Communication Strategies of Startups During the Natural Catastrophe of the 2024 DANA: Impact on Public Opinion and Business Reputation

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

1. The author employs GPT-4 to classify the sentiment of social media posts, but the validation process for these classifications has not been clearly explained. It is recommended that the author provide details on who conducted the validation, the proportion of data that was reviewed, and whether any reliability testing of the classification results was performed.
2. The author is advised to define the criteria for selecting the 60 posts with “high engagement.” In addition, information regarding the number and characteristics of interview participants, such as their business sectors, company scale, and recruitment methods, should be included.
3. The author should explain the exclusion of professional platforms such as LinkedIn and X (Twitter), which are arguably relevant to startups.
4. The second research question concerning organizational resilience has not yet been operationalized. Is resilience assessed solely based on interviewee perceptions? The author is advised to elaborate on the indicators or frameworks used to define this concept more measurably.
5. The Discussion section reiterates previous findings and does not explicitly connect empirical results to theoretical frameworks. To deepen the analysis and interpretation, the author should consider integrating additional theories, such as Media Richness Theory, Uses and Gratifications Theory, Stakeholder Theory, Organizational Learning Theory, and Agenda-setting Theory.
6. Several figures and tables do not include specific numbers or percentages, making them difficult to interpret and less informative quantitatively.
7. The interview findings are presented only in generalized summaries. Given the qualitative approach adopted, the author is encouraged to include anonymized direct quotes from respondents to enhance the depth and credibility of the findings.
8. The finding that TikTok exhibits a narrative disconnection from the crisis context has not been analytically discussed. The author is advised to explore further the causes of this phenomenon and its implications for the effectiveness of cross-platform communication strategies.

Comments on the Quality of English Language

There are several instances of inconsistent tense usage and awkward grammatical structures. For example: “Only few days later the same strategy was used by...” This should be revised to:
“Only a few days later, the same strategy was used by...”

Additionally, some sentences are overly narrative or informal for an academic article. For instance: “This was a very emotional post and triggered several comments of solidarity.”
A more appropriate academic phrasing would be: “This post carried a strong emotional tone and generated numerous comments expressing solidarity.”

The authors are encouraged to revise the manuscript for grammatical consistency, tense alignment, and academic style throughout the text.

Author Response

We sincerely appreciate the time and effort dedicated to evaluating our manuscript. Your constructive comments have been invaluable in refining and enhancing the quality of our work. We have carefully considered all your suggestions and made the necessary revisions to address the concerns you raised. Below, we provide detailed responses to each point, outlining the changes implemented in the revised manuscript.

We trust that these revisions have strengthened the paper and that it now represents a more robust and meaningful contribution to the field. Please do not hesitate to contact us if you need any further clarifications or if you believe additional modifications are necessary.

The authors

 

PEER REVIEW COMMENTS

 

1. Validation of GPT-4 sentiment analysis

The author employs GPT-4 to classify the sentiment of social media posts, but the validation process for these classifications has not been clearly explained. It is recommended that the author provide details on who conducted the validation, the proportion of data that was reviewed, and whether any reliability testing of the classification results was performed.

We have added a detailed explanation in line 256:

“The manual validation process was conducted by researchers with expertise in digital communication and qualitative analysis. A total of 100% of the GPT-4 sentiment classifications were reviewed manually to ensure consistency with the context of each post and the overall tone.  While no inter-coder reliability test was formally applied due to the exploratory nature of this study, the high level of agreement observed between reviewers reinforces the reliability of the sentiment analysis.”

 

2. Criteria for post selection and participant details

The author is advised to define the criteria for selecting the 60 posts with “high engagement.” In addition, information regarding the number and characteristics of interview participants, such as their business sectors, company scale, and recruitment methods, should be included.

This has been addressed in two sections of the revised manuscript.

In line 265:

“A total of 60 high-engagement posts were selected for the sentiment analysis. These included the top 10 posts from each platform: Instagram, TikTok, and Facebook, during two distinct timeframes: the critical period (28 October to 15 November 2024) and the extended period (16 November 2024 to 31 January 2025). This approach ensured a balanced representation of emotionally relevant and high-impact content for each stage of the crisis, resulting in 30 posts per period. Selection was based on engagement metrics provided by the Fanpage Karma tool, including reactions, comments, shares, and overall interaction levels.”

 

Additionally, in line 280:

“The responses were gathered through an open-ended online questionnaire designed to encourage narrative expression, rather than binary or closed answers. The participating startups represented diverse sectors, including mobility, healthtech, and digital services, and were characterised as micro-enterprises with fewer than 50 employees. Recruitment was carried out via purposive sampling, through direct invitations sent via email to startups that had shown active social media presence during the DANA crisis. Although the sample size is modest, the depth and thematic consistency of the responses support the exploratory nature and interpretive aims of this study.”

3. Exclusion of LinkedIn and X (Twitter)

The author should explain the exclusion of professional platforms such as LinkedIn and X (Twitter), which are arguably relevant to startups.

This justification was already included in line 216.

 

4. Operationalisation of organisational resilience

The second research question concerning organizational resilience has not yet been operationalized. Is resilience assessed solely based on interviewee perceptions? The author is advised to elaborate on the indicators or frameworks used to define this concept more measurably.

We have addressed this in the conclusion, line 601:

“Regarding the second research question, the study found that organisational resilience in startups was expressed through the ability to adapt communication strategies quickly, ensure operational continuity, and show empathetic leadership during the crisis. These traits were identified through the participants’ narratives and reflect how startups managed to maintain internal stability and public trust, despite their limited resources. Rather than relying on quantitative indicators, resilience was understood as a communicative and cultural capacity to respond constructively in adverse situations.”

5. Weak theoretical connection in the Discussion.

The Discussion section reiterates previous findings and does not explicitly connect empirical results to theoretical frameworks. To deepen the analysis and interpretation, the author should consider integrating additional theories, such as Media Richness Theory, Uses and Gratifications Theory, Stakeholder Theory, Organizational Learning Theory, and Agenda-setting Theory.

In order to justify this in the discussion, changes have first been added to the theoretical framework, and then to the discussion section:

 

Line 126:

“In addition to SCCT, other theoretical perspectives can help frame the communicative dynamics observed during crisis scenarios. Media Richness Theory offers insights into the effectiveness of emotionally charged and visually rich platforms, such as Instagram and Facebook, when immediacy and contextual clarity are required (Liu et al., 2016). Uses and Gratifications Theory contributes to understanding why certain types of content resonate more strongly with audiences in times of uncertainty, based on users’ active search for emotional connection, information, or reassurance (Hossain, 2019). From a strategic standpoint, Stakeholder Theory provides a relevant framework for analysing how startups adapt their messaging to address the expectations of different audiences, especially in contexts of reputational vulnerability (Gossel, 2022).”

Line 585:

“These findings can be further interpreted through the theoretical lenses: from a Media Richness Theory perspective, the use of platforms like Instagram and Facebook reflects an intentional choice to communicate through emotionally rich and immediate formats (Liu et al., 2016). Uses and Gratifications Theory helps explain why audiences responded to content offering emotional connection and community support (Hossain, 2019). Stakeholder Theory offers a strategic framework to understand how startups tailored their messages to maintain engagement with key publics and manage reputational expectations (Gossel, 2022).”

Also, we have included additional references:

 

Liu, B. F., Fraustino, J. D., & Jin, Y. (2016). Social Media Use During Disasters: How Information Form and Source Influence Intended Behavioral Responses. Communication Research, 43(5), 626–646. https://doi.org/10.1177/0093650214565917

 

Hossain, Md. A. (2019). Effects of uses and gratifications on social media use: The Facebook case with multiple mediator analysis. PSU Research Review, 3(1), 16–28. https://doi.org/10.1108/prr-07-2018-0023

 

Gossel, B. M. (2022). Analogies in Entrepreneurial Communication and Strategic Communication: Definition, Delimitation of Research Programs and Future Research. International Journal of Strategic Communication, 16(2), 134–156. https://doi.org/10.1080/1553118x.2021.2015689

 

 

6. Lack of quantitative detail in figures and tables.

Several figures and tables do not include specific numbers or percentages, making them difficult to interpret and less informative quantitatively.

Figures 5, 7, 9, and 10 have been updated to include numerical data (frequencies and/or percentages), thereby enhancing the clarity and interpretability of the visual elements.

 

7. Inclusion of direct quotes from interviews

The interview findings are presented only in generalized summaries. Given the qualitative approach adopted, the author is encouraged to include anonymized direct quotes from respondents to enhance the depth and credibility of the findings.

We have incorporated anonymised direct quotes:

 

Line 527:

One participant explained, “We didn’t have a risk management plan, but from the first moment we knew we had to act quickly, especially through social media.” Another added, “Our main communication channel was WhatsApp, and that allowed us to stay agile and aligned.”

 

Line 537:

“Our reputation came out stronger. People saw our human side and our ability to act fast.” It was also noted that the structural flexibility of startups can become an advantage when responding swiftly to adverse scenarios. As one respondent put it, “Not having rigid structures allowed us to adapt quickly, even though we lacked predefined protocols.”

 

Line 544:

 One participant summed it up as follows: “Empathy was the most powerful tool we had to connect—with the team, with our clients, with those affected.”

 

8. Analytical discussion regarding TikTok

The finding that TikTok exhibits a narrative disconnection from the crisis context has not been analytically discussed. The author is advised to explore further the causes of this phenomenon and its implications for the effectiveness of cross-platform communication strategies.

We have addressed this in line 610:

“The absence of crisis-related content on TikTok may reflect uncertainty about how to communicate effectively on entertainment-focused platforms, or a lack of strategic alignment between communications teams and their brand tone. However, neglecting to integrate all active platforms into a unifying crisis narrative can fragment communication efforts and weaken public perception, particularly among younger audiences who rely on these channels for information and engagement.”

 

9. Language and style issues

We have reviewed the manuscript thoroughly to correct inconsistent tenses and improve sentence structure. Specific examples cited by the reviewer (e.g., “Only few days later...”) have been revised for grammatical accuracy and academic tone.

 

 

Reviewer 2 Report

Comments and Suggestions for Authors

Your study presents an insightful and timely contribution to the growing body of literature on crisis communication, particularly in the context of startup resilience and digital media engagement.

The integration of quantitative social media analytics, sentiment analysis using GPT-4, and qualitative interviews results in a well-rounded and methodologically sound investigation. The manuscript is clearly written, well-structured, and supported by a robust theoretical framework. Your findings offer both academic and practical value, especially in demonstrating how emotionally adaptive communication can enhance public trust and organizational legitimacy during crises.

That said, I recommend minor revisions prior to acceptance. These include a more explicit discussion on the ethical considerations and limitations of using AI-based sentiment analysis, greater elaboration on the lack of crisis-related content on TikTok, and more nuanced theoretical engagement with concepts such as emotional branding or symbolic resilience. Additionally, minor editorial refinements would improve the manuscript's clarity and conciseness.

Overall, I commend your work and believe that, with these adjustments, it will make a valuable contribution to the journal.

Comments for author File: Comments.pdf

Author Response

We sincerely appreciate the time and effort dedicated to evaluating our manuscript. Your constructive comments have been invaluable in refining and enhancing the quality of our work. We have carefully considered all your suggestions and made the necessary revisions to address the concerns you raised. Below, we provide detailed responses to each point, outlining the changes implemented in the revised manuscript.

We trust that these revisions have strengthened the paper and that it now represents a more robust and meaningful contribution to the field. Please do not hesitate to contact us if you need any further clarifications or if you believe additional modifications are necessary.

The authors

 

PEER REVIEW COMMENTS

 

1. Ethical considerations and limitations of AI-based sentiment analysis

I recommend a more explicit discussion on the ethical considerations and limitations of using AI-based sentiment analysis.

In response, we have included line 632:

“Moreover, while the use of AI-based sentiment analysis tools such as GPT-4 improves processing power and efficiency, it can also have limitations. These include the potential for contextual misinterpretation, lack of cultural nuance and reliance on algorithmic assumptions.”

Line 620:

“…the last part of which is understood as the communicative capacity of organisations to project strength, legitimacy and emotional alignment with their audiences in times of crisis, while also acknowledging that digital platforms serve not only strategic goals but also act as relational and cultural spaces where trust and collective meaning are co-constructed.”

2. Crisis-related content on TikTok

Greater elaboration on the lack of crisis-related content on TikTok is recommended.

This suggestion has been incorporated into line 610:

The absence of crisis-related content on TikTok may reflect uncertainty about how to communicate effectively on entertainment-focused platforms, or a lack of strategic alignment between communications teams and their brand tone. However, neglecting to integrate all active platforms into a unifying crisis narrative can fragment communication efforts and weaken public perception, particularly among younger audiences who rely on these channels for information and engagement.

Reviewer 3 Report

Comments and Suggestions for Authors

Dear Author(s)

I am pleased that I got the opportunity to review this paper, which deals with a relevant and contemporary topic. The combination of quantitative content analysis, sentiment analysis using GPT-4 and supplementary interviews represents a methodologically interesting framework that I believe has the potential to contribute to the understanding of crisis communication in the digital environment.

Also, I find that the paper is well structured and the results are clearly presented.

However, in order for this study to reach its full potential, I believe there are several important aspects that need to be considered in revision:

The references used are mostly contemporary and thematically relevant, which is a strong point of the work. But there is room for a stronger theoretical-analytical connection. For example, certain theoretical concepts (especially SCCT and organizational resilience) remain more descriptive than analytical in the function of interpreting findings or are absent. I think that a stronger connection between theory and results would contribute to the overall academic credibility of the paper.

While organizational resilience is only superficially mentioned in the analysis, SCCT (Situational Crisis Communication Theory) is not applied in the analytical segment of the paper. There is no classification of strategies according to the SCCT model (rebuild, deny, diminish), nor is there an attempt to identify how the degree of perceived responsibility affects the choice of communication tactics. This is why there is a discontinuity between the theoretical framework and practical analysis. Because of all this, I recommend that SCCT or:

- remove from the theoretical part if it is not possible to introduce it methodologically into the analysis,
or
- include at least a basic categorization of communication examples according to this theory.

Although the interviews were methodologically justified and ethically conducted, in the analysis itself I did not see any direct quotes or illustrative insights that would strengthen the interpretation. The inclusion of a few illustrative statements could perhaps contribute to the depth of insight and understanding of the narrative from the perspective of the actors themselves.

The paper clearly identifies that the content on TikTok was thematically detached from the context of the crisis, which is a significant finding. However, it has not been sufficiently reflected whether it is a failure in the crisis strategy or a conscious decision based on the nature of the platform itself. I believe that adding a brief explanation or discussion of that issue would improve the overall discussion.

This research has a high degree of originality and brings a methodologically interesting framework for the study of crisis communication in the context of startups. However, in order to preserve the coherence between theory and analysis, as well as to strengthen the depth of interpretation, I think that we should work on the aspects I recommended. I genuinely believe that, with some targeted improvements, this paper could make a strong and meaningful contribution to the field of crisis communication.

Author Response

We sincerely appreciate the time and effort dedicated to evaluating our manuscript. Your constructive comments have been invaluable in refining and enhancing the quality of our work. We have carefully considered all your suggestions and made the necessary revisions to address the concerns you raised. Below, we provide detailed responses to each point, outlining the changes implemented in the revised manuscript.

We trust that these revisions have strengthened the paper and that it now represents a more robust and meaningful contribution to the field. Please do not hesitate to contact us if you need any further clarifications or if you believe additional modifications are necessary.

The authors

 

PEER REVIEW COMMENTS

 

1. Theoretical-analytical connection between SCCT and findings

Certain theoretical concepts (especially SCCT and organizational resilience) remain more descriptive than analytical in the function of interpreting findings or are absent.

Changes have first been added to the theoretical framework, and then to the discussion section:

Line 126:

“In addition to SCCT, other theoretical perspectives can help frame the communicative dynamics observed during crisis scenarios. Media Richness Theory offers insights into the effectiveness of emotionally charged and visually rich platforms, such as Instagram and Facebook, when immediacy and contextual clarity are required (Liu et al., 2016). Uses and Gratifications Theory contributes to understanding why certain types of content resonate more strongly with audiences in times of uncertainty, based on users’ active search for emotional connection, information, or reassurance (Hossain, 2019). From a strategic standpoint, Stakeholder Theory provides a relevant framework for analysing how startups adapt their messaging to address the expectations of different audiences, especially in contexts of reputational vulnerability (Gossel, 2022).”

Line 585:

“These findings can be further interpreted through the theoretical lenses: from a Media Richness Theory perspective, the use of platforms like Instagram and Facebook reflects an intentional choice to communicate through emotionally rich and immediate formats (Liu et al., 2016). Uses and Gratifications Theory helps explain why audiences responded to content offering emotional connection and community support (Hossain, 2019). Stakeholder Theory offers a strategic framework to understand how startups tailored their messages to maintain engagement with key publics and manage reputational expectations (Gossel, 2022).”

Also, we have included additional references:

 

Liu, B. F., Fraustino, J. D., & Jin, Y. (2016). Social Media Use During Disasters: How Information Form and Source Influence Intended Behavioral Responses. Communication Research, 43(5), 626–646. https://doi.org/10.1177/0093650214565917

 

Hossain, Md. A. (2019). Effects of uses and gratifications on social media use: The Facebook case with multiple mediator analysis. PSU Research Review, 3(1), 16–28. https://doi.org/10.1108/prr-07-2018-0023

 

Gossel, B. M. (2022). Analogies in Entrepreneurial Communication and Strategic Communication: Definition, Delimitation of Research Programs and Future Research. International Journal of Strategic Communication, 16(2), 134–156. https://doi.org/10.1080/1553118x.2021.2015689

 

2. Interview quotes

We have incorporated anonymised direct quotes:

 

Line 527:

One participant explained, “We didn’t have a risk management plan, but from the first moment we knew we had to act quickly, especially through social media.” Another added, “Our main communication channel was WhatsApp, and that allowed us to stay agile and aligned.”

 

Line 537:

“Our reputation came out stronger. People saw our human side and our ability to act fast.” It was also noted that the structural flexibility of startups can become an advantage when responding swiftly to adverse scenarios. As one respondent put it, “Not having rigid structures allowed us to adapt quickly, even though we lacked predefined protocols.”

 

Line 544:

 One participant summed it up as follows: “Empathy was the most powerful tool we had to connect—with the team, with our clients, with those affected.”

 

3. TikTok's narrative disconnection

…it has not been sufficiently reflected whether it is a failure in the crisis strategy or a conscious decision based on the nature of the platform itself. I believe that adding a brief explanation or discussion of that issue would improve the overall discussion.

We have added line 610:

 

The absence of crisis-related content on TikTok may reflect uncertainty about how to communicate effectively on entertainment-focused platforms, or a lack of strategic alignment between communications teams and their brand tone. However, neglecting to integrate all active platforms into a unifying crisis narrative can fragment communication efforts and weaken public perception, particularly among younger audiences who rely on these channels for information and engagement.

 

Reviewer 4 Report

Comments and Suggestions for Authors

The paper addresses the communication strategies adopted by startups during and after an extreme event that occurred in Valencia in 2024. It provides an acceptable reconstruction of the literature on crisis communication and the methodologies used; however, these elements are not sufficient to support a solid theoretical framework. In my opinion, the main limitations are as follows:

1. The flood case study is treated in an overly instrumental manner: the cultural, social, and symbolic context of the crisis is not adequately explored. This is particularly problematic given that the paper employs sentiment analysis. Without a deeper examination of the socio-cultural dimension of the event, it is difficult to grasp the true emotional and symbolic meaning of the communications analyzed.
2. The results section is overly descriptive and lacks analytical depth: paragraphs 5 and 6 (with a numbering error in paragraph 5) are addressed too briefly and do not include adequate interpretive reflection. This weakness seems to stem from a certain vagueness in the research question and in the focus of the phenomenon under study.
3. The implicit assumption that digital platforms are merely tools for marketing and brand identity construction appears reductive. This perspective overlooks the complex social, relational, and cultural implications of digital media usage, which are widely discussed in the recent media studies literature.

For these reasons, I believe the paper requires substantial and thorough revision before it can be considered for publication. I particularly recommend:

1. Reformulating the research question in a more specific way;
2. Integrating a stronger theoretical framework, also from a sociological and media studies perspective;
3. Strengthening the analysis of qualitative data and better contextualizing the case study in relation to the crisis experienced by the local population.

Author Response

We sincerely appreciate the time and effort dedicated to evaluating our manuscript. Your constructive comments have been invaluable in refining and enhancing the quality of our work. We have carefully considered all your suggestions and made the necessary revisions to address the concerns you raised. Below, we provide detailed responses to each point, outlining the changes implemented in the revised manuscript.

We trust that these revisions have strengthened the paper and that it now represents a more robust and meaningful contribution to the field. Please do not hesitate to contact us if you need any further clarifications or if you believe additional modifications are necessary.

The authors

 

PEER REVIEW COMMENTS

 

1. Cultural, social, and symbolic contextualisation of the flood case

We have added line 546:

“These responses reveal not only operational agility, but also a deeper cultural responsiveness to the emotional climate that surrounded the DANA. For many startups, communication became a way of expressing solidarity and reaffirming their role within a community that mourned, mobilised and sought support. This symbolic and affective dimension is essential to fully understand the resonance of the messages analysed”

2. Descriptive nature of the results section

In response to this comment, both sections have been revised. The discussion now includes a more structured interpretation of the findings and integrates multiple theoretical frames, which define startups' communicative behaviour in a more analytically way. The conclusions section has also been expanded to clearly address the two research questions and incorporate reflections on multi-platform coherence, symbolic resilience and the socio-emotional context of the DANA crisis.

3. Digital platforms treated as instrumental tools

We have added line 620:

“…the last part of which is understood as the communicative capacity of organisations to project strength, legitimacy and emotional alignment with their audiences in times of crisis, while also acknowledging that digital platforms serve not only strategic goals but also act as relational and cultural spaces where trust and collective meaning are co-constructed.”

4. Stronger theoretical integration

We have expanded the theoretical framework to incorporate concepts from media and communication studies, including Media Richness Theory, Uses and Gratifications Theory, Stakeholder Theory… These theories are explicitly applied in the discussion section to interpret the communicative behaviour of startups across platforms.

 

 

Round 2

Reviewer 4 Report

Comments and Suggestions for Authors

I appreciated the efforts made to address my comments. I therefore consider the article to be sufficiently aligned with the scientific standards required for publication.