By Martin P. Block, Professor Emeritus, Medill Integrated Marketing Communications and Yingying Chenb, Department Chair and Professor of Electrical and Computer Engineering and the Peter D. Cherasia Faculty Scholar at Rutgers University
Previous cross-cultural studies have provided theories and evidence of distinctive patterns in consumer attitudes, behavior, and outcomes of consuming specific media channels, such as social media, eWOMs, and traditional media (Goodrich and de Mooij 2014; Nam and Kannan 2020; Shankar et al. 2022). As omnichannel marketing increasingly emphasizes the integration of digital and traditional media to create seamless and coordinated brand communication (Verhoef, Kannan, and Inman 2015), existing studies are still limited in understanding the influence of both online and offline media channels and their joint effects on consumers’ purchase decisions in different cultural contexts (Nam and Kannan 2020). Adding a cultural dimension to understanding the allocation of various media channels and their effects becomes critical for marketing researchers and professionals.
The challenges of comparing a variety of media channels include the gathering of large samples from different cultural contexts and the consistent measurement of media influence that allows comparison across a variety of media types (Block and Schultz 2009). Additionally, as media multitasking becomes more prevalent and media channels exert synergies on marketing communications, simultaneity—the interdependence among media channels becomes increasingly relevant in media usage (Schultz, Block, and Viswanathan 2018). However, existing literature provides limited knowledge about the role of cultural differences in media contingencies. To address the challenges, we examine the influence and interdependence of a comprehensive list of online and offline media on consumers in China and the United States. The two countries represent distinctive cultural contexts regarding Hofstede’s cultural dimensions (1980). Specifically, based on a large sample of US adults (n=17,430) and Chinese adults (n=6,058) collected during the first quarter of 2022 from the commercial syndicated database, we first delineate the influence of 25 online and offline media channels. We then use Bayesian network analysis to compare media contingencies in US and China. Our study extends previous cross-cultural comparison studies to understand the ecologies of media channels and their influence on buying decisions. We also provide methodological innovations by integrating Bayesian network analysis to study media interdependence.
Cultural Dimensions in Media Influence and Contingencies
Hofstede’s cultural dimensions (1980) lay the foundation for cross-culture comparisons. Power distance, individualism/collectivism, and uncertainty avoidance are particularly relevant in explaining the distinctive media usage patterns. For instance, consumers from high uncertainty avoidance cultures are more likely to rely on expert opinions (de Mooij and Hofstede 2011), seek information endorsed by other social media users (Sweeney, Soutar, and Johnson 1999), or trust information from friends and peers over objective sources such as magazines, and productive reviews (Dawar, Parker, and Price 1996). In contrast, consumers from individualist cultures often seek information from a wider range of sources, including traditional media, and rely less on interpersonal communications (de Mooij and Hofstede 2011) than those from collectivist cultures. Consumers from collectivist and high-power distance cultures are also more likely to seek information on social media and online recommendations, and rely on personal contacts for information gathering, whereas those from individualistic and low-power distance cultures tend to seek information from both interpersonal communication channels and traditional media (Goodrich and de Mooij 2014).
The influence of media also varies across different cultural contexts. Consumers from collectivist cultures are more influenced by interpersonal communications, such as word of mouth (Money, Gilly, and Graham 1998), and by feelings and social influence rather than hard facts (Goodrich and de Mooij 2014). Consumers from high uncertainty avoidance and collectivist cultures are more likely to stick to a smaller set of channels, making it more expensive for them to switch channels (Kumar and Pansari 2016). In collectivist cultures, consumers often remain loyal to retail channels with which they have established deep relations with retailers, while in individualist cultures, consumers frequently switch channels for convenience (Pick and Eisend 2016). The effects of omnichannel communication on consumer experience are stronger in high-uncertainty avoidance cultures than in low-uncertainty avoidance cultures (Nam and Kannan 2020).
Despite many studies showing cultural differences in media influence, how media channels are contingent on each other in different cultural contexts remains largely unexplored. The media contingencies are relevant to omnichannel strategies because consumers often seek information and make purchase decisions under the influence of a communication ecology. Instead of relying on one single media as the source of information, individuals tend to rely on multi-model communication connections, including interpersonal communication, mass media, digital media, and community information sources, to seek and construct knowledge to realize goals (Broad et al., 2013). Communication ecology theory proposes that individuals construct a network of communication resource relations to pursue a goal in their communication environment (Kim and Ball-Rokeach 2006). The communication ecology approach assumes that information sources are interdependent in a media system (Altheide, 1995). This assumption is particularly supported by the literature in consumer-driven multitasking with media (Schultz, Block, and Viswanathan 2018) and media synergies (Authors, 2018). Thus, the question for this study is whether and how cultural difference links to media contingencies that influence consumers to make purchase decisions.
Given a limited cross-cultural comparison of both online and offline media influence and contingencies in the US and China, we ask the following research questions: How do online and offline media channels differently influence an individual consumer’s purchase decision in the US and China (RQ1)? Which set of media channels has a greater influence on making purchase decisions by Chinese consumers than on US consumers (RQ2)? How are media channels contingent on each other differently in the US and China (RQ3)?
Data and Methods
The data on Chinese consumers are drawn from the China Quarterly Studies (CQS), an ongoing, syndicated research study developed and maintained by Prosper Technologies through its ProsperChina™ division. The data on US adults is primarily from the ongoing Media Behavior and Influence (MBI) Study. The study uses a double-opt-in methodology. Each group of respondents is balanced against US Census data (age and gender) to ensure it represents the nation at large. The CQS used here is limited to the first quarter to correspond to the time (January-February 2022) of the annual MBI. The US data (n=11,122) here has been reduced to only 18-54 ages to facilitate comparison with the China data (n= 6,058). The Appendix provides a more detailed description of the samples (p.19 in the Appendix).
Media influence (RQ1, RQ2) is measured by asking if the type of media influences the purchase decision across a variety of categories including apparel/clothing, automobiles, eating out, electronics, grocery, medicines, telecommunication services/mobile, and financial services with a yes or no response. The options contain a comprehensive list of 25 media channels (see Appendix p.20). We calculated an average purchase influence across eight categories.
To delineate the contingencies between media (RQ3), we used Bayesian network analysis. Bayesian networks (BNs) were chosen as the modeling platform because of their capacity to handle large datasets and nonlinearities, make inferences, and examine uncertainties. Bayesian networks graphically encode a joint probability distribution among variables (Pearl 1988). A BN represents the conditional independence between the variables in the probability distribution while the variables connected to one another each have a conditional probability distribution to locally represent the strength of their connections. BNs are often used when uncertainties exist about the likely outcome for a variable in a system model and when inferences are optimally considered in an omnidirectional fashion (Conrady and Jouffe 2015). In our study, BN analysis calculates the conditional probability of the influence of one media given the influence of another. Each node in the network is a media channel. Each arc (edge) represents the strongest mutual information (MI). The Maximum Weight Spanning Tree (MWST) algorithm was used to identify the strongest MI link. The arc force was calculated using Kullback-Leibler Divergence. The direction of an arc represents the flow of influence from one node to another. The node force, the sum of the arc forces of all arcs connected to the node, was calculated to identify the most important media in the entire network. All BN methods were implemented in Bayesialab 10.2 (Bayesia SAS 2022). The Appendix (p. 20) lists more details.
Data from Chinese consumers indicate a greater level of media influence on making purchase decisions in every category (Table 1). The overall average across the twenty-five types is nearly double for China compared to the US. The top six media channels that generate greater influence on Chinese consumers are videos on the mobile phone (China=30.4%, US=11.4%), outdoor (China=22.5%, US=7.6%), product placement(China=23.0%, US=8.6%), web radio(China=21.3%, US=7.1%), read articles(China=24.9%, US=10.8%), and mobile apps (China=26.0%, US=12.0%). The closest category is the internet (China=30.5%, US=26.8%), the top influence media in both countries. Word of mouth appears to have a similar influence in both countries (China=28.0%, US=21.5%). Coupons, social media, and TV are near the tops for Americans. Notably, on average, all media types are more influential in China.
For the strength of dyadic relations in terms of their relative influence (Table 2), the strongest pair for both countries are social media and mobile (KL Divergence: China=0.376, US=0.558). The second most important for the US is TV and Cable TV (KL=0.439), but in China is the word of mouth and in-store (KL=0.330). Mobile and reading an article are high among both (China=0.239, US=0.420). In China, coupons are associated with outdoor (KL=0.170), whereas in the US coupons are associated with email (KL=0.290). Comparing the strength of mutual influence from each node (Table 3), web radio has the highest node force in China (total node force=1.628), indicating that web radio has the strongest connection to all other media influence in and out. Web radio is driving relative influence in other media types as shown by the outgoing node force. In China, yellow pages, mobile, video games, and product reviews complete the top five. Social media shows only outgoing force. The US shows blogs as the highest, with web radio in second place. Reading an article, cable TV and mobile complete the top five. Social media, unlike China, shows only incoming force. The biggest differences appear to be cable TV and the role of social media.
Discussion and Implications
In this study, we conducted a cross-cultural comparison of the media influence and contingencies between consumers from the US and China. We found differences and similarities in media influence and contingencies. Particularly, we showed similarities in the influence of the internet, which is the most influential type of media for Chinese and Americans to make purchase decisions. This finding perhaps relates to a similar trend of increasing media allocation on digital in both countries (Lee, Yalcinkaya, and Griffith 2022). We also found that word of mouth and product reviews are among the top media types for both countries.
Other findings point to the difference between China and US. We highlighted three major differences. First, the media influence is higher in China than in US. This is perhaps because in a collectivist, high-power distance culture rooted in Confucianism, consumers rely more on interpersonal connections in building their own media channels. Eventually, they stick to the channels as a ritual in daily life. Our finding also suggests that marketers may consider investing in omnichannel media in China to optimize the consumer experience (Nam and Kannan 2020). Second, traditional and digital media are more interconnected in China, whereas in the US, the traditional and digital cluster together. Perhaps this finding relates to the popularity of QR codes, which connects online and offline media channels in China. We found that mobile media (mobile apps and videos) are more influential among Chinese consumers than their US counterparts. Mobile media has also become a hub connecting traditional and online media in China but not the US. This is perhaps because US consumers show higher risk aversion when using mobile applications, such as mobile payment and have more concerns about privacy than Chinese consumers (Zhang et al. 2018). A larger media group supported by the internet, read articles, word of mouth, and product reviews cluster in the media network of US but not China, suggesting that US consumers might spend more time seeking information from those media channels to avoid uncertainties. Lastly, we found that social media plays different roles in China and US. In China, social media influence other media channels, but other media influence social media in the US. Perhaps in a communal society like China, consumers engage with eWOMs on social media to select information from other media channels. In an individualistic society like the US, consumers seek information from various media channels, which leads to posting their own perceptions of products on social media. Marketers should therefore consider the unique affordances of social media in different cultural contexts.
To read the research article in its entirety with the list of references, please email the editor of the Retail Tech Bulletin.