Campaign participation, spreading electronic word of mouth, purchase: how to optimise corporate social responsibility, CSR, effectiveness via social media?

Purpose – The purpose of this paper is to investigate how to optimise corporate social responsibility (CSR) when communicating via social media. In particular, the communication type, cause proximity and CSR motives are addressed facing the increased demand for transparency and the grown consumers ’ expectations towards socially responsible brands. Design/methodology/approach – Online survey was developed, based on a profound literature review and a field research we did on the actual social media behaviour of socially responsible brands. Consumers attitudes and behavioural reactions in terms of willingness to participate in a campaign, to spread e-WOM and to purchase were investigated, as a function of CSR motives (value vs performance vs value and performance) and cause proximity (national vs international), respectively, with monologue (study 1) and dialogue communication type (study 2). Findings – Cause proximity enhanced the campaign participation, and this effect was pronounced for both, monologue and dialogue type of communication. CSR motives modulated the willingness to spread electronic word-of-mouth, and this holds for both, monologue and dialogue communication. Attitudes and purchase intention were highest when value- and performance-driven motives were communicated, but these effects appeared only when the message was in a dialogue form of communication. Message credibility and CSR motives credibility perception further modulated consumers response. Practical implications – The outcomes could be used in developing marketing (communication) strategies leading to values and revenues optimisation. Originality/value – The results are discussed in a framework of how CSR resonates via social media.


Introduction
Corporate social responsibility (CSR) was recognised as an efficient tool to enhance attitudes towards a company (Bhattacharya and Sen, 2004), its financial performance and market value (Du et al., 2010). Defined as commitment to improve community well-being through discretionary business practices and contributions of corporate resources (Kotler and Lee, 2005), CSR has been increasingly incorporated in companies' portfolio over the past decade (e.g. Badenes-Rocha et al., 2019;Bign e et al., 2010;Du et al., 2010;Groza et al., 2011).
Note, however, that launching CSR initiatives, and even communicating these, e.g. via the company's web page, does not necessarily mean that consumers respond to these initiatives. A recent study indeed reported that the overall CSR consumer discourse is still very limited (D'Acunto et al., 2020), despite general interest in CSR (Ettinger et al., 2018). An auditing of the literature even showed that incorporating CSR into company's strategies is paradoxically associated with corporate social irresponsibility (Riera and Iborra, 2017). The above outcomes concerning the way consumers perceive corporate activities are a warning call to the expanding number of CSR initiatives. Namely, there is increased demand for transparency and emergent need to better understand the effectiveness of CSR.
Furthermore, the drastic change of the communication landscape forced companies to change the way they interact with their customers (Kumar et al., 2016;Colicev et al., 2018) when communicating about the brand (Stojanovic et al., 2018), the company  and CSR initiatives (Badenes-Rocha et al., 2019). In particular, social media (e.g. Facebook, Twitter) are increasingly used to faster and more efficiently reach customers.
Social media implementation, nevertheless, raised consumers' expectations regarding CSR. Recent papers, auditing relevant literature, acknowledged the need for digitalisation in CSR communication (Verk et al., 2019), as well as the consumers' sensitivity by asking: do CSR messages resonate via social media (Saxton et al., 2019). The burning question however is: How CSR messages resonate via social media, and thus, how to optimise CSR effectiveness via social media?
The current paper addresses this question, providing in-depth exploration on various parameters hypothesised as key drivers of CSR effectiveness. We complement the existing body of literature bringing together theories of CSR motives (e.g. Groza et al., 2011;Ellen et al., 2006), cause proximity effect (e.g. Grau;Folse, 2007;Groza et al., 2011), communication type (e.g. Du et al., 2010;Korschun and Du, 2013), credibility (e.g. Flanagin and Metzger, 2000;Goldsmith et al., 2000) and social media impact (e.g. Badenes-Rocha et al., 2019;Bign e et al., 2019) in a single framework. With this framework, we aim at shedding much needed light on how to maximise the CSR effectiveness.
First, we look at consumers' attitudes and how these influence behavioural reactions in terms of willingness to participate in a campaign, to spread electronic word-of-mouth (e-WOM) and to purchase, when consumers are exposed to social media posting about CSR. Second, we explore how the above parameters resonate as a function of cause proximity (national vs international), CSR motives (value vs performance vs value and performance) and communication type (monologue vs dialogue), emerging as major determinants of consumer response. Third, we investigate how perceived message credibility and CSR motives credibility can influence consumers' attitudes and behavioural responses. Forth, the outcomes are summarised in a framework of how CSR messages resonate via social media.
The paper is organised as follows: introducing the theoretical background that inspired the empirical research. Based on a profound literature review and a field research we did on the actual social media behaviour of socially responsible brands, an online survey was developed, addressing how to optimise the CSR impact via social media. Two studies are reported, respectively, when consumers are exposed to monologue (study 1) and dialogue (study 2) form of communication. The results are discussed in a framework of how CSR resonates via social media. The paper sums up with practical implications suggesting development of marketing (communication) strategies leading to values and revenues optimisation.
2. Theoretical background 2.1 Cause proximity Cause proximity reflects the distance between the CSR activity and the consumer (Grau and Folse, 2007) and was associated with local, regional, national and international initiatives (Varadarajan and Menon, 1988). Note, however, previous studies are not univocal on proximity EJMBE effect when it comes to cause-related marketing. While local donations and positive message framing served as effective cues to generate favourable attitudes (Grau and Folse, 2007), the hypothesised proximity effect was not significant on attitudes (Ross et al., 1992). A plausible explanation for the discrepancies in findings could be the fact that these earlier studies employed different scales and measurements. A study re-examining the proximity effect reported a moderate role of spatial proximity of the CSR initiative on values-driven attributions (Groza et al., 2011). It was also found that proximity (local vs national vs international) affects consumer identification with the cause and the corporate image evaluation. Previous findings are a cornerstone to build around and to further look at how proximity reflects in behaviour reactions.
In the present context, addressing the effect of social media, it is important to investigate how proximity influences consumers' attitudes and response in terms of willingness to participate in a campaign (Grau and Folse, 2007), to spread word-of-mouth (Zeithaml et al., 1996), online (Stojanovic et al., 2017) and to buy a product (Groza et al., 2011). We hypothesise: H1a. Attitudes will be more positive when national (than international) cause is initiated and communicated to consumers.
H1b. Willingness to participate in a campaign will be higher when national (than international) cause is initiated and communicated to consumers.
H1c. Willingness to spread e-WOM will be higher when national (than international) cause is initiated and communicated to consumers.
H1d. Willingness to buy a product will be higher when national (than international) cause is initiated and communicated to consumers.
Following the classical theory, that attitudes are beliefs about an object/person that affect intention to behave and actual behaviour (theory of planned behaviour (TPB), Ajzen, 1991), we are interested to know how beliefs about various corporate motives may modulate consumer's response.

Corporate social responsibility motives
Three types of CSR motives were classified in a taxonomy, in line with the company's actions, as performance-, stakeholders-or value-driven motives (Groza et al., 2011). The performancedriven motives are defined as an instrument to achieve a company's performance objectives in terms of sales raise, profitability and investments return (Swanson, 1995). While stakeholder-driven motives reflect the implementation of CSR initiatives, which align to stakeholders' norms (Maignan and Ralston, 2002), the value-driven motives reflect the CSR initiatives businesses generate to positively impact the society (Hooghiemstra, 2000). CSR motives underlying a company's actions have also been addressed from the perspective of consumers (Groza et al., 2011;Ellen et al., 2006). When consumers perceive that a company focuses on itself, i.e. to achieve its performance objectives (Ellen et al., 2006), this reflects performance-driven CSR motives (in terms of Swanson, 1995). When a company focuses on others than itself, i.e. stakeholders, this is comparable to stakeholder-driven CSR (as defined by Maignan and Ralston, 2002). Respectively, the society-oriented activities are attributed to the value-driven CSR motives.
CSR motives were reported to significantly influence consumers' perception and thus response (Sen et al., 2009;Vlachos et al., 2009). Not surprisingly then, in various studies, different motives were extensively explored. In general, the nature of CSR initiatives (value-, strategicstakeholder-driven) reflects consumer attribution effects, and these attributions are mediators in consumers' responses to CSR (Groza et al., 2011). Consumers responded most positively to CSR initiatives judged as values driven and strategic, while CSR efforts perceived as egoistic received negative reactions (Ellen et al., 2006). Overall attitude was greater in socially motivated CSR effectiveness via social media than in the profit-motivated conditions (Becker-Olsen et al., 2006), but this effect was only pronounced with high fit between the firm and social initiative. In the low fit condition, both, profit and social motives influenced attitudes positively. The authors explained these results with consumers' beliefs that a firm could be profit motivated. However, consumers are sceptical in case of discrepancy between stated objectives and the firm's actual actions. In other words, performance-strategic and social-value motives are not mutually exclusive. One might even argue that combining value-and performance-driven motives could generate most positive response. Whether this is the case, we test: H2. When a message combines value-and performance-driven CSR motives (in comparison to a message communicating only value-driven or only performance-driven CSR motive), H2a. attitudes will be more positive, and there will be: H2b. higher willingness to participate in a campaign; H2c. higher willingness to spread e-WOM; and H2d. increased buying intention.
We have to point out here that attitudes towards company (Grau and Folse, 2007;Ellen et al., 2006) and purchase intention (Bianchi and Bruno, 2019;Goldsmith et al., 2000;Groza et al., 2011) received attention in the earlier CSR literature. Furthermore, positive attitudes influence positively the purchase intention, as defined with the brand equity theory (Aaker, 1996;Keller, 1993). What is much needed and more interesting in the present context is to explore how CSR resonates via social media the interplay between attitudes and behaviour. Knowing that social media could change brand awareness (Colicev et al., 2018) and participating in a firm's social media is a function of the customers' attitudes (Kumar et al., 2016), it is worth looking at causal relationships. In particular, we are interested to know how attitudes influence e-WOM spread and willingness to participate in CSR campaign. A recent study exploring the effects of social media use reported that intention to develop e-WOM depends on brand image and awareness (Stojanovic et al., 2018), in line with the brand equity theory. A study in the CSR context also showed that attitudes influence purchase intention and intention to spread e-WOM (Chu and Chen, 2019). Based on the previous findings, we assume: H3. The more positive the attitudes are, the higher will be: H3a. the willingness to participate in a campaign; H3b. the willingness to spread e-WOM; and H3c. the purchase intention.
Note that initial attitudes towards company and CSR perception may change over time (Bign e, et al., 2010) and through communication . Therefore, we could expect that the above hypothesised effects might be modulated over time by further communication.

Communication type
Although combining user-and company-generated content was acknowledged in communicating CSR and in stimulating consumer engagement (Badenes-Rocha et al., 2019), social networking platforms and sites are most commonly utilised in a traditional oneway company-to-consumer communication (Saxton et al., 2019), referred to as monologue form of communication. Message in a one-way ad format was found to generating better consumer responses to the company, in comparison to a CSR message posted by another Facebook user (Kim and Xu, 2019). It was even reported that communicating CSR initiatives EJMBE internally elicited greater perceptions of trust and satisfaction towards the organisation than the external CSR message and the chief executive officer's (CEO's) personal life message (Wang and Huang, 2018).
In the digital age, however, public could not be kept away. By contrast, there is an emergent need to bring the public in and to examine the dialogic form of communication in CSR efforts. A dialogue type of communication we refer to as two-way communication between a company and its stakeholders and/or customers. Participating in dialogues on virtual platforms can lead to active stakeholder engagement (Korschun and Du, 2013) and might overcome stakeholder's scepticism towards company's CSR activities (Badenes-Rocha et al., 2019).
Furthermore, bringing the consumers in, i.e. engaging in dialogues on social media, by proactively seeking their feedback to a company's activities can encourage perceiving the CSR as more favourable (Du et al., 2010). Building favourable image (Bign e et al., 2010) and development of sustainable consumer-stakeholder's connection (Badenes-Rocha et al., 2019) is very important, as it could bring extra value to the company (Kumar et al., 2010), and is a prerequisite for building strong brands (Keller, 1993).
Therefore, the focus should be on exploring constitutive approaches to best communicate CSR. Assuming that interrogative (than narrative) message framing will have high potential to generate dialogues, we hypothesise that active stakeholder engagement on virtual platforms may emerge. CSR messages in a dialogue form indeed led to consumers' more positive attitudinal and behavioural reactions, in comparison to when a CSR message was communicated in a monologue form (Bialkova et al., 2015). Consumers were more willing to participate in a campaign, to spread e-WOM and to purchase when engaged in dialogue (than monologue) form of communication. The authors further noted that message credibility is a factor in CSR communication effectiveness (Bialkova et al., 2015).

Credibility
Information credibility is a crucial component in consumer perception, as reported at the infancy of internet communication (Flanagin and Metzger, 2000), and in adoption of e-WOM on social networks (Aghakhani et al., 2018). Therefore, it is worth paying attention to: How credibility of information reflects in a consumer's response when CSR is communicated via social media?
This investigation is especially relevant, taken that increased message credibility could lift customers trust and thus engagement (Badenes-Rocha et al., 2019). Consumers trust was further associated with brand authenticity (Hernandez-Fernandez and Lewis, 2019), which enhanced brand credibility and value perception (Del Barrio-Garcia and Prados-Peña, 2019). Moreover, corporate credibility, reflecting believability of company's intentions and communications at a particular moment in time, influences attitudes and purchase intention (Goldsmith et al., 2000). Consumers respond most positively to corporate initiatives perceived to be oriented to help others while responding negatively to a company's motivation perceived as stakeholders oriented (Ellen et al., 2006). Put differently, motives credibility perception seems to play a role in consumer response towards a company's CSR efforts. Taken the abovementioned, we expect: H4. Message credibility as well as CSR motives credibility perception will mediate the effect of attitudes on willingness to participate in a campaign (H4a), to spread e-WOM (H4b) and to purchase (H4c).
Based on the previous findings that monologue and dialogue forms of communication might have different effectiveness (Bialkova et al., 2015), and that user-and company-generated contents reflect differently consumer's engagement (Badenes-Rocha et al., 2019), we test the above hypotheses, respectively, with monologue (study 1) and dialogue (study 2) CSR messages communicated via social media. The two empirical studies are described in detail below. CSR effectiveness via social media 3. Method section 3.1 Study 1 3.1.1 Participants. In total, 120 participants (38 men) took part in the study. They were native Dutch, age between 18 and 66 years old. In total, 62% had university education. Participants were randomly assigned to one of the experimental conditions as explained hereafter.
3.1.2 Stimuli and design. A 2 (Proximity: national vs international cause) 3 3 (CSR motives: value vs performance vs value and performance) between participants design was employed.
Facebook posts have been created for each of the conditions, based on a field study exploring the actual brand behaviour on social media (Facebook and Twitter) of a well-known cosmetic brand operating internationally. Each Facebook post encompassed the brand logo, a picture, a text and a cause logo. Half of the participants were presented with a logo communicating a national cause, and the other half a logo communicating an international cause. The picture was the same for all participants, i.e. presenting an animal. The text varied between participants, namely, one-third of the participants were presented with a message communicating value, one-third with a message communicating performance and one-third with a message combining value-and performance-driven motives. In study 1, the message was in monologue (narrative) form. See Table A1 for examples on message framing. Table A2 presents a summary of the main manipulated factors.
3.1.3 Procedure. The study was conducted online, i.e. a questionnaire was distributed via e-mails and Facebook. Information about the research purpose and a consent form were provided prior to the study.
Participants were first presented with the Facebook post. Afterwards, they had to answer questions addressing the message effectiveness in terms of attitude towards the company, willingness to participate in CSR campaign, spreading electronically word of mouth (e-WOM), purchase intention. Perceived message credibility and perceived CSR motivation were also measured. The final part of the survey encompassed questions addressing demographics.
3.1.4 Instrument. Scales from previous relevant work were adopted for the current context, to measure the constructs of interest (see Table 1 for summary of the constructs used, and Table A3 for details). The scale for attitude towards the company encompassed three items and was derived from MacKenzie and Lutz (1989). An example of an item is "My attitude towards the company is" (1 5 unfavourable, 7 5 favourable). The scale for willingness to participate in the CSR campaign had three items, used by Grau and Folse (2007) (1989), Zeithaml et al. (1996), Grau and Folse (2007), Flanagin and Metzger (2000), Groza et al. (2011), Ellen et al. (2006; See Appendix, Table A3 for details

EJMBE
in the context of cause-related marketing. An example of an item is "I would be willing to participate in this campaign" (1 5 strongly disagree, 7 5 strongly agree). The willingness to spread e-WOM encompassed three items, e.g. "How likely is it that you say positive things online about the company to other people?" (1 5 not at all likely, 7 5 extremely likely). The scale was adopted from Zeithaml et al. (1996) for the current context of e-WOM. The scale for purchase intention encompassed four items, e.g. "What is the likelihood to purchase a product from the company" (1 5 very unlikely, 7 5 very likely). The scale was previously used by Groza et al. (2011) to study consumers' response to CSR. The scale for perceived credibility of the message included four items and was derived from Flanagin and Metzger (2000). An example of an item is "The information provided in the message was" (1 5 not at all believable, 7 5 extremely believable). The scale for perceived CSR motivation included eight items and was adopted from Ellen et al. (2006). An example of an item is "The company is making the offer because feels morally obliged to help" (1 5 strongly disagree, 7 5 strongly agree). All scales used had high internal validity, all Cronbach's alpha > 0.70 (Table 1). T-tests were conducted to check that all dependent variables are significantly different from neutral.
3.1.5 Analytical procedure. T-tests were conducted to probe for the proximity effect. ANOVAs were conducted to test the CSR motives impact. Regression models were performed to explore the interplay between attitudes, consumers' willingness to participate in CSR campaign, to spread e-WOM, to purchase. Further regression modelling tested whether and how perceived credibility of the message and perceived CSR motivation interplay in changing attitudes and behaviour responses. In the following, we present only the results substantiated statistically.
3.2 Results (study 1) 3.2.1 Cause proximity. Cause proximity (national vs international) influenced the willingness to participate in a CSR campaign (see Figure 1, left panel). Participants were more likely to participate in a campaign with national than international cause (M 5 4.93 vs M 5 3.39), as revealed by a significant T-test outcome, t(118) 5 5.38, p < 0.0001. Attitudes towards the brand (M 5 5.01), purchase intention (M 5 4.27) and willingness to spread e-WOM (M 5 3.49) were relatively high. Cause proximity, however, did not modulate these parameters, all p-values > 0.1.
3.2.2 Corporate social responsibility motives. Participants were most likely to spread e-WOM when the messages communicated performance and value (M 5 4.23) than only value (M 5 3.15) or only performance-driven motives (M 5 3.08), F(2, 117) 5 6.68, p < 0.005. Perceived message credibility as well as perceived CSR motivation were also highest when participants saw a performance-and value-driven motives message, p-values < 0.01. There was a tendency for most positive attitude towards the brand when the message combined performance and value (M 5 5.44), than communicating only one of these motives (respectively M 5 4.80 and M 5 4.76). Purchase intention was also highest when the message communicated performance and value (M 5 4.68) than only value-(M 5 4.18) or only performance-driven motives (M 5 3.96). However, none of these effects was substantiated statistically, all p-values > 0.1. Figure 2 (left panel) provides an overview on the effects of CSR motives.

Note(s): NB. With asterisk, significant differences
3.2.3 Credibility. How perceived message credibility and CSR motives credibility can influence consumers' responses was investigated. Regression models explored the interplay between attitudes and behaviour reactions in terms of campaign participation, willingness to spread e-WOM and to purchase. The modelling is conducted following Baron and Kenny's (1986) definition of mediators' influences. Therefore, consequent analyses are done to test: (1) the effect of message and CSR credibility on attitude; (2) the effect of message and CSR credibility on campaign participation, purchase intention, willingness to spread e-WOM; (3) the effect of attitudes on campaign participation, purchase intention, willingness to spread e-WOM; (4) the effect of message and CSR credibility on campaign participation, purchase intention, e-WOM, when including attitude as a mediator. Table 2 summarises the outcomes of model 4 (for study 1), and details on statistics for other models are presented below.
Message credibility explained 65% of the variance in attitudes. The model testing the effect of message credibility and CSR motivation on attitudes also explained 65% of the variance, R 2 5 0.65, F(2, 117) 5 109.60, p < 0.0001.
The model encompassing attitudes, message credibility and perception of CSR motives had a relatively low explanatory power concerning the campaign participation. The models concerning the willingness to spread e-WOM and the purchase intention had a good explanatory power, see Table 2 for exact statistical outcomes.
3.3 Study 2 3.3.1 Design, procedure, instrument were the same as in study 1. The only difference concerned the type of communication, i.e. in study 2, the message was in a dialogue (interrogative) form, see Table A1.
3.3.2 Participants. In total, 120 participants (43 men) took part in the study. They were native Dutch, age between 18 and 64 years old. And, 66% had university education. Participants were randomly assigned to one of the experimental conditions. 3.4.2 Corporate social responsibility motives. CSR motives played a major role in consumers' response (see Figure 2, right panel). Attitude towards the brand was most positive when the message communicated performance and value (M 5 5.99) than only value-(M 5 4.95) or only performance-driven motives (M 5 4.53), F(2, 117) 5 15.99, p < 0.0001. Purchase intention was also highest when the message communicated performance and value (M 5 5.53) than only value-(M 5 4.57) or only performance-driven motives (M 5 4.06), F(2, 117) 5 13.62, p < 0.0001. Participants were most likely to spread e-WOM when presented with a message communicating performance and value combined (M 5 4.84) than with a message communicating one of these motives separately (respectively M 5 3.52 vs. M 5 3.17), F(2, 117) 5 13.26, p < 0.0001.
3.4.3 Credibility. The same modelling procedure as in study 1 was followed for study 2.
Message credibility explained 37% of the variance in attitudes. Adding perception of CSR motivation to the model increased the explanatory power to 46%, R 2 5 0.46, F(2, 117) 5 49.35, p < 0.0001.
The regression modelling further reported that the higher the perceived message credibility (β 5 0.27) and perceived CSR motivations were (β 5 0.39), the higher was the purchase intention, R 2 5 0.36, F(2, 117) 5 33.40, p < 0.0001. Willingness to spread e-WOM also increased when the message (β 5 0.41) and the CSR motivations (β 5 0.23) were perceived as more credible, R 2 5 0.35, F(2, 117) 5 31.55, p < 0.0001. Concerning the campaign participation, the message credibility explained only 8% of the variance in the model. Adding perception of CSR motivation to the model did not change its explanatory power.
The more positive the attitude towards the company was, the higher was the willingness to take part in the company's CSR campaign (R 2 5 0.12, F(1, 118) 5 15.53, p < 0.005); to spread CSR effectiveness via social media e-WOM (R 2 5 0.27, F(1, 118) 5 43.15, p < 0.0001); and the purchase intention (R 2 5 0.49, F(1, 118) 5 117.51, p < 0.001). The model encompassing attitudes, message credibility and perception of CSR motives had a relatively low explanatory power concerning the campaign participation. The models concerning the willingness to spread e-WOM and the purchase intention had a good explanatory power, see Table 3 for exact statistical outcomes.

General discussion
The goal of the present paper was to understand: How CSR messages resonate via social media, and thus, how to optimise the CSR effectiveness via social media? Addressing this question, we complement the existing body of literature bringing together theories of CSR motives (Groza et al., 2011;Ellen et al., 2006), cause proximity effect (Grau and Folse, 2007;Groza et al., 2011), communication type (Du et al., 2010;Korschun and Du, 2013), credibility (Flanagin and Metzger, 2000;Goldsmith et al., 2000) and The results are clear in showing that cause proximity, CSR motives and communication type are crucial determinants for consumers' attitudes and behaviour reactions in terms of willingness to participate in a campaign, to spread e-WOM and to purchase (see Figure 3 for a model summary). Message credibility and CSR motives credibility perception further modulated consumers response, as described in detail below.

Cause proximity
Willingness to participate in a CSR campaign was higher when national than international cause was launched (H1b confirmed), and this effect was pronounced for both, monologue (study 1) and dialogue message framing (study 2). Cause proximity, however, did not modulate neither attitudes (H1a rejected), neither willingness to spread e-WOM (H1c rejected) nor purchase intention (H1d rejected), see Figure 1.
Note that previous studies were not univocal on the effect of proximity. While hypothesised proximity effect was not significant on attitudes in some earlier studies Campaign participation e-WOM Purchase intention R 2 5 0.123 R 2 5 0.377  Table 3. Summary of the statistics for model 4 (study2) Figure 3. Conceptual framework EJMBE (Ross et al., 1992), later it was shown (Grau and Folse, 2007;Groza et al., 2011) that cause proximity modulates the attitudes, consumer identification with the cause and thus the corporate image evaluation. A plausible explanation for differences in findings could be the difference in manipulated factors, e.g. cause type, CSR motives, as well as the brand perception. Another explanation could be the nature of research itself, i.e. while the abovementioned studies explored the effect of the manipulated factors without implementing media, hereby we specifically explore the role of social media in CSR effectiveness. Further explanation could be that consumers have favourable attitudes towards the company (Bhattacharya and Sen, 2004) and/or social media positively influenced the brand image (Stojanovic et al., 2018). If this is the case, one could expect that any type of CSR activity is positively evaluated, irrespective of cause proximity. Hereby, consumers indeed reported very positive attitudes towards the brand, and relatively high purchase intention, irrespective of cause proximity.
Taken the abovementioned, we could say that proximity modulates behaviour responses, e.g. campaign participation, willingness to spread e-WOM, purchase intention with different magnitude. This is very interesting finding, opening new avenues to explore proximity effect, also when combined with other cause attributes.

Corporate social responsibility motives
Campaign participation did not differ with respect to CSR motives, opposite to our expectation in H2b. This is another important outcome, taken that, in general, our respondents were very much interested to participate in the campaign. It might be the case that people would like to take part in the campaign as they consider that CSR means "doing good", and thus reflecting the moral aspects of behaviour. A recent study found that people indeed are very willing to behave in environmentally and socially responsible manner (Hosta and Zabkar, 2020).
Spreading e-WOM and purchasing, by contrast, are not necessarily associated with moral, but rather with financial aspects as related to marketing objectives. Willingness to spread e-WOM hereby was highest when message combined value-and performance-driven motives (H2c confirmed), and this effect appeared for both monologue and dialogue type of communication. Attitudes (in line with H2a) and purchase intention (in line with H2d) also changed, being highest when message combined value-and performance-driven motives. However, these effects were pronounced only for dialogue type of communication. This is a very important outcome, showing that CSR messages not just resonate via social media (Saxton et al., 2019), but have different resonance as a function of CSR motives and communication format. The current results nicely cohere with other studies that consumer response may change over time (Bign e, et al., 2010) and through communication .
Furthermore, the regression modelling reported a strong relation between attitudes and behavioural consumers' response (in line with TPB, Ajzen, 1991). The more positive the attitudes were, the higher was the willingness to spread e-WOM (H3b supported) and to purchase (H3c supported). These findings provide a direct link between attitudes towards company (Grau and Folse, 2007;Ellen et al., 2006), purchase intention (Bianchi and Bruno, 2019;Goldsmith et al., 2000;Groza et al., 2011), spreading e-WOM (Chu and Chen, 2019;Stojanovic et al., 2018) and thus being essential building blocks in our framework (Figure 3).
We have to also point out here that the more positive the attitudes were, the higher was the willingness to participate in a campaign (in line with H3a). Note, however, that the positive correlation between attitudes and campaign participation was only pronounced for dialogue message framing. This is another important finding, confirming the crucial role of communication type in optimising the CSR effectiveness.

Communication and credibility
Monologue vs dialogue message framing resonated differently on CSR effectiveness when it comes to attitudes and purchase intention (Figure 2). The current findings cohere with previous research that dialogue form of communication enhances attitudes and thus increases purchase intention (Bialkova et al., 2015). A plausible explanation for the positive effect of dialogue could be the better stakeholders (Korschun and Du, 2013) and consumers' engagement (Badenes-Rocha et al., 2019), in comparison to one-way communication.
A further explanation for the more favourable CSR perception when consumers being engaged in a dialogue could be found in the credibility effect. Information credibility is a crucial component of consumer perception in internet communication (Flanagin and Metzger, 2000) and in adoption of e-WOM on social networks (Aghakhani et al., 2018).
Hereby, message credibility was higher with dialogue than monologue communication type, as substantiated statistically with additional T-test run across both studies t(238) 5 2.10, p < 0.05. Furthermore, the more credible the message was perceived, the more positive were the attitudes. These outcomes are in line with previous research that increased message credibility could lift customers trust and engagement (Badenes-Rocha et al., 2019), improve brand perceived value (del Barrio-Garcia and Prados-Peña, 2019; Hernandez-Fernandez and Lewis, 2019) and enhance attitudes (Bialkova et al., 2015).
Interestingly, however, hereby in monologue type of communication message credibility explained the total effect on attitudes, while in dialogue type of communication attitude's formation was modulated by message credibility and CSR motives credibility perception. This is another important contribution in understanding how CSR resonates via social media. Namely, in the one-way communication, message credibility is the fundament. In the two-way communication, engaging consumers on social media, by proactively seeking their feedback seems to overcome their scepticism towards message credibility and thus can enhance the CSR credibility perception. The current findings nicely support previous work that engaging consumers in virtual dialogues lead to perceiving the CSR as more favourable (Du et al., 2010) and thus to building a favourable brand image (Bign e, et al., 2010).
Moreover, message credibility as well as CSR motives credibility perception mediated the effect of attitudes on willingness to spread e-WOM (supporting H4b). Note, however, this effect was with different magnitude for monologue (Table 2) and dialogue (Table 3) message framing. CSR motives credibility perception mediated the effect of attitudes on purchase (H4c supported partially), and this effect was pronounced for both, monologue and dialogue message framing. A slightly different look at these data (Tables 2 and 3) show that message credibility mediated the effect of attitudes on e-WOM, while CSR credibility perception mediated the effect of attitudes on purchase intention. These outcomes are new and crucial in understanding the way CSR resonates via social media. They are a cornerstone to build around, to satisfy the increasing demand for transparency and the growing consumers' expectations regarding CSR.

Managerial implications and future research
Any brand wants to know what consumers think about the brand, what influences costumers' attitudes and behaviour toward the brand. Consumers' attitudes and behaviour, however, could change as a function of CSR communicated via social media, as the present study shows. The current outcomes could be taken by managers to: First, carefully select and honestly communicate their CSR motives. Combining value-and performance-driven CSR motives (than communicating only value or only performance) led to favourable attitudes and highest purchase intention, and this effect was better pronounced with dialogue message framing. Thus, it seems that consumers can accept that a firm is profit oriented. However, this performance-driven motivation should be well framed.

EJMBE
What seems to be much more important is the believability of a company's intention and communication at a particular moment in time (as recognised a while ago, e.g. Becker-Olsen et al., 2006;Goldsmith et al., 2000). In this respect, dialogue type of communication provides opportunity to directly address the company's CSR initiatives, and thus to rise transparency.
In this respect, the second recommendation: brands should prioritise the type of communication, i.e. brining public in, involving in dialogue, especially taken the digital era we are living in. Such dialogue type of communicating CSR could stimulate consumers' engagement (Badenes-Rocha et al., 2019), but also will increase the message credibility perception, as demonstrated hereby. A further enhancement of consumers-brand engagement could possibly be achieved when launching communications via different channels, i.e. a follow-up research could compare the role of Facebook vs Twitter, in CSR effectiveness. And, another study could even explore different platforms effect (e.g. digital vs augmented reality vs virtual reality). This might be a challenging task, but will bring extra value, if incorporated in the portfolio of socially responsible brands.
Third, as the cause proximity increased the willingness to participate in a CSR campaign, alliance between a cause and consumers should be soughed in CSR initiatives. Although previous studies reported that cause proximity modulates the attitudes (Grau and Folse, 2007;Groza et al., 2011), hereby we have not confirmed such relationship. Note that the current study explored the CSR effectiveness via social media, while abovementioned studies did not look at media impact. Also, the study was conducted in The Netherlands, a country with traditions in social initiatives. Thus, it would be interesting to conduct a follow-up study in a country where CSR initiatives are not that popular. A further cross-cultural comparison (e.g. European Union (EU) vs USA vs emerging markets) could explore how CSR resonates via social media, depending on consumers' nationality and culture.
Last, but not least, the message credibility should be guaranteed. Information credibility is a crucial component in consumer perception (Flanagin and Metzger, 2000), and in adoption of e-WOM on social networks (Aghakhani et al., 2018). As demonstrated hereby, the more credible the message was perceived the more positive the attitudes towards the company were, and the higher was the willing to spread e-WOM.
Therefore, by incorporating appropriate communication strategies, a CSR initiative might reflect in spreading favourable e-WOM, increased campaign participation and purchase grow. Such positive outcomes of CSR activities could build long-term consumerstakeholder's connection. As known, sustainable consumer-stakeholder's connections (Badenes-Rocha et al., 2019) bring extra value to the company (Kumar et al., 2010) and predetermine brand success (Keller, 1993).

Conclusions
Addressing the change of customer-brand interactions in the digital era and the increased demand for transparency in CSR communication, the current paper explored how CSR resonates via social media, and thus providing understanding on how to optimise the CSR effectiveness via social media.
Based on a profound literature review and a field research we did on the actual social media behaviour of socially responsible brands, an online survey was developed. In particular, consumers' attitudes and behavioural reactions in terms of willingness to participate in a campaign, to spread e-WOM and to purchase were explored.
The results are clear in showing that: (1) cause proximity enhanced the campaign participation, and this effect was pronounced for both, monologue and dialogue message framing.
(2) Combining value-and performance-driven motives increased the willingness to spread e-WOM, and this effect appeared for both, monologue and dialogue communication. (3) Attitudes and purchase intention were highest when value-and performance-driven motives CSR effectiveness via social media were communicated together, but these effects appeared only when the message was in a dialogue framing. (4) Message credibility and CSR credibility perception mediated the effect of attitudes on willingness to spread e-WOM, and to purchase, and these effects had different magnitude as a function of communication type. (5) The more credible the message was perceived the more positive the attitudes towards the company were, and the higher was the willingness to spread e-WOM. (6) The more credible the CSR motives were perceived, the more positive the attitudes towards the company were, and the higher was the purchase intention. In sum, the parameters emerging hereby as key drivers of message optimisation provide insights on how to be most appropriately combined to enhance the CSR impact. The current outcomes could be directly implemented in a company's social media campaigns to amplify values and maximise business returns to CSR. Purchase intention What is the likelihood to purchase a product from X Groza et al. (2011) in the CSR context 1 5 very unlikely -7 5 very likely 1 5 improbable -7 5 probable 1 5 impossible -7 5 possible 1 5 low intent to purchase -7 5 high intent to purchase Message credibility The information provided in the message was Flanagin and Metzger (2000) 1 5 not at all believable -7 5 extremely believable 1 5 not at all accurate -7 5 extremely accurate 1 5 not at all trustworthy -7 5 extremely trustworthy 1 5 not at all biased -7 5 extremely biased 1 5 not at all complete -7 5 extremely complete CSR motives perception X is making the offer because Ellen et al. (2006) They feel morally obliged to help They have a long-term interest in the community Their owners of employees believe in this cause They want to make it easier for consumers who care about the cause to support it They are trying to give something back to the community They will get more customers by making this offer They will keep more of their customers by making this offer They hope to increase profits by making this offer 1 5 strongly disagree -7 5 strongly agree