Digital content marketing in B2B contextsthe drivers of engagement and electronic word of mouth communication at international trade shows

Resumen

PURPOSE The potential that social networks represent as a marketing tool is unquestionable, they allow sharing information, building networks, and influencing consumers and brands. Furthermore, social networks contribute to business-to-business (B2B for its acronym in English) engagement, offering a perfect environment for its development. In this sense, the marketing literature has recognized the usefulness of social networks in the promotion of B2B international trade shows (ITSs for its acronym in English). Additionally, in times of crisis such as COVID-19, communication is one of the main challenges in emergency management, and social networks are a key issue in this context. For the B2B marketing sphere, the international trade shows are essential and relevant events for the development of engagement, recognized as ideal vehicles to present sales opportunities to companies, strengthen contacts, maintain brand image, and access new markets. Additionally, they are valuable for establishing relationships in new countries. Despite the relevance of these events, there is a call for more research in B2B contexts, specifically studies that use analytics to reach significant conclusions. In social networks, on the one hand, engagement is an essential construct for his research, given the influential nature of these networks. From a general view, engagement is formed by three components: cognitive, emotional, and behavioral, which can be translated into the virtual context under other denominations. This perspective allows us to analyze 'virtual' engagement as a multidimensional construct made up of commitment, virality, and popularity. On the other hand, electronic word of mouth (eWOM for its acronym in English) is a marketing construct that influences consumer behavior and is a result of the communication process in social networks. This construct has recently been defined in the literature as consumer-generated communication that uses digital tools and is primarily directed at other consumers. eWOM communication allows multidirectional exchanges of information between communicators and receivers, with an unprecedented speed of diffusion. In social networks, the attitude towards information is one of the key drivers of eWOM. Thus, in this context, where digital content marketing creates, communicates, distributes, and exchanges content that has value to stakeholders, companies use social media to generate positive marketing results. In this sense, the literature has pointed out that social networks are a topic that requires continuous academic research, given the power they have. However, there is a limited number of investigations that empirically study the adoption, use and effect of social networks in the context described. Considering engagement and eWOM as results of the online communication process in social network contexts, this doctoral thesis aims to discover their predictors in international B2B contexts. At the same time, it hopes to contribute theoretically and empirically to scientific knowledge in the field of digital content marketing, providing findings related to the adoption, use and effect of social networks, and contributing to a greater understanding of how they can be used to promote engagement and eWOM. In this context, the general objective of this doctoral thesis is "to identify the drivers of engagement and electronic word of mouth communication in digital content marketing in international B2B contexts". To achieve the general objective, on the one hand, three global goals are proposed. First, to identify which are the key variables that drive engagement and eWOM in international B2B contexts, second, to study the influence of these variables on engagement and eWOM, and third, and last, to single out the managerial implications. Along with this, a research strategy is proposed to achieve the general objective and the three global goals. The research strategy comprises three studies with specific objectives. With the previous premises and the review of the literature, the variables to be considered in the studies are proposed. The dependent variables analyzed were engagement and volume of eWOM. The independent variables analyzed were value co-creation practices, content strategies, message characteristics, time frame, presence of emojis, number of emojis, and countries. More specifically: First, engagement is an essential construct in social media research. According to the literature, more studies are needed to analyze the relationship between engagement and joint value creation in B2B contexts. Thus, engagement was selected as the dependent variable to be explained in the first study. Engagement is made up of three components: cognitive, emotional, and behavioral. This perspective allows us to divide engagement into a multidimensional construct formed by popularity, commitment, and virality. These dimensions were operationalized according to the following criteria: (i) popularity, measured through the number of favorites in the tweets; (ii) commitment, measured through the number of comments on the tweets; and (iii) virality, measured through the number of retweets in the publications. Second, eWOM is a marketing construct and is one of the results of the communication process in social networks such as Twitter, which influences consumer behavior. Furthermore, the literature highlights the eWOM's ability to interact with multiple cultures and societies. Specifically, in social networks, the attitude towards information is one of the key drivers of eWOM that influences the purchase intentions of consumers. Furthermore, some authors examined the impact of eWOM on sales and concluded that eWOM volume has a stronger impact than eWOM valence. Thus, the volume of eWOM was selected as the variable to explain, in the second and third studies. The eWOM was operationalized through the volume of retweets on Twitter. Third, value co-creation practices have been mentioned in the literature as an emerging topic. To date, there are few studies on the relationship between value co-creation practices and engagement in social networks. The research calls are related to the need for more research especially in the B2B context. Given the above, the first study analyzed the extent to which value co-creation practices in social networks influence engagement in a B2B context. Thus, value co-creation practices were selected as the independent variables of the first study. As people and brands interact and engage, co-creating value practices are generated, and their experiences become the new basis for co-creating value. The process of joint creation of value is facilitated by social networks because they empower people outside of companies, offering a perfect setting for the development of co-creation of value. More specifically, on the one hand, the linking practices defined in the literature are the practices of co-diagnosis, co-ideation and co-evaluation: (i) co-diagnosis practices are related to collecting and organizing information for collaborative use (in the first study there were no tweets classified in this category); (ii) co-ideation practices are related to generating and suggesting ideas (communicating, sharing, and involving the generation and suggestion of ideas); and (iii) co-evaluation practices are related to commenting and selecting ideas. On the other hand, the materializing practices defined in the literature are co-design, co-testing and co-launching practices: (i) co-design practices are aimed at developing concepts and knowledge; (ii) the co-testing practices are aimed at testing prototypes, improving the offer, and giving feedback; and (iii) the co-launching practices are aimed at creating and managing information, advertising, marketing and dissemination of information. Fourth, social media content strategies play a key role in promoting eWOM. The research calls are related to the need to know the most effective content strategies with respect to different platforms and industries. It is highlighted that it is relevant for companies to develop digital content marketing strategies, considering that B2B buyers depend on internet information in the first stage of the purchase process. Thus, social media content strategies were selected as one of the independent variables of the second study. Given the above, the second study looked at the extent to which content strategies influence eWOM volume in a B2B context. According to the literature review, social media content strategies can be classified into (i) informative strategies, (ii) transformational strategies, (iii) interactive strategies, and (iv) promotional strategies. More specifically: the (i) informative strategies are related to: (a) product attribute themes, and (b) product or service knowledge themes; (ii) transformational strategies are related to: (a) sentimental themes, (b) brand attachment themes, and (c) experiential themes; (iii) interactive strategies are related to: (a) consumer participation themes, and (b) customer relationship themes; and (iv) promotional strategies are related to: (a) non-monetary promotion themes, and (b) monetary promotion themes. Fifth, the social media message features allow us to analyze the content of the messages of the social networks. In this sense, the research calls are related to the type of content (textual, image/video based) that consumers prefer in particular contexts. Additionally, more research has been called for in the literature on the impact of social media message features on engagement, across various social media platforms and contexts. Thus, the features of the message were selected as one of the independent variables of the second study. Given the above, the second study looked at the extent to which message features influence eWOM volume in a B2B context. According to the literature review, the characteristics of social network messages can be classified into: (i) interactive features, (ii) textual features, and (iii) visual features. More specifically, the (i) interactive features are related to: (a) hashtags, and (b) mentions; (ii) the textual features are related to: (a) the retweet request, and (b) the presence of the word 'please'; and (iii) the visual features are related to: (a) the photo, and (b) the video. Sixth, the time frame has been defined in the literature as the moment messages are posted. This variable is related to the relevance of social networks in crisis management when communication is a challenge. COVID-19 is a unique case due to the speed with which it has spread throughout the world and the havoc it has caused in the international business context. The research calls are related to the need for research on the impacts of COVID-19 in the international business context, and in general theoretical and empirical research on the impact of COVID-19. Thus, the time frame was selected as one of the independent variables in the second and third studies. Given the above, the second and third studies analyzed to what extent the time frame in social networks influences the volume of eWOM in a B2B context. Seventh, the presence of emojis plays an important role in social interactions, both in initiating and maintaining an interaction, and can bring clarity to text messages, reducing ambiguity. Also, specific emojis can be used with dissonant customers. Emojis are visual symbols and static forms, with their own Unicode that include facial expressions, emotions, plants, animals, activities, abstract concepts, objects, among others. Emojis are used frequently, they increase engagement as a visual stimulus, they complement or replace written language, as a language that transcends borders and is the fastest growing language globally. The literature calls for more research on the use of emojis in real-world communication in different contexts. Thus, the presence of emojis was selected as one of the independent variables of the third study, therefore, the presence or absence of emojis in each tweet was coded. Given the above, the third study analyzed to what extent the presence of emojis in tweets influenced the volume of eWOM in a B2B context. Eighth, a greater number of emojis increases engagement as a visual stimulus. The number of emojis is related to purchase intent, volume of eWOM on Twitter, and attention paid to a message. From a general point of view, it has been found that a greater number of emojis increases engagement as a visual stimulus. The research call described in the literature is related to the need for more research on the effects that different amounts of emojis can have in a message, given that the amount of emojis positively influences the volume of eWOM on Twitter. Thus, the number of emojis was selected as one of the independent variables of the third study. Given the above, the third study analyzed to what extent the number of emojis in messages on social networks influences the volume of eWOM in a B2B context. Nineth, the countries were considered as a variable of interest because in the field of business, culture could influence the use of social networks by companies. In addition, emojis as a language are a global practice that transcends borders. Several authors have discussed the eWOM's ability to interact with multiple cultures and societies. The research call is related to the identification of differences between countries in the use of social networks. Thus, countries were selected as one of the independent variables of the third study. Given the above, the third study looked at whether there are differences between countries in the presence of emojis, the number of emojis, the time frame, and the volume of eWOM, in a B2B context. Considering the afore mentioned, this doctoral thesis through its research strategy addresses the general objective, the variables described above, and the research calls described in the literature, through three empirical studies that address the following specific objectives: The objectives of the first study focused on identifying: (1) if linking practices (co-diagnosis, co-ideation, and co-evaluation) influence engagement; (2) if materializing practices (co-design, joint testing, and co-launch) influence engagement; and (3) if there are differences between linking practices and materializing practices regarding the production of engagement, in an international B2B context. The objectives of the second study focused on identifying: (1) if social media content strategies influence eWOM volume; (2) if social media message features influence eWOM volume; and (3) if the time frame (Ex-Ante COVID-19 publication date or Ex-Post COVID-19 publication date) influences eWOM volume, in an international B2B context. The objectives of the third study focused on identifying: (1) if the emoji presence, the number of emojis, and the time frame (Ex-Ante COVID-19 publication date or Ex-Post COVID-19 publication date) influence eWOM volume; and (2) if there are differences between countries in terms of emoji presence, number of emojis, time frame, and volume of eWOM, in international B2B contexts in 5 countries. METHOD In general terms, to achieve the stated objectives, the methodology mainly considered: (1) the use of a non-experimental methodology; (2) the social network Twitter and the tweet as the unit of analysis; (3) as a B2B context, international trade shows specializing in food and beverages; (4) content analysis to understand communication on Twitter; (5) the Kolmogorov-Smirnov Test to verify the distribution of the data; (6) the Kruskal Wallis Test to verify if there are significant differences between the data groups; and (7) the Optimal Scaling Regression Analysis (CATREG) for each dependent variable, to test the hypotheses considering a statistical significance level of p-value <0.05 (α) and a 95% confidence interval. Specifically, to study the influence of predictor variables on engagement and eWOM volume, a regression analysis with optimal scaling was performed using the CATREG algorithm. This methodology allows quantifying categorical variables and optimizing the multiple regression coefficients in a regression model with categorical data. The CATREG algorithm was created by the Data Theory Scaling System Group (DTSS). This method has been used successfully to perform categorical regression analysis on marketing data before. In general, there are some advantages to using CATREG models. For example, they do not need to assume normality of the residuals, nor do they require that the data be organized as interval-scaled variables, nor do they need to be linearly related, or normally distributed. Before interpreting the results, intercorrelations between predictors were evaluated for both untransformed and transformed predictors. Thus, the three studies analyzed the following samples: The first study analyzed a sample of 1,608 tweets published during the year 2019, compiled from the Twitter account "@Specialty_Food" belonging to the company "Specialty Food Association" (SFA) that owns and organizes the "Fancy Food Show", which are ranked as the top five food and beverage fairs in the United States. The research period ran from January 1st, 2019, to December 31, 2019, analyzing 1,608 tweets. Data were collected through keyhole.co. an accessible freeware in its basic version. The second study analyzed a sample of 3,482 tweets from the same Twitter account “@Specialty_Food” belonging to the company “Specialty Food Association” (SFA), owner and organizer of the “Fancy Food Show”. The research period comprised two periods: period one (Ex-Ante COVID-19) from January 1st, 2019, to March 10, 2020, analyzing 1,930 tweets, and period two (Ex-Post COVID-19) from March 11, 2020, to August 31, 2021, analyzing 1,552 tweets. COVID-19 was declared a pandemic on March 11, 2020, by the World Health Organization. A third-party service known as TrackMyHashtag was used to collect the data. The third study analyzed a sample of 9,329 tweets from the Twitter accounts of 10 international trade shows specializing in food and beverages, located in 5 countries (Spain, the United States, France, Mexico, and the United Kingdom). The research period comprised 2 periods: period one (Ex-Ante COVID-19) from June 21, 2019, to March 10, 2020, analyzing 3,566 tweets, and period two (Ex-Post COVID-19) from March 11, 2020, to January 31, 2022, analyzing 5,763 tweets. For data collection, a third-party service called Twitonomy was used. CONCLUSIONS The conclusions of the first study highlight the relevance of value co-creation practices (co-ideation, co-evaluation, co-design, co-testing and co-launching) in the production of engagement in a social network such as Twitter. The findings show that value co-creation practices positively influence engagement and its dimensions (popularity, commitment, virality, and global engagement) in B2B contexts. In addition, the study confirms the existence of differences between linking practices and materializing practices in terms of the production of engagement. Likewise, the study contributes to the understanding of value co-creation practices in the context of social networks by providing answers related to the identification, the use of these practices by B2B actors and the effect on engagement on Twitter. Overall, these findings demonstrate the usefulness of value co-creation practices to drive engagement in virtual B2B contexts. The conclusions of the second study highlight, on the one hand, the effectiveness of social media content strategies to predict the volume of eWOM. The most relevant are informative strategies and transformational strategies, compared to interactive and promotional strategies. More specifically, informative strategies increase the volume of eWOM (specifically the use of product attributes and the non-use of themes related to product knowledge, the latter related to the ways of operating the product or service); transformational strategies (specifically the use of experiential themes and the non-use of sentimental themes); and interactional strategies (specifically the non-use of customer relationship themes, that is, tweets that request responses and comments from customers about products and services). On the other hand, regarding the messages features we can conclude that the use of videos, mentions and photos are useful to predict the volume of eWOM. Specifically, video has the most power, followed by mentions and photos. Hashtags, URL links, retweet request and use of “please” are not predictors of eWOM volume. Finally, regarding the time frame, we can say that it influences the volume of eWOM. Tweets published before the COVID-19 pandemic positively influence eWOM volume in this study. In the conclusions of the third study, we can highlight that the emoji presence, the number of emojis and the time frame explain the volume of eWOM on Twitter in the sample that included the five countries together (n = 9,329), which supports the effectiveness of emojis in the social network Twitter, in B2B communications. Regarding the differences between countries, the relationships were not always significant in the subsamples (France, Mexico, Spain, United Kingdom, United States). From a general point of view, the findings confirm the existence of differences between countries in terms of the emoji presence, the number of emojis, the time frame and the volume of eWOM on Twitter. Therefore, the effect that the predictor variables have on the volume of eWOM is significantly moderated by the country. More specifically, we can conclude the following: In the sample from the United States, alone, all predictors (emoji presence, number of emojis, and time frame) were significant in eWOM volume production, although the US sample had the lower presence of emojis (only 21% of the tweets had emojis), in relation to the other countries. For the French sample, two predictors significantly influenced eWOM volume, number of emojis, and time frame. However, for the Spanish and UK samples, only one predictor influenced eWOM volume, number of emojis, and time frame, respectively. Regarding the time frame, the behavior of France and the United States was similar; that is, tweets published before COVID-19 increased the volume of eWOM. However, in the British sample the trend is stronger and contrary to other countries (tweets published Ex-post COVID-19 increase the volume of eWOM). Overall, these findings demonstrate the usefulness of emojis and time frame in driving eWOM volume in B2B contexts. In the case of the Mexican sample, no predictor was significant. MANAGERIAL IMPLICATIONS In the first study, the managerial implications allow us to suggest to marketing directors and managers of B2B companies that organize fairs or virtual B2B events, the use of co-design, co-evaluation and co-ideation practices, prioritizing them over others, given that are powerful motivators of engagement and its dimensions on Twitter. It is suggested to generate publications that are framed in value creation practices of co-design, co-evaluation and co-ideation. It is recommended, in order of importance, to promote co-creation practices aimed at: (i) developing concepts and knowledge in face-to-face and online events (co-design practices); (ii) share valuable business ideas (co-evaluation practices); and (iii) share business cases (co-ideation practices). From a general point of view, to generate more engagement on social networks in B2B contexts, it is recommended to prioritize publications that incorporate events based on collaborative and dynamic human interactions. Research has shown that, in the case of Twitter, tweets posting live events and online events, where concepts and insights are developed (co-design practices), produce the most engagement. In the second study, the managerial implications allow us to suggest the following: First, since the results highlight the relevance of the use of the experiential themes and the product attribute themes to predict the volume of eWOM, it is recommended that the marketing managers of ITSs that prioritize these strategies on Twitter. On the other hand, it is not recommended to prioritize strategies that involve product knowledge themes and sentimental themes, as well as messages based on themes of brand attachment, consumer participation, non-monetary and monetary promotions, since all of them they did not show to be relevant to predict the volume of eWOM. Second, since the most important features in social media messaging are visuals and mentions, we recommend marketing managers prioritize videos, mentions, and photos in social media communication, since these features can positively influence the volume of eWOM. Third, it is important to understand the time frame in which the communication occurs, since this influences the volume of eWOM. In this study, only tweets published before the COVID-19 pandemic positively influence eWOM volume. It is inferred that this result comes from the three cancellations of face-to-face ITS in the ex-post COVID-19 period, developing virtual versions, as a new way in the B2B context. The study has shown that crises affect the shared digital context in different ways, therefore, the research suggests deepening the analysis of digital communication in times of crisis. In the third study, the managerial implications allow us to suggest that B2B companies integrate emojis in social media communications to boost digital marketing content through eWOM volume. The study suggests the use of emojis in the field of international business. Likewise, the research suggests, in the case of the United States, to analyze other industries to explore if a greater number of emojis results in a greater volume of eWOM. In relation to the time frame, crises have been shown to affect shared digital content in different ways. Therefore, the study suggests deepening the analysis of digital communication in times of crisis. In this study, only tweets published before the COVID-19 pandemic significantly and positively influenced eWOM volume. It is suggested that in times of crisis, B2B companies make an additional effort to reach the target audience, which is probably inserted in the specific problem of the crisis and pays less attention to the different publications. In addition, we suggest B2B companies adopt emojis in social media communication in most countries, especially in the United States, given the effectiveness of emojis and their low percentage of use in tweets in this country. LIMITATIONS AND FUTURE LINES OF RESEARCH The limitations of the first study are related to the analysis of a single case, a single social network, and a single country. Consequently, the findings should not be generalized to contexts with different settings when studying. We suggest the development of new studies applied in different sectors, platforms, and countries to confirm the effect of value co-creation practices on engagement. Additionally, there are low percentages of explained variance in the results, as is usual in social science studies. This means that engagement and its dimensions are also explained by other variables, not only by co-creation practices. The influence of other variables on engagement in B2B contexts could be analyzed in new studies. It would also be interesting to study what other variables are influenced by co-creation practices. Finally, the use of the co-creation practice in other ITS can be studied. The limitations of the second study are also related to the analysis of a single case, a single social network, and a single country. Consequently, the findings should not be generalized to contexts with different settings when studying. The development of new studies applied in different industries, platforms, and in different countries is suggested to confirm the effects of content strategies, message features, and time frame on eWOM volume. In addition, it was suggested to carry out studies that analyze other predictors of eWOM volume, such as the effect of emojis. The unexpected result related to the time frame, namely that tweets published before the COVID-19 pandemic positively influenced eWOM volume, could also be verified in other contexts. The limitations of the third study relate to the analysis of a single industry, the international food and beverage trade shows, and the analysis of a social network. Consequently, the findings should not be generalized to contexts with settings other than those of this study. We suggest the development of new studies applied in different sectors, platforms and in more countries, to confirm the effects of the presence of emojis, the number of emojis, and the time frame (date of publication Ex-Ante COVID-19 or date of Ex-Post COVID-19 publication) in the eWOM volume. Finally, we suggest studying the use of emojis in other contexts.