made incremental improvements, the revised manuscript continues to exhibit major theoretical, methodological, and empirical limitations. The study's core validity problems-conceptual model absence, weak measurement, inadequate sample, limited ecological validity, incomplete statistical evaluation, and insufficient data transparency-remain unresolved to a degree that materially affects the scientific soundness of the work.

 

Comment 10:

At minimum, the manuscript requires a major revision involving:

• development of an explicit conceptual framework,
• improved measurement and attribute operationalization,
• clearer and statistically defensible modeling,
• substantial tempering of claims,
• and compliance with open-science data availability

Only after such fundamental concerns are addressed could the manuscript be reconsidered for publication.

Response 10:

These issues have been addressed separately above.

Reviewer 4 Report

Comments and Suggestions for Authors

• The paper addresses an important and emerging topic in sustainability and digital marketing, making a strong contribution to contemporary research.

• The mixed-methods approach integrating expert interviews and a discrete choice experiment is well designed and strengthens the reliability and depth of findings.

• The study clearly demonstrates the practical relevance of sustainable web design features, such as loading speed, green hosting, and sustainability seals, for consumer perception and purchase intention.

• The manuscript is well structured, coherent, and easy to follow, with a comprehensive literature foundation supporting the research framework.

• The discussion and recommendations for practitioners are particularly valuable and offer actionable insights for website designers and digital marketers.

• Consider shortening or consolidating some parts of the literature review to avoid repetition and improve flow.

• Provide additional clarity on how minimalistic vs. sophisticated design was defined and operationalized in the choice experiment.

• Enhancing discussion on cross-cultural applicability and broader industry contexts may strengthen generalizability in future research.

•  

Author Response

Comment 1:

The paper addresses an important and emerging topic in sustainability and digital marketing, making a strong contribution to contemporary research.

The mixed-methods approach integrating expert interviews and a discrete choice experiment is well designed and strengthens the reliability and depth of findings.

The study clearly demonstrates the practical relevance of sustainable web design features, such as loading speed, green hosting, and sustainability seals, for consumer perception and purchase intention.

The manuscript is well structured, coherent, and easy to follow, with a comprehensive literature foundation supporting the research framework.

The discussion and recommendations for practitioners are particularly valuable and offer actionable insights for website designers and digital marketers.

Response 1:

Thank you for this assessment.

 

Comment 2:

Consider shortening or consolidating some parts of the literature review to avoid repetition and improve flow.

Response 2:

Where possible the literature review has been reworked to make it more concise while conforming with the previous comments of the other reviewers.

 

Comment 3:

Provide additional clarity on how minimalistic vs. sophisticated design was defined and operationalized in the choice experiment.

Response 3:

Thank you for this fitting proposal. The participants were given additional information about the attributes and levels. In this context they also got a description what construes minimalistic and sophisticated designs.

 

Comment 4:

Enhancing discussion on cross-cultural applicability and broader industry contexts may strengthen generalizability in future research.

Response 4:

We fully agree with these aspects and have added additional comments in the limitations section. Some additional comments were also added to the overall discussion.

 

Round 3

Reviewer 3 Report

Comments and Suggestions for Authors

 

Dear Authors,

Thank you for submitting the revised version of your manuscript. While I acknowledge several improvements (clearer attribute definitions, added conceptual discussion/figure, and inclusion of the choice cards and Apollo code in the appendix), the revision does not yet resolve multiple scientific and methodological concerns that materially affect validity, interpretability, and reproducibility. In addition, the revision package is procedurally incomplete. Tracked-changes file required. You did not provide a version with track changes enabled. This prevents reviewers and editors from reliably verifying what has been modified and whether changes introduced new inconsistencies. Hence, I'm returning the manuscript without further review.