1. Introduction

Corporate Social Responsibility (CSR) has evolved from a peripheral activity into a core component of corporate strategy [1]. Effective communication of CSR activities not only enhances public trust but also establishes lasting brand advantages. However, the complexity and challenges of CSR communication have grown significantly with the rapid development of social media [2]. The bidirectional interactive features of social media provide new opportunities for CSR communication while placing higher demands on corporate communication strategies. Studies have shown that different CSR models yield varying communication effects [3]. For example, Corrective CSR emphasizes tangible improvements, which are most convincing to the public, while Cultivating Goodwill CSR tends to enhance brand image through simple announcements. However, these theoretical assumptions lack validation in practice, especially in the specific cultural context of China. This study aims to fill this gap by analyzing CSR communication content on the Weibo platform and examining the matching effectiveness of different CSR models and communication strategies. The research employs a content analysis approach, using CSR-related posts from the official Weibo accounts of 30 leading Chinese companies recognized for their CSR performance. These posts are categorized according to Shiu’s CSR models (Cultivating Goodwill, Compensating, and Corrective) and Etter’s communication strategies (Broadcasting, Reacting, and Engaging) to evaluate their matching effectiveness. The findings contribute to a deeper theoretical understanding of CSR communication and provide practical insights for companies to develop more effective communication strategies.

2. Literature Review

Research on Corporate Social Responsibility (CSR) has undergone extensive classification and expansion. Bowen (1953) first came up with the idea of corporate social responsibility (CSR) [4]. In 2021, Shiu divided CSR into three models based on its main goals and how it is carried out: Cultivating Goodwill CSR aims to improve brand image through donations to charities or sponsorships of cultural events; Compensating CSR indirectly balances out negative effects through donations to the environment or support for sustainable development; and Corrective CSR directly lowers negative effects through changes in business processes or product designs [5]. These models represent a progression from image enhancement to problem resolution, with the choice depending on industry characteristics and public expectations [1]. The effectiveness of CSR communication is influenced by communication strategies. Etter (2014) proposed three types of strategies tailored to the features of social media: Broadcasting, characterized by one-way dissemination of information, is suitable for Cultivating Goodwill CSR; Reacting, which involves responsive interaction through addressing public comments, fits Compensating CSR; and Engaging, which invites proactive public participation, aligns with Corrective CSR [3][6]. Although Engaging strategies are more effective in building trust and participation, most companies still prefer Broadcasting strategies, which limits the potential value of communication [2]. Furthermore, matching specific CSR models with suitable strategies is crucial for communication effectiveness. For example, Corrective CSR enhances participation and trust through Engaging strategies [7], Compensating CSR mitigates negative impacts through Reacting strategies [3], while Cultivating Goodwill CSR reinforces brand identity through Broadcasting strategies [8]. As China’s leading social media platform, Weibo provides a significant perspective for studying CSR communication [9].

3. Research Methodology

This study employs a content analysis approach using Weibo as the data source to explore the match between CSR models and communication strategies. Based on the Blue Book of Corporate Social Responsibility (2023), the top 30 companies recognized for their CSR performance in China were selected. These companies represent leaders in the field of CSR within the Chinese market. To ensure the data's representativeness and operational feasibility, official and active corporate Weibo accounts were further filtered as the focus of analysis. CSR-related posts from January 2023 to January 2024 were collected, resulting in a dataset of 143 posts.

3.1. Coding Framework

Firstly, a CSR model classification framework defined by Shiu (2021) will be employed, categorizing posts into three models: Cultivating Goodwill CSR, which involves activities such as charitable actions or cultural sponsorships aimed at enhancing brand image; Compensating CSR, which indirectly addresses negative impacts through environmental donations or sustainability initiatives; and Corrective CSR, which includes direct actions like process changes or product redesigns to mitigate negative impacts [5]. Secondly, the communication strategies proposed by Etter (2014) will be applied to classify posts into three categories: Broadcasting, characterized by one-way dissemination of information; Reacting, which involves responsive communication triggered by user comments; and Engaging, which promotes proactive public engagement through interactive discussions [3]. Finally, these classifications will serve as the foundation for analyzing the alignment between CSR models and communication strategies, enabling a deeper understanding of their matching effectiveness [5][3].

3.2. Data Analysis

Firstly, a scoring system was implemented to evaluate the effectiveness of different CSR model and communication strategy combinations, assigning 1 point for Broadcasting, 3 points for Reacting, and 5 points for Engaging [3]. Secondly, descriptive statistics were calculated to determine the average matching scores for each combination, and one-way ANOVA was conducted to identify significant differences in matching scores, followed by post hoc Tukey tests for specific pairwise comparisons [9]. Finally, the coders were trained thoroughly to make sure they all had the same understanding of CSR models and strategies. Cohen's Kappa was used to check the reliability of the categorizations [9].

4. Results

Table 1 and Table 2 are presented below to illustrate the key findings of this study. As you can see in Table 1, the matching scores for different CSR model-strategy combinations are spread out with their mean and standard deviation. This shows that they are not all equally effective. Table 2, on the other hand, presents the results of the one-way ANOVA, showing the significant differences in matching scores across combinations and supporting the hypothesis regarding optimal matching effectiveness.

Table 1: Statistical description

Table 2: ANOVA

4.1. Descriptive Statistics

Table 1 summarizes the mean and standard deviation of the matching scores for different CSR model-strategy combinations, highlighting significant variations in their effectiveness. The results show that Cultivating Goodwill CSR paired with Broadcasting scored the lowest, with a mean of 2.08 (SD = 1.54), significantly lower than other combinations. Compensating CSR paired with Reacting scored moderately, with a mean of 3.00 (SD = 1.59), while Corrective CSR paired with Engaging scored the highest, with a mean of 4.00 (SD = 1.60). These results suggest that the effectiveness of CSR model-strategy combinations varies significantly, with Corrective CSR benefiting most from its pairing with Engaging strategies. To visually represent these differences, Fig.1 compares the average matching scores of Cultivating Goodwill, Compensating, and Corrective CSR models. The figure illustrates that Corrective CSR achieves the highest score among the three models, significantly outperforming Compensating CSR and Cultivating Goodwill CSR, with statistical significance levels of p < 0.001 and p < 0.05, respectively. Cultivating Goodwill CSR, on the other hand, recorded the lowest score among all models, further emphasizing its relatively limited effectiveness when paired with Broadcasting strategies. These findings, supported by both the descriptive statistics in Table 1 and the visual representation in Fig.1, underline the importance of aligning CSR models with suitable communication strategies to achieve optimal effectiveness.

Figure 1: Comparison of CSR scores

4.2. Analysis of Variance (ANOVA)

One-way ANOVA results indicate significant differences in matching scores across CSR model-strategy combinations (F = 9.760, p < 0.001). Specifically, the match between Compensating CSR and Reacting strategies scored significantly higher than other combinations, with mean differences being statistically significant (p < 0.05). Corrective CSR paired with Engaging strategies ranked second, but it still significantly outperformed Cultivating Goodwill CSR paired with Broadcasting. Cultivating Goodwill CSR paired with Broadcasting scored the lowest and showed significant differences compared to other combinations (p < 0.05). These results support the hypothesis that Compensating CSR performs best when paired with Reacting strategies, Corrective CSR benefits significantly from Engaging strategies, while the effectiveness of Cultivating Goodwill CSR is limited when paired with Broadcasting strategies.

5. Discussion

5.1. Theoretical Contributions

Through the analysis of quantitative data, this study confirms that there are connections between various CSR models and communication strategies. It also shows the variety and complexity of how well CSR communication works by using actual scores. The results show that Corrective CSR achieved the highest scores (4.00), indicating that Engaging strategies have significant advantages in promoting CSR activities that directly address social or environmental issues. This finding supports Porter and Kramer’s (2011) “shared value” concept, emphasizing that substantial actions effectively enhance public trust and communication outcomes.

The moderate score of Compensating CSR (3.00) aligns with the hypothesis and further confirms the applicability of Reacting strategies in the communication of Compensating CSR. By promptly responding to public concerns, companies can maintain a certain level of public trust and achieve positive communication effects without directly modifying their operations.

Although Cultivating Goodwill CSR achieved the lowest scores (2.08), it significantly outperformed the hypothesized value of 1 point. This result suggests that Broadcasting strategies, despite their lack of interactivity, still fulfill basic public expectations by efficiently transmitting philanthropic information. This finding expands Chu et al.’s (2020) discussion on the limitations of Broadcasting strategies by providing empirical support for their potential value in specific contexts.

5.2. Matching Effectiveness Discussion

The analysis of CSR model-strategy matching effectiveness further deepens the understanding of communication outcomes across different combinations: By directly addressing social or environmental problems, corrective CSR shows that companies are responsible, and engaging strategies increase public awareness and participation through two-way communication. This combination gets the highest scores (4.00), which supports the idea that engaging strategies are very important for promoting corrective CSR and shows that people are more likely to trust companies that take proactive problem-solving actions, with engaging strategies making this effect stronger. Compensating CSR achieved moderate matching scores (3.00), supporting the theoretical assumption that Reacting strategies build public trust effectively; while Compensating CSR does not directly adjust corporate operations, timely responses to public inquiries significantly enhance communication outcomes, as Reacting strategies effectively meet public demands for transparency, mitigating dissatisfaction and fostering trust without addressing underlying issues. Although Cultivating Goodwill CSR scored the lowest in matching effectiveness (2.08), it significantly outperformed the hypothesized value of 1 point, indicating that Broadcasting strategies, despite their limitations in interactivity, still provide foundational value by transmitting clear philanthropic messages, as the efficient coverage of information by Broadcasting strategies likely explains why public perceptions of corporate goodwill exceed minimum expectations, but compared to more interactive strategies, Broadcasting remains less effective, explaining its relative disadvantage in matching effectiveness.

6. Conclusion

This study investigates the communication effectiveness of different CSR practices in the social media environment by examining the matching relationships between CSR models and communication strategies. The results show that there are strong connections between CSR models (Corrective, Compensating, and Cultivating Goodwill) and communication strategies (Engaging, Reacting, and Broadcasting) that have an effect on how communication works. Corrective CSR paired with Engaging strategies demonstrates the highest communication effectiveness by fostering public trust and participation through interactive formats, such as real-time updates and social media discussions. Compensating CSR with Reacting strategies is effective in addressing public concerns and enhancing transparency through timely responses and progress disclosures. Although Cultivating Goodwill CSR paired with Broadcasting strategies shows limited effectiveness, the use of multimedia and light interaction methods can still achieve foundational communication value. These findings enhance the understanding of CSR communication mechanisms and offer actionable recommendations for companies to optimize their strategies.

Despite its contributions, the study has limitations. The exclusive focus on Weibo as the primary platform, while representative of China’s social media landscape, restricts the generalizability of the findings to other platforms and cultural contexts. Furthermore, this research focuses on matching relationships without deeply exploring their impact on consumer behaviors, such as purchase intentions or brand loyalty. Future studies could expand to other social media platforms, adopt cross-cultural perspectives, and use experimental or longitudinal methods to validate and deepen these findings.

In summary, this study highlights the critical importance of aligning CSR models with appropriate communication strategies to optimize communication outcomes. By bridging theoretical validation with practical insights, the research provides valuable guidance for companies to enhance stakeholder trust and engagement while opening new directions for academic exploration. CSR communication is a dynamic and evolving practice, and future investigations will further support the balance between corporate social responsibility and business value.