Setting the future of digital and social media marketing research: Perspectives and research propositions

3.1.1. Conceptual framework and research proposition

In Fig. 1, we propose a humanity framework. Various facets of individuals enacting humanity are depicted in clouds (certain ones are at the forefront); the lower level (such as information seekers) shows dimensions which are more tied to consumption while the upper level ones (such as knowledge creators) are more tied to creation.

Fig. 1. Digital Marketing and Humanity Conceptual Model.

3.2.1. Lack of validated scales

One of the challenges of researching social media marketing is the lack of appropriate constructs (Cooper, Stavros, & Dobele, 2019) and validated scales (Tran et al., 2019) to test the potential models of how social media usage impacts consumption behaviors. While existing scales may be adapted for the context of social media, this approach is not always successful (Tran et al., 2019). The scales currently available do not fully capture and measure the complexity of social media. For example, no scale currently exists to measure the level of perceived connectedness via social media between an individual to others, as well as perceived connectedness via social media between an individual to a brand. It would be interesting to explore these relationships and the link between perceived connectedness and behavioral outcomes, such as brand awareness, electronic word-of-mouth intentions and purchase intentions. One recommendation for future research is the development of scales germane to social media and the capabilities of social media.

3.2.2. Changes in social media

With social media platforms in perpetual beta mode, which is the consistent release of new functions, change is a constant in social media platforms. This makes it challenging to research social media marketing and the metrics associated with social media. While some social media metrics may remain consistent, such as number of followers, likes and shares, other metrics emerge that may be useful. For example, social media influencers, also known as influencers, are individuals with the ability to influence others by promoting and recommending brands and market offerings on social media. The influencer marketing industry is expected to be $15 billion by 2022 (Schomer, 2019) and metrics such as customer influence effect, stickiness index and customer influence value (Kumar & Mirchandani, 2012) have emerged as key factors in evaluating possible influencers to share information about the brand, increase brand awareness and promote electronic word-of-mouth conversations about the brand. Another concern is the problem of decreasing organic reach, which is the number of people who have viewed the content at no cost to the marketer (Tuten & Solomon, 2018). Between 2013 and 2018, the expected organic reach of a brand’s Facebook post dropped from 12 % to 5 % (Tips for Increasing Organic Reach, 2019). With the improving algorithms and surge in number of posts, brands must work harder to gain attention and boost organic reach, which is proving to be more problematic. Beyond the continual changes in existing social media platforms, another concern is the development of new social media platforms.

New social media platforms are introduced and begin to grow in popularity and gain new followers, representing both opportunities and challenges for social media marketing. With the different social media platforms, there may be varying capabilities for brands to interact with individuals, which means social media marketing managers must learn to adapt to effectively use the platform to reach consumers. Furthermore, perhaps new marketing strategies for the new social media platforms must be created to gain customer leads and enhance consumer engagement. For academics researching social media marketing, the novel social media platforms also present opportunities for additional research, including potentially comparing the different platforms, exploring how and why individuals may use the various social media for distinct purposes and developing or modifying metrics for measuring return on investment.

3.2.3. Social network analysis

Social network analysis involves studying networks of people, where each individual is a node. The social structure of connections and ties between nodes are investigated and characterized in social network analysis (Wasserman & Faust, 1995). Social network analysis has been researched in relation to electronic word-of-mouth (Sohn, 2009), identifying influencers (Zhang & Li, 2014) and investigating how individuals influence others, the relationship between influence and tie strength and the flow of information through social media (Gandomi & Haider, 2015). However, there are still gaps in the social network analysis of social media usage literature.

One possible avenue for research may be to explore how consumers’ motivations for sharing information via social media impact others’ perceptions of the message, which is particularly interesting given the phenomenal growth of the influencer industry. An extension of this research is to examine if and how influencers’ use of paid advertisements, sponsorships and partnerships affects followers in the social network. Experimental or quasi-experimental designs may be employed to determine if nodes pay attention to the paid advertisements on social media, and how this may interact with the nodes’ reactions, shares and comments on the influencers’ non-organic posts. Beyond influencers, it would be fruitful to research the potential negative effects of what happens when a node views the same message on social media multiple times. Is there a point where there is a negative effect of exposure to the same information shared by too many nodes? Perhaps the information about the number of shares, comments and reactions on social media may dilute the impact of the relevant information (Nisbett et al., 1981). The potential findings of such research may have both theoretical and managerial implications for understanding consumers.

Often, both academics and practitioners focus on understanding how positive electronic word-of-mouth messages travel through nodes in social networks. However, there is also value in examining negative electronic word-of-mouth in social networks, as recommended by Pfeffer, Zorbach, and Carley (2014). These researchers defined the term online firestorm as the “phenomenon involving waves of negative indignation on social media platforms,” which has affected brands, organizations, politicians and celebrities following the release of less than desirable information (Pfeffer et al., 2014, p. 125). Future research may explore how the technical capabilities of specific social networking sites, an individual’s social identity, altruistic tendencies and commitment towards the organization impacts the individual’s propensity to share negative word-of-mouth messages. In addition, it would be interesting to use social network analysis to investigate how in-group and out-group members are likely to transmit negative word-of-mouth information, as well as the potential influence of negative messages on these nodes (Balaji et al., 2016).

It would be intriguing to compare how negative versus positive electronic word-of-mouth travels through a social network online. The velocity of information flow, volume of information shared, network clusters and cross-posts on different social media may be analyzed and compared for negative and positive electronic word-of-mouth. To conduct this research, a social networking platform that allows all viewers to see the nodes, ties and frequency of reactions and comments may be used. A quasi-experimental design approach may be used to select the nodes; an event or noteworthy news article would be shared to these nodes. Data on the number of times the node viewed, shared and commented on the event and the impact of these numbers may be explored to better understand how viral marketing works. It would be interesting to potentially measure the speed of the information flow, which may be possible with the advanced technical designs of the social media platform. Based on the research limitations and gaps outlined above, the two propositions are formulated hereafter to help guide future research on this topic.

3.2.4. Research propositions

When individuals make judgments, the dilution effect occurs when people fail to use diagnostic information in the presence of nondiagnostic information (Nisbett et al., 1981). For social media marketing, the dilution effect may occur when consumers view the number of reactions and comments about a message, which serve as the nondiagnostic details. The message is the diagnostic information. However, if the number of reactions and comments on the social media platform is too high, perhaps that may serve to dilute the effect of the diagnostic message details and serve to negatively impact the individual’s judgment of the information. The following is proposed.

Proposition: On social media platforms, a) the number of reactions and comments positively impacts the individual’s perception of the message to a peak point; b) beyond the peak point, the number of reactions and comments negatively impacts the individual’s perception of the message.

The immediacy of social media makes it easier for people to act on their altruistic impulse and use it to spread messages and propagate goodness (Tuten & Solomon, 2018). On the other hand, individuals may also decide to use social media for the purpose of altruistic punishment, which is bringing attention to a person or organization whose behavior is socially unacceptable (Rost et al., 2016). The altruistic punishment available through the use of negative messages on social media allows enforcement of the group’s social norms. To enforce the social norms on the offending person or organization, the negative word-of-mouth information is spread more quickly through a social network, when compared to positive word-of-mouth information. Thus, the following is postulated.

Proposition: In social networks, negative word-of-mouth information flows faster than positive word-of-mouth information.

3.2.5. Conclusion

With more consumers turning to social media use daily for activities such as reading the news, researching products and enjoying entertainment (GlobalWebIndex Social Flagship Report, 2019), organizations must strategically use social media marketing to appeal to their target audiences. However, challenges in using social media to reach consumers include lack of appropriate scales to measure and investigate constructs of interest, the constant changes in current and emerging social media platforms and the application of social network analysis to research the flow of electronic word-of-mouth messages and the influence on consumers’ attitude and behaviors of such information. We encourage researchers to conduct further research in these areas to better understand the phenomena of social media marketing to benefit both academics and practitioners.

3.3.1. Challenges and opportunities for research

The first challenge relates to customer experience (CX) design. Research emanating across IS and marketing literatures do indicate that CX-design of D&SMT plays a fundamental role in encouraging favorable customer behaviors towards brands (Carlson et al., 2018; Kamboj et al., 2018). Nevertheless, consumers expect a seamless, integrated and holistic customer experience, regardless of the channel. Offering multi- or omnichannel consumption experiences, whereby interactions with a variety of digital channels and with real people (either phone or in-person) complement rather than compete with each other, enhances the overall customer experience (Bolton et al., 2018). However, despite channel integration efforts by organisations, recent market reports show that 54 % of UK customers are disappointed with their most recent experiences (Temkin Group, 2017). As such, the design and measurement of CX initiatives involving D&SMT and their integration will become paramount in order to unlock the full extent of CEBs and well-being to maximize mutual value fulfilment for the customer and the firm.

The second challenge concerns personalization. D&SMT provides significant opportunities for highly personalized communications. Here, brands can offer content and customized product recommendations resulting in greater customer satisfaction and CEB outcomes. Despite this, several challenges surface for consideration. First, given the growing availability of personalization options through D&SMT, customer brand evaluations are expected to increase. This has important implications for CX management as when customers develop habits to using specific technologies, their initial customer delight is expected to transfer to their realm of expectation (c.f. Rust and Oliver, 2000). A second challenge is managing the personalization/privacy paradox. Numerous corporate data breaches have led to government legislation for increased privacy e.g. The European Union General Data Protection Regulation (GDPR), tighter data security and the consumer right to erase their information which has the degree of difficulty for marketers to build meaningful and authentic relationships with consumers through personalized experiences. Furthermore, when brands leverage collected customer data to provide relevant, personalized content, it might heighten some customers’ sense that the brand has some manipulative intent, which could spur privacy concerns (Aguirre et al., 2016). This has negative consequences for cultivating CEBs and their well-being where consumers are less likely to participate, for example, in sharing their motivations, preferences and desires with brands. Despite these challenges, the opportunity therein lies in understanding how to overcome these issues so that consumers are willing to disclose this information for personalization to more readily occur so that satisfied and loyal customers can be achieved.

The third challenge is associated with segmentation. With the increasing rate of media fragmentation and channel multiplicity, segmenting the marketplace has become vital. However, as D&SMT adoption intensifies in the marketplace, careful consideration is needed as to the development of technology-specific user segmentation. While many customers will use familiar technology (e.g. websites, mobile apps), technologies that are new or more peripheral to the market offering (e.g. interactive virtual assistance, social media to register complaints, self-serve technologies) will likely see varying adoption levels across customer segments, including in terms of demographics, psychographics, or brand- or marketing-related preferences. This will become a critical issue looking forward as by 2050, demographic trends indicate that society will contain growing numbers of aged population and consumers with disabilities, changing family roles and structures, and global migration (Fisk et al., 2018). As such, the opportunity arises for customer experiences through D&SMT to be designed for inclusion to account for the myriad of ways in which different segments of consumers can access and engage with brands.

The fourth challenge relates to innovation and collaboration. The use of D&SMT has enabled customers new opportunities for interacting and collaborating with brands in the innovation process (e.g. Carlson et al., 2018; Kamboj et al., 2018). For instance, the increasing sophistication of online brand communities on social media have enabled opportunities for customers to become active collaborators that generate new ideas within these communities (e.g. My Starbucks Idea). Here, advancements in social media monitoring, text and image analysis techniques that “listen” to, and capture, customer-generated content enable ideation, sharing, product development and brand experience improvements (Humphreys and Wang 2017; Villarroel Ordenes et al., 2018). As such, this requires organisations to prepare for and invest in their own adaptive capability to capture, manage and exploit these opportunities from customers to gain a deeper understanding of customers’ available resources and those they are willing to invest in particular brand-related interactions. This understanding can, in turn, be translated into customization tools to optimally cater for specific customer needs, wants or preferences (e.g. Lego Ideas customer co-creation platform).

3.3.2. Agenda for research and practice, and research propositions

The following paragraphs advance a set of research questions relating to the challenges and opportunities outlined above.

3.3.3. Concluding statement

IS and Marketing scholars have an essential role to play in shaping the agenda of how D&SMT can cultivate CEBs in the future and generate greater value for the profit and non-profit organisation. However, to contribute meaningfully, it must be noted that greater interdisciplinary awareness between IS and marketing scholarship is required to unveil novel insights to achieve CEB objectives. It is anticipated that the research agenda will be useful for scholars to stimulate new insights that inform and guide how practitioners may harness technology to deliver greater benefits for the customer, their well-being, and ultimately stronger customer-brand relationships. To date, there has been growing research activity in the literature related to CEB, and many important and noteworthy contributions to knowledge have been made. To advance the knowledge base further forward, particularly given the fast-moving nature of digital settings, research that attempts to broaden our understandings of CEB, and examines new and evolving technologies that enhance the customer experience that impacts society as a whole will be most valuable.

3.4.1. Theoretical advances

Computers and artificial intelligence have been used for decades to help digitalize business processes in and within companies in the buyer-seller relationships as well as to attract and keep customers. Martínez-López and Casillas (2013) provide complete review of industrial marketing deployment of artificial intelligence e.g. in segmenting and pricing from 1970s to 2013. Continuing from that Syam and Sharma (2018) provide a detailed overview impacts of machine learning (ML) and artificial intelligence (AI) on personal sales and sales management. They also detail an overview of worthy research areas that are still relevant. In addition, ML and IoT has been focused from a more systemic perspective i.e. business model innovation (Leminen et al., 2019), where authors found four different types of architypes of business models that are utilized by different B2B companies. Also IoT and ML studies are reviewed in detail by Leminen et al. (2019). Since, the publication of Leminen et al. (2019), Suppatvech, Godsell and Day (2019) provide a literature review on how IoT is linked to industrial services. They provide four architype models with add on, sharing, usage-based and solution oriented where degree and intensity of IoT usage varies. At a more fine grained level, Liu (2019) utilizes big data analysis and ML methods to show that negative user-generated content – customer sentiment has negative impact on stock performance. The ever increasing popularity of big data research and big data skills have spilled over to an industrial context. Sun, Hall and Cegielski (2019) build an integrated view to explain the decision to adopt big data. They show that relative advantage and technology competence are key factors when planning the adoption. B2B companies face challenges in digital marketing technology investment decisions (Sena and Ozdemir, 2019) as well as on deciding how and what skills to develop (Sousa and Rocha 2019). Data, big or small, is analyzed in multiple ways by B2B companies. Different types of analytics are used (big data, web and social media among others) to identify possibilities to increase marketing performance (Järvinen and Karjaluoto, 2015), automate marketing activities (Järvinen and Taiminen 2016) and better social media marketing (SMM) (Sivarajah et al. 2019). Besides big data, IoT and ML also virtual and augmented reality trials have been popular in the industrial field as well which have also resulted in some research. Laurell et al. (2019) study barriers to adopt virtual reality which are – lack of sufficient technological performance and limited amount of applications.

SMM in the context of business markets has witnessed a steady increase in the number of publications as the review of literature indicates (Salo, 2017). Since the publication of Salo (2017) and other related systematic literature reviews on sales (Ancillai et al. 2019) and advertising (Cortez et al., 2019), several scholars have approached industrial SMM. Subsequent to the publication of the aforementioned reviews on SMM, traditional areas of B2B marketing such as advertising, marketing communications, strategic management, sales and NPD have been of interest to those who utilize social media lenses in their research. Within advertising, two recent publications are discussed. First, with help of a hierarchy-of-effects theory Juntunen, Ismagilova, and Oikarinen (2019) study the use of Twitter by world’s leading B2B companies. They identify from an advertising perspective that companies use Twitter for creating awareness, knowledge and trust, interest and liking amongst their followers by both strategic and tactical means. Interestingly, purchase, preference and conviction are much less advertised. Twitter feed scraping is gaining popularity among B2B marketers as a similar study was conducted by McShane, Pancer and Poole (2019) while focusing on a fluency perspective. From a marketing communications point of view, Magno and Cassia (2019) survey 160 companies and find that thought leadership (similar to influencer marketing within a business to consumer context) via SMM, has a positive influence on brand performance and customer performance. Additionally, they highlight that specific thought leadership capabilities could emerge that support social media capabilities. Similarly to Magno and Cassia (2019), Foltean, Trif and Tuleu (2019) are interested in these capabilities and survey 149 companies. It was shown that social media technology improves the CRM capabilities of a B2B company. Besides capabilities, acquiring new customers i.e. sales is pertinent function in the B2B marketing field as indicated by the two reviews. Iankova et al. (2019) identify that SMM is used in all stages of the customer engagement cycle for acquiring new customers. From the new product development point of view, Cheng and Krumwiede (2018) studied supply chains. By surveying 367 Taiwanese companies they found that market and technological-knowledge processing capabilities strengthens the positive effect of using social media in the new product development process with the involved supplier.

3.4.2. Limitations of research and practice

Some limitations of research and practice can be identified from the existing research. First, one can see that previous studies still to some extent, rely on comparison with B2C and B2B which is rather unorthodox method and provides limited information for industrial marketing scholars. Second, several studies still rely on relatively small sample sizes (about 75 companies) in surveys as well as a relatively small amount of key-account interviews (e.g. seven experts) conducted. Third, even though web scraping and utilizing e.g. netnography in qualitative research is gaining popularity, these need to be linked with objective measures. A Limited amount of research in general, exists that utilizes objective measures, which can correspond to a lack of real-world relevance. Fourth, it seems that sales and advertising are still pertinent areas of focus while the scholars could also deepen our understanding in other areas of industrial marketing (e.g. buyer-seller relationships).

3.4.3. Agenda for future research

Based on the meta-analysis of the existing systematic literature reviews and research, one can highlight some future research areas. First, B2B research will see in the future B2B companies using more of their resources to content marketing that is geared toward lead generation e.g. via A/B tested email campaigns. Second, many B2B companies are still lacking skills in SEM especially SEO which links to customer insights creation with 360-degree video as well as immersive (AR-VR) content or any interactive content for that matter. Third, in the realm of website design predictions are that customers are increasingly focused on page speed download times among other factors which leads to website experience optimization unless 5 G will be adopted at faster rates. Fourth, B2B companies have increasingly adopted different types of website and social media analytics tools which desperately needs research especially on effectiveness of current data dashboard and visualizations, are we focusing on the right issues? As is the case with consumer markets, B2B market companies will increasingly use predictive or behavioral marketing tactics leading to an increased use of learning e.g. with bacterial algorithms. Fifth, some industries such as finance and food are likely to be interested in applying digital technologies such as blockchain, cloud computing, thin clients, remote monitoring and sensors that alter not only the buyer-seller relationship, but also customer behavior. All these technologies can be used in different B2B categories (Salo, 2017). Considering the above discussion and help guidance for future research, the following two propositions are offered.

3.4.4. Research propositions

It can be synthesized, with some hesitation from the previous B2B digital and social media marketing review, and proposed as a core of future research that digital technologies in broad (including AI, platforms, IoT, Machine learning, big data, different digital analytic and visualization tools) are changing in ever increasing pace B2B digital and social media marketing. This poses a challenge to theory testing and building as different types of digital technologies influence marketing sub-domains (Reid and Plank 2000) and empirical contexts (industries) in different ways. Besides, these challenges to theoretical advancements are also an ethical dilemma as part of the general opinion, with some parts of the research community having woken up to the dark side of the ever-increasing volume of data and opaque predictive algorithms.

Proposition: Digital technologies are increasingly changing digital and social media marketing in the context of B2B.

In order to extend the previously mentioned core research focus where theoretical advances are sought for in B2B Digital and social media marketing one has also adapt novel research methodologies. Both case studies and surveys have been traditionally popular method of gaining new insights in the field of B2B marketing and business studies at large. Still, relatively little of the digital research methods, that offer, in many situations objective data are fully utilized. Digital analytics tools (Järvinen 2016), neuroscience (Bear et al. 2020), motion, voice and other sensors (e.g. Nguyen 1997; Heikenfeld 2016), netnography (Kozinets 2010), data scraping and mining (Munzert et al. 2014) provide novel data and data triangulation possibilities (Patton 1987). Hence, it is proposed that in future we embrace the plurality of research methods, be it used for mixed or single method research.

Proposition: Digital research methods with objective data are to be widely employed in understanding digital and social media marketing in the context of B2B.

3.5.1. Research propositions

In this section, two propositions are developed to encapsulate the theoretical and practical arguments outlined in the previous section. Their development draws on further elaboration of the concept of Digital Content Marketing (DCM) and how customer experience and customer journey comes into play when it comes to delivering on relevant, specific and targeted DCM from the brand.

For instance, Hollebeek and Macky (2019) suggest that consumer engagement in the DCM context includes elements of multi-tier interaction between the brand and its consumers. These intra-interaction consequences such as consumers’ cognitive, emotional, and behavioral engagement with the brand thus trigger extra-interaction consequences of brand trust and attitude thus developing brand equity through the DCM strategy.

Lamberton and Stephen’s (2016) comprehensive examination of the literature suggests that a DCM offers several benefits to customers including exploratory and positive behaviors that directly benefit the brand (cf. Appel et al. 2019; Dabbous & Barakat, 2020; Leeflang et al. 2014). Drawing upon and synthesizing the aforementioned works in marketing and information systems literature examining customers’ interactions with brand-related stimuli (Digital Content in the paid, owned or earned media) on a digital platform. Customer experience with Digital Content refers to a customer’s perception of their interactive and integrative participation with a brand’s content in any digital media (Judy & Bather, 2019).

3.6.1. Fake reviews and trust in eWOM

Many business reports in the last years document that an increasing number of consumers rely on online reviews and ratings to make purchase decisions. Online reviews have become a primary information source in consumer information search and their extraordinary success is due to their reliability and usefulness. Initially, eWOM was perceived as a reliable form of information about a brand and its products, particularly compared to marketing information sources. However, mass media all over the world in the last number of years, has reported several scandals within the online reviews industry. This has been a factor within the tourism sector, revealing the practice of some managers posting promotional reviews about their business and offering discounts or freebies to consumers in exchange for glowing reviews (Filieri, 2016). For instance, an executive at Accor Group hotels in the Asia-Pacific region was caught posting several positive reviews for its hotels acting as a customer and Australian property developer Meriton, has been ordered to pay $3 m for manipulating TripAdvisor reviews about its serviced apartments (The Guardian, 2018). TripAdvisor, Amazon, Yelp have often made the news for not being able to prevent fake or promotional reviews to be posted and appear on their website. The surge of fake reviews has stressed the importance of the reliability of OCRs (Filieri, 2016).

Research on eWOM has paid huge attention to the concept of source credibility and on the influence that credible communicators have on consumer behavior such as purchase intention (e.g. Hovland et al., 1953). However, research on trust in eWOM is in its infancy. From a theoretical perspective, source credibility theory (Hovland et al., 1953) is one of the most popular theories that are used to understand what makes a communicator an influential source of communication. The source credibility model established by Hovland et al. (1953) analyze the factors that can affect the receiver’s acceptance of a message and rest on two core variables: source expertise and source trustworthiness. A source is considered as expert when he/she is knowledgeable about the subject, while trustworthiness refers to the source honesty, sincerity, and believability (Hovland et al., 1953). Source credibility is important in eWOM as expert sources influence information helpfulness (Filieri, 2015), information adoption (Cheung et al., 2009), and purchase intention (Filieri et al., 2018; Ismagilova et al., 2019).

Some scholars have started to investigate the determinants of OCR trustworthiness (e.g. Qiu, Pang, and Kim, 2012; Hu et al., 2012; Filieri, 2016). Scholars have investigated when and for what type of vendor fake reviews are more likely to appear (e.g. Hu et al., 2012), the role of credible information on consumer perceptions and behavior (e.g. Cheung et al., 2009) while other studies have focused on the impact of various elements (source, message, receiver) on perceived information credibility (e.g. Filieri, 2016). eWOM message credibility is defined as the extent to which one perceives a recommendation/ review as believable, true, or factual (Cheung et al., 2009). For instance – Cheung et al. (2009) investigate the determinants of e-WOM perceived credibility in China and found that source credibility, confirmation of prior belief, recommendation consistency, recommendation rating, and argument strength influence perceived e-WOM review credibility. Qiu, Pang, and Kim (2012) used experiments with students and found that a conflicting aggregated rating decreases review credibility and diagnosticity for positive reviews but not for negative reviews via the mediating effect of review attribution. Other studies take a more holistic approach and consider the various elements that can affect trust and mistrust perceptions in eWOM considering the source, the message, the receiver, the format, the pattern, and the valence, of the review (Filieri, 2016). For example, Filieri (2016) use a qualitative approach to analyze consumers’ perception of trust and mistrust in online reviews and build a theoretical framework explaining trustworthiness and persuasion in e-WOM communications. They reveal that some consumers – namely the more experienced one in the use of online reviews and those with higher levels of involvement in the purchase; adopt a set of criteria to assess review trustworthiness. The study reveal that factors that affect trust and mistrust in online reviews, which refer, in order of importance, to the content and writing style of a review (degree of detail, type of information, length, writing style, and presence or not of picture), review extremity and valence, the source of communication, the pattern emerging from reading several reviews. Ketron (2017) found that reviews with higher quality of grammar and mechanics (spelling, punctuation, capitalization, and organizational elements of writing such as paragraphs) have higher perceived credibility and exert a stronger influence on consumers’ purchase intentions. Drawing upon Filieri (2016)’s framework, Baker and Kim (2019) found that readers of exaggerated fake posts find the utility and trustworthiness of the reviews diminished as a result of exaggerated emotions and language present.

3.6.2. Research gaps

Although valuable studies have been conducted, more research is needed on the topic of consumers’ perception of eWOM trust and mistrust. In order to provide some guidelines for future research we have to first distinguish among the three different elements that can affect the persuasion and influence of a communication: the source, the receiver, the message, the medium, and the context (Wathen and Burkell, 2002). As highlighted above, the majority of studies have focused on the textual format of eWOM, specifically focusing on reviews and ratings posted by anonymous reviewers (with no followers) in third-party retailers such as Amazon, TripAdvisor, and Yelp. Research has established that consumers pay more attention to the credibility of the message rather than to the source (Filieri, 2016), which comes into play rarely and only for high involvement decisions. Thus, scholars in the future should consider information quality as a potential mediator or moderator of the effects of source credibility dimensions in eWOM contexts and with high versus low involvement decisions to understand when the message and source are more important in the evaluation of review trustworthiness.

Furthermore, few studies have focused on the characteristics of the receiver of eWOM messages, the perceived credibility of the medium where reviews are hosted, and the context in which the information is processed. Accordingly, eWOM can manifest in different formats and channels. For instance, eWOM can manifest though textual, verbal, visual messages or a combination of them (e.g. consumers’ photo and comment in Instagram). Moreover, eWOM can appear in different channels such as online consumer communities (i.e. TripAdvisor), social networking websites (i.e. Facebook, WeChat), instant messaging apps (i.e. WeChat, WhatsApp, Snapchat), blogging and microblogging websites (i.e. Twitter, Weibo), photo sharing social networking (i.e. Instagram), and video-sharing platforms (i.e. YouTube).

Future research could assess whether consumers trust some formats when compared to others. For instance, a picture is a thousand words and provides evidence of the purchase and actual use/consumption of products (e.g. photo of a dish in a restaurant) increasing review trustworthiness (Filieri, 2016). Scholars could analyze whether the combination of different formats (e.g. video + text, picture + text, video + picture, picture or text only) can increase review trustworthiness. Experimental methods can be employed for this purpose.

Furthermore, research for instance has not considered if the medium where the review is published affects perceived eWOM trustworthiness. Medium credibility focuses on the perceived credibility of the channel through which content is delivered rather than the sender (or senders) or the content of the message. Consumers may have different perception of credibility of the reviews and ratings published depending on the channel where they are published. For instance, consumers may trust more the reviews and ratings published in Google or Facebook rather than in Tripadvisor.com, because the reviewer can be a person in their social network, it has a public profile, and he/she generally provide more information about himself/herself. Different variables can affect perceived trust and reliability of a medium such as the reputation of the company (i.e. brand reputation) and consumers’ attitude towards it, the frequency of use or habit, the quality of the information provided, the ease of accessibility of them, and usefulness of the information provided and so on.

3.7.1. Challenges

There are four main challenges facing research in social media marketing, each of which also has consequences for practice: (1) the speed of development of both practice and research in social media marketing; (2) the interdisciplinary nature of the field; (3) the diversity of research questions; and, (4) the wide range of theoretical perspectives and research methods.

3.7.2. Proposals and propositions for future research

In developing their proposal for future research, Lamberton and Stephen (2016) coupled social media marketing with digital, and mobile marketing. Although the focus in this short piece is on social media marketing, it is important that future research agendas in social media marketing acknowledge its contextualization relative to the wider digital marketing field, including mobile marketing, and the plethora of other marketing channels. In addition, due to the interfaces between the different marketing and social media channels and the rapid evolution of both research and practice in the field, the boundaries of any agenda for future research are likely to be blurred and dynamic. This offers unique challenges in proposing a comprehensive agenda for future research into social media marketing. Indeed, there is evidence that previous researchers have proposed that, for example, more research be conducted on a specific action on a social media platform, only for the action to be removed from the platform soon after the article was published. Hence, this article offers a selective and personal perspective on the key areas for future research. This agenda covers the following broad themes: Systematic literature reviews; Social media marketing management; Social media consumer behavior; and, Social media context. In all of these areas, there has been some previous research, but there is considerable scope for exploring these topics further.

3.8.1. Understanding the concepts: defining augmented reality

There is general agreement that Augmented Reality (AR) is a “medium in which digital information is overlaid on the physical world that is in both spatial and temporal registration with the physical world and that is interactive in time” (Craig, 2013, p.20). There is not so much agreement, however, about how it differs from related concepts such as Virtual Reality (VR), Mixed Reality (MR) or Assisted Reality. In the 1990s, Milgram et al. (1995) developed the famous “Reality-Virtuality-Continuum”, which means that everything between VR (where people are completely isolated from reality) and the real world is called Mixed Reality, with the two subcategories of Augmented Reality and Augmented Virtuality. When screening academic and industry-driven publications, however, readers may soon find that the terms are not defined uniformly, and that this is a limitation. More specifically, since the announcement of the HoloLens device by Microsoft in 2016, the common understanding of MR has developed into a synonym for “very realistic AR”, i.e., content that is not only superimposed, but also realistically integrated into the environment. In contrast, a device that clearly does not deliver realistic content, like Google Glass, is not typically called a Mixed Reality (although it probably would be according to the Milgram Continuum). A common term for this minimalist, typically purely functional/textual, superimposed (mostly 2D) content is “Assisted Reality”. Here, users get supporting content, ideally contextual, in their field of vision.

Finally, there is a superordinate term – XR – that represents all of the above concepts. There is no agreement about whether the “X” stands for “extended”, “expanded” or only serves as a variable X for “anything” about new and innovative forms of realities. Fig. 2 summarizes this alternative to the Milgram Continuum, a more industry-specific classification of new realities. Accordingly, XR serves as a generic term with two independent subcategories, AR and VR. AR is a generic term for a continuum ranging from Assisted Reality to Mixed Reality. This means that in a perfect Mixed Reality, users cannot distinguish between virtual elements and real objects.

 

Fig. 2. Augmented Reality, Assisted Reality, Mixed Reality and Virtual Reality.

Numerous forecasts from many institutes indicate a future in which consumers will consistently be exposed to Augmented Reality. In particular, specific glasses-like devices, so called “Augmented Reality Smart Glasses” (e.g., Google Glass, Microsoft HoloLens, or MagicLeap One), are expected to soon move from being in a proof-of-concept phase or an enterprise tool to being a mass-market technology. Almost all the companies concerned, including Apple, Samsung, Sony, and many start-ups, are announcing the arrival of such glasses. As Apple CEO Tim Cook has stated, Augmented Reality “will happen in a big way, and we will wonder when it does, how we ever lived without it. Like we wonder how we lived without our phone today” (quoted in Leswing (2016)). Forecasted numbers from BCG (2018) and other companies support this quote.

When Augmented Reality is used in marketing, it is called “Augmented Reality Marketing” (Rauschnabel et al., 2019). Although still at an early stage and potentially futuristic, this strategic concept already raises many fundamental questions to which academics are encouraged to provide answers to.

Before discussing four broad areas of research directions, I will start with a short discussion on how and where Augmented Reality can influence marketing, in particular the marketing mix (product, place, price, and promotion). Fig. 3 shows a very simple representation of a supply chain from suppliers to consumers. Traditional e-commerce (e.g., online shops, branded shopping apps, etc.) have established direct links between many brands and consumers, disrupting established supply chains over the last two decades. Augmented Reality functions can further improve and expand these channels. In B2B marketing, Augmented Reality can be an effective tool for marketing products or services to other companies, usually as an advertising or sales tool (e.g., a tool that visualizes a new robot in a customer’s factory). In B2C marketing, Augmented Reality can create direct interactions between manufacturers and consumers and thus offer new opportunities for communication (“promotion”) and sales (“place”). Furthermore, Augmented Reality can be used to expand core products and services (“product”). For instance, companies can use it to extend their range of services or physical product with AR components (e.g., manuals in AR or additional features). Brands can expand their product portfolio with virtual holographic products (e.g., a home decor brand could offer additional holographic decoration items that the user can only experience in AR). Third party vendors can also use Augmented Reality to enable their customers to virtually try out the products they offer from multiple vendors (e.g., Amazon’s Try at Home feature for selected products), suggesting that these vendors may need new forms of data (e.g., 3D models) from manufacturers. Finally, as indicated by the gray arrow at the bottom of the figure, companies can also obtain information from users – through interaction as well as through context data, as will be explained later.

Fig. 3. Augmented Reality Marketing along the supply chain.

3.8.2. Research direction 1: strategic augmented reality marketing

Augmented Reality Marketing is a strategic concept that requires adequate resources and competencies, including (but not limited to) budget, knowledge of 3D visualization, technologies, or user behavior. This also means that Augmented Reality Marketing is interdisciplinary. For example, marketers must consider collaboration with PR, sales, innovation management, IT, law, etc., to increase the chances of success.

Strategic scholarly projects could build on prior research on strategic marketing, organizational behavior, dynamic capabilities and related theories. Potentially relevant questions include:

•

Which dynamic capabilities drive the success of Augmented Reality Marketing? Which competencies do Augmented Reality marketers need? How should these requirements be integrated into digital marketing curricula?

•

How should Augmented Reality Marketing be organized and implemented? Which departments and functions should be involved?

•

How can the success of Augmented Reality Marketing be measured? What are relevant KPIs?

3.8.3. Research direction 2: the role of augmented reality in marketing

When people think of Augmented Reality Marketing, many probably immediately associate it with the promotion mix. Certainly, many companies have shown good examples of how AR communications/applications can attract attention (e.g., Burger King’s app, which allows users to virtually “burn” competitors’ ads to get a free Whopper, or Pepsi, who harassed people in a bus shelter by adding monsters, animals and meteors on a screen that looked like a window). A recent BCG study (2018) reflects this view and shows that most marketers see Augmented Reality in the early stages of the customer journey (e.g., building brand awareness), but expect this to change (e.g., generating revenue).

Augmented Reality can also be used outside the promotion mix, as shown by Lego, for example. The toy brand offers Augmented Reality functions for many of its products. In Lego’s case, customers can add virtual fires to castles or play with virtual Lego figures. In the future, companies might also use Augmented Reality to improve the perceived value of their products (e.g., remote maintenance services or additional functions, as Lego is already doing; Hinsch, Felix & Rauschnabel (2020)). In addition to adding Augmented Reality to the product value, it is also imaginable that Augmented Reality can even replace physical products. Only recently, Microsoft has been offering MSOffice applications for its HoloLens device and showing what future offices can look like without screens and hardware. This could also point to new virtual competitors. With a few exceptions (e.g., Carrozzi et al., 2019), the idea of treating Augmented Reality content as a new product category is new. An industry study by Ericsson (2017) concluded that even Augmented Reality ads could be a threat to products, since these ads “could in essence turn into free versions of the products or services themselves”. The impact on marketing, such as 3D-scanned holographic counterfeits of branded products, can be enormous, but these possibilities have been largely overlooked in the literature. Virtual products can also have an impact on pricing decisions (e.g., questions can arise about the willingness to pay for virtual products). Finally, AR apps can serve as a further direct-to-consumer channel.

Some unanswered questions that are both theoretically and managerially relevant are:

•

Product: How do consumers interact with virtual products in their perceived real world, compared to real products? Consumers can become attached to real products, but can they also do so with virtual ones? What advantages and disadvantages do consumers see in virtual versus real products?

•

Price: Are consumers willing to pay for virtual products? Do added services (e.g., specific functions added via Augmented Reality) increase the willingness to pay? Are consumers willing to pay higher prices for products if they can experience them through Augmented Reality before making a purchase decision?

•

Place: In which situations do which consumers prefer Augmented Reality channels to browse or purchase products? What are the advantages and disadvantages?

•

Promotion: What drives the effectiveness of promotional messages in Augmented Reality? How does good content marketing or good storytelling in Augmented Reality look like? Does the ability to convince consumers to buy a product after trying it out in real life lead to better decisions and lower return rates?

3.8.4. Research direction 3: understanding the users

To increase the chances of success in Augmented Reality Marketing, managers need to understand how users interact in AR (Han et al., 2018). Research in MIS, marketing, psychology, communication sciences and other disciplines has significantly improved consumer interaction in reality (e.g., supermarkets) and in virtual environments (e.g., social media). Augmented Reality, on the other hand, is different. For instance, Rauschnabel et al. (2019) argue that Augmented Reality content must be integrated into the real world well in order to generate inspirational user experiences. In another study, I showed that people expect Augmented Reality to be used to change and improve the real world, as it would be with decorative objects, but also with things that people do not own or cannot afford or that just do not exist (e.g., fairies). This construct – Desired Enhancement of Reality – is specific to the Augmented Reality context and does not apply at all to any other technology (Rauschnabel, 2018). Finally, Rauschnabel et al. (2018) discuss that data protection research has typically focused on the protection of a user’s privacy. Since AR must by definition track and understand the real world, this also affects the privacy of other people.

Fruitful avenues for future research in Augmented Reality lead beyond testing established models such as the Technology Acceptance Model. In contrast, future research should pay particular attention to the unique characteristics of Augmented Reality. That is to say, it should either deepen our understanding of the recently identified AR specific constructs or even identify and study other distinctive characteristics.

There are many theories on media and technology acceptance that scientists can apply to the context of Augmented Reality Marketing. However, I encourage academics to begin with a broader, perhaps even explorative or theory-building approach that always aims to identify and understand more conceptual differences between Augmented Reality and other media formats. This is relevant both at the construct level (i.e., the identification of specific Augmented Reality constructs) and at the process level (i.e., which psychological processes affect users in Augmented Reality compared to other media?). Methodically, qualitative research is an effective way to identify these differences, and experimental designs can compare Augmented Reality with other media formats. Longitudinal studies as well as field-experiments can further deepen and validate these findings.

Some relevant questions include:

•

What drives the adoption of Augmented Reality? Who are the early adopters of Augmented Reality? What are potential moderators or segmentation variables for detecting and understanding consumer heterogeneity in the answers to these questions?

•

What exactly are the dimensions of consumer experience in Augmented Reality?

•

What risks and fears do people perceive when interacting with Augmented Reality? How do these factors change over time?

•

Are people willing to integrate the brands they like or love into their homes, as they would with merchandising? If so, what are the conceptual differences?

•

How can virtual products – including virtual counterfeits – impact marketing? How does it impact consumer-brand relationships, for instance, if consumers 3D-scan branded products and replicate them as holograms?

3.8.5. Research direction 4: the role of data in Augmented Reality Marketing

Many companies today can use a variety of data to automate or at least support marketing decisions. However, this data is typically generated from behavioral data by tracking consumer behavior in digital environments. However, many behaviors happen “offline” and are thus invisible to marketers. For instance, marketers do not typically know too much about a person’s lifestyle or products that are bought offline or received as gifts (unless they can make some good predictions based on data).

Augmented Reality must by definition “understand” not only users, but also their environments. Cameras, microphones, depth sensors and other instruments therefore continuously track, map and interpret a user’s environment. Augmented Reality technology could, for example, recognize a small 10-year-old television in a user’s large living room. It could then recommend a new TV by overlaying the old screen with a new and bigger one. The user could then immediately see how much better a larger screen would be. It is maybe not surprising that threats to both users’ and other people’s privacy (Rauschnabel et al., 2018) might play a dominant role in these discussions.

Future research will probably overcome the technological hurdles (e.g., the analysis of streaming data, including object recognition) to make use of this new form of data, but many new questions will remain and require interdisciplinary research:

•

What is legally possible?

•

What are potential ethical concerns?

•

What reaction do consumers show when their environments are continuously streamed, analyzed and complemented by marketing messages?

3.8.6. Research propositions

As discussed, forecasts predict consistently growing markets for Augmented Reality and indicate that it will become a mass medium in a few years. Recent research and case studies show the enormous potential of Augmented Reality Marketing. The possibilities range from communicative gimmicks to strategically developed investments, in almost all industries, including consumer goods, B2B, tourism, services and so on. Today it is almost impossible to market products or services purely offline. In the future we can expect a similar role for Augmented Reality. Consumers will live in a world that is consistently enriched with virtual content by marketers and other players. Companies must adapt to these developments, and Augmented Reality will probably play a similar role in almost all industries as online channels do today.

Proposition: Augmented Reality will be as prevalent in the marketing of the future as the Internet is today.

Since the role of Augmented Reality goes far beyond communicative aspects, dealing with it is very complex. Managers need to better understand the unique aspects of consumer behavior, specific goals, KPIs, technological challenges and so on. Building on this, they need supporting tools to develop Augmented Reality strategies. As scientists and educators it is our duty to support managers with insights and to include relevant topics in our curricula. This includes relevant concepts (e.g. theories and strategies), tools (e.g. 3D modelling), boundary conditions (e.g. legal and ethical aspects) and technological fundamentals (e.g. tracking or mapping).

Proposition: The management of Augmented Reality is highly complex. Therefore, companies must develop specific skills and scholars should support this.

3.8.7. Conclusion

Augmented Reality is still in its infancy, but most probably about to experience a breakthrough. As discussed in this section, Augmented Reality Marketing is much more than just a communicative gimmick, it might become a new paradigm in management research and practice. However, surprisingly little research has been conducted in this field. I hope that this section inspires scholars from a range of disciplines to research Augmented Reality and how it can positively impact marketing, businesses, consumers, and societies as a whole.

3.9.1. Responsible AI-driven social media marketing

Responsible AI is defined as the integration of ethical and responsible use of AI into the strategic implementation and business planning process (Wang et al., 2020). It can be viewed as a guardian framework that is used to design ethical, transparent and accountable AI solutions that create better service provision. Such solutions harness, deploy, evaluate and monitor AI machines, thus helping to maintain individual trust and minimize privacy invasion. Responsible AI places humans at the centre and meets stakeholder expectations and applicable regulations and laws. The ultimate goal of responsible AI is to strike a balance between satisfying customer needs with the responsible use of AI and attaining firms’ long-term profitability.

Responsible AI can be conceptualized from the perspectives of economic and social sustainability, in which the role of AI is highlighted as an enabler of social responsibility and business long-term success. From the economic sustainability perspective, AI tends to be viewed as a powerful tool not only for maximizing economic value but also for satisfying customer needs with fewer ethical concerns and dilemmas. In the social media context, firms should place AI ethics and standards at the core of their social media marketing strategy by formulating ethics policies on harnessing social media data and by considering socially favorable AI approaches and algorithms to create effective marketing communications for consumers. From the social sustainability perspective, the societal impact of AI on the well-being of humans and environments requires considerable attention. Developing AI solutions by considering human rights, fairness, inclusion, employment and equality can lead to potential gains in credibility for products and brands.

From these two perspectives, three responsible AI initiatives for social media marketing are identified, namely: (1) Ethical AI design and governance; (2) Risk control by human and AI coordination; and (3) Ethical AI mindset and culture.

3.9.2. Research questions

The initiatives mentioned above offer substantial potential to extend our understanding of responsible use of AI in social media marketing. They also provide new avenues for research in areas that demand attention from scholars; more specifically, they suggest the need for in-depth studies in the context of social media marketing. Below, I suggest relevant research questions with regard to ethical AI design and governance, risk control by human and AI coordination, and ethical AI mindset and culture.

3.9.3. Research propositions

In order to implement responsible AI it is necessary to develop ethical AI design and governance and to minimize potential risks and biases. First, an ethical AI design should take into account data transparency and privacy protection in order to retain consumer trust. For example, Capital One has created QuickCheck, a criteria system that provides an initial decision with transparent and comprehensive explanation before customers make a formal application for a credit card online. This responsible AI initiative is featured in their “Check it, don’t chance it” social media marketing campaign aimed at creating customer awareness.

Second, it is essential to minimize potential biases (e.g., algorithmic, racial, ideological and gender biases) during the ethical AI solution development stage. By doing so, it will be possible to improve the accuracy of product recommendations and targeted promotions and advertisements (Rai, 2020). For instance, to prevent racial and gender biases, L’Oréal uses an AI-powered spot diagnosis algorithm developed by modelling more than 6000 patient photos with consideration of different ethnicities, age, gender, and skin conditions. Another example is Oral B’s smart toothbrush, which tracks thousands of brushing styles and behaviors from a range of users so as to provide personalized feedback and advice to its customers.

When AI-enabled products are designed ethically and responsibly, taking into account customer needs and values, consumers will perceive the product as highly trustworthy, and this will increase their engagement levels with marketing campaigns. Meanwhile, AI tools developed by diverse and bias-free data sets can provide a deeper understanding of customers and thus drive more effective social media marketing spend. In view of this, we suggest the following proposition:

Proposition: The implementation of responsible AI will drive social media marketing performance.

By adopting an effective marketing communication approach, firms should be able to gain high market share and brand reputation with less investment in advertising and promotion (Luo & Donthu, 2006). Particularly in social media environments, where consumers receive large numbers of AI-generated marketing messages from various platforms on a daily basis, effective marketing communication driven by AI can lead to positive word-of-mouth and brand image and the formation of purchase intention. Research by the Direct Marketing Association (2018) found that 57 % of surveyed consumers are comfortable with receiving personalized marketing messages generated by marketing automation tools that may surprise them and help them make purchasing decisions. However, in some cases, the collection and utilization of social media data by AI algorithms can lead to consumers receiving irrelevant or redundant marketing appeals, delivered indiscriminately via social media platforms. In order to avoid this issue, which can cause frustration for customers and the loss of credibility for products and brands, it is essential that firms devise appropriate AI methods (Wang et al., 2020).

Explainable AI has the potential to address this challenge, as it can achieve targeted marketing. Explainable AI provides a precise rationale of the marketing decision-making process and provides marketing managers with a human-understandable explanation (Overgoor et al., 2019; Rai, 2020). For instance, explainable AI could be used to inform marketing managers why a specific discount voucher is sent to a consumer and though which social media channel a specific product is targeted to a prospective consumer. This would increase the accuracy of marketing efforts and reduce the likelihood of relatively undifferentiated advertising executions.

The above discussion suggests that social media marketing communication can be optimized to achieve better market returns for firms that strive to implement explainable AI in their marketing practices. Therefore, we suggest the following proposition:

Proposition: Social media marketing communication has a stronger influence on firm’s market returns in firms that highly engage in explainable AI development.

3.10.1. Innovative advertising

For a digital marketing campaign to be successful, the essential and most important aspect is to connect with the right audience. With AI powered advertising, it can be simple and straight forward for marketers to target the right audience. AI tools can gather information, analyze the data and predict future behaviors. With this data, marketers can target advertisements as per the interest and predicted consumption patterns of the audience. Moreover, advertisements can be enhanced with Augmented Reality (AR) and Virtual Reality (VR) thus collecting more information from the customers and AI can eventually utilize the data for a more engaging and personalized advertisement experience.

3.10.2. AI powered websites

AI powered website builders can now build websites based on user data, thus, reducing the cost and time of creating an interactive website from scratch. Currently, the majority of these AI-powered builders are still in the early stages of development. However, these AI driven processes are likely to transform digital marketing in the future.

3.10.3. Enhanced shopping experience

AI powered marketing is going to change the traditional way of shopping by improving the online shopping experience. A number of brands are trying different possibilities with various formats of AI to upgrade the shopping experience for customers, allowing them to interact with products via the use of AR before purchase.

3.10.4. Genuine product recommendations

The granularity of the data analysis via AI algorithms are far beyond human imagination. AI based digital marketing strategies are now being used to assist marketers in ascertaining the behavior, interests, needs, habits of customers at every stage of the sales cycle providing product purchase suggestions to customers in real time.

3.10.5. Chatbots

Chatbots are widely being integrated on initial touchpoints with customers and are programmed in such a way so as to answer customer’s questions in real time. Chatbots are likely to perform a greater number of marketing related tasks and develop increasing levels of sophistication as well as data analysis capability.

3.10.6. Here are a number of ways in which AI is changing the face of digital world

The ability of AI to analyze, monitor and utilize data, removes the need for humans to be involved in many of the mundane roles within data analysis. Brands and marketers are leveraging AI digital marketing for detailed and improved customer’s experiences. AI based digital marketing is helping marketers gain a competitive edge in numerous ways.

3.11.1. Type of mobile advertising

We explain the types of mobile advertising through location based, informative, credible, entertaining, incentive based and irritating formats. First, location-based advertising (LBA) helps the consumers to receive the information on the go. LBA facilitates comparison among stores, location, entertainment or shopping and helps the marketers as well as consumers by connecting the “when” and “where” elements. Further, the weather associated with a location also have significant role in determining effectiveness of mobile advertising on consumers (Lin et al., 2016). Second, informative nature of advertising, which conveys the messages about the products and services helps in increasing the effectiveness (Smith, 2019). Third important dimension is credibility, which refers to the truthfulness and believability of the ads. Both advertising credibility and advertiser credibility determines consumer confidence and consumer expectation. (Lin and Bautista, 2018) Credibility influences the advertising value (Ducoffe, 1995) along with consumers’ attitude and behavior. Fourth is entertainment, which is the ability of the ad to provide enjoyment and satisfy consumers’ diversion and aesthetic needs. These help to develop positive moods, attitudes and attention (Elliott and Speck, 1998). Entertainment leads to increase in advertising effectiveness and perceived value. Fifth, incentives provided to the consumer via advertising have positive affect on the consumers. Incentives relates to the advertising value, perceived value and perceived usefulness. This is because shoppers always yearn for “a good deal” and “being a smart shopper” (Park et al., 2018). Sixth, irritation, which is that type of advertising where the marketers annoy, insult or offend the consumers leading to negative, impatient and displeasure towards the ads (Smith, 2019). This intrusion interrupts the activities of the consumers, leading to negative associations, which significantly affects the advertising value (Ketelaar et al., 2018).

Proposition: types of advertising primarily, location based, informative, credible, entertaining, incentive based and irritating influences the effectiveness of mobile advertisements.

3.11.2. Content of mobile advertising

The key content related elements of mobile ads include style & personalization, user control, functional, relevance &contextual, short and emotive. First, the content preferred by the consumer especially needs to be stylish and personalized for positive consumer responses. Positive responses lead to perceived value and perceived usefulness (Shareef et al., 2017). Second, consumers also want the content to be functional (Smith, 2019). They avoid useless and intrusive content using “swipe” or “ask me later” options. Further, content has to depict “real life”. Third, users should have control over the content and it should be rigorous enough to deliver the message in five seconds (Smith, 2019). Fourth, content of the mobile advertising has to be contextual and relevant as they positively influence the perceived value (Lin et al., 2016). Contextual content is premised on the interests, preferences, and consumption patterns of the consumers, which increases the willingness of the individuals to engage with the ads (Lin et al., 2016). Fifth is relevance, which refers to ad content’s congruency (Shareef et al., 2017) with the lifestyle and interest of the consumers, resulting in positive attitude towards ads. Sixth, content of the advertising has to be emotive. Emotional content would generate positive feelings, enhancing the advertising value and perceived value.

Proposition: content of mobile advertising specially style, personalization, user control, functional, relevance, contextual, short and emotive dimensions significantly influence the effectiveness of mobile advertising.

3.11.3. Organism: processing of mobile advertising

Consumers process mobile advertising as the part of internal evaluation represented by Organism in the SOR model. This process focusses on three elements; advertising value and structure, user interface, and perceived value. According to the elaboration likelihood model (ELM), mobile advertising affects the peripheral processing as the emphasis is on the existing needs and creation of new needs for the products and the services (Shankar and Balasubramanian, 2009). The customer interactions in terms of the overall structure and user interface is imperative for processing mobile advertising. ELM also works with hedonic or utilitarian elements of products and emotional involvement of products and services. Thus, cognitive and affective drivers act while consumers process mobile advertising. This leads to interactivity, involvement and repurchase of products and services (Lu et al., 2018). Second, user interface, which primarily includes visual design where the size, color, music and animations play an important role in affecting consumers’ responses. Third, perceived value is the overall benefit that the products and services offer through mobile advertising, when compared to competitors. Simply put, it is the “subjective evaluation of the advertising” (Ducoffe, 1995). This associates with the perception about the company’s competitive advantage that forecast purchase intention (Grewal et al., 2016). Thus, perceived value affects consumers’ purchase intentions, attitudes (Lin et al., 2016) and behavior.

Proposition: Type of advertising and content of advertising significantly influences the processing of mobile advertisements in terms of advertising value and structure, user interface and perceived value.

3.11.4. Moderator: social influence

Consumers’ responses are moderated by social influences (Shareef et al., 2017) as they process mobile advertising. Social impact theory also states that numbers of peers and their proximity increases the social impact on the individuals and thus, the subsequent responses for mobile advertisements are influenced. Social influences also affect the trustworthiness and motivation of consumer responses. Thus, social influences affect consumers’ attitudes and responses towards mobile advertisements. Social influences are related to the subjective norms based on the theory of planned behavior, aspirational and associative reference groups that enhance the trust of the consumers and eventually leads to the positive consumer responses.

Proposition: consumers’ processing of the mobile advertisements especially with respect to advertising value and structure, user interface and perceived value is moderated by social influences primarily for the consumer response.

3.11.5. Consumer response

Consumer responses to mobile ads are in the form of interactivity, involvement, attitude towards ads and brands, and purchase intention. Interactivity relates to the online navigational mechanism, which provides immediate feedback to consumers, providing better communication. Interactivity has multiple dimensions, as it is two-way communication with advertising context and user control. Interactivity is a favorable consumer behavior in the context of mobile advertising (Lu et al. 2019). Interaction leads to involvement, which is the subjective experience of the consumers, related to affective state. Mobile advertising can extensively involve consumers with contextual messages that can engage and stimulate. The personal involvement, physical involvement and situational involvement, which reflects relevance for the consumers leads to advertising value (Lu et al., 2019). Thus, mobile advertising can involve consumers, develop interactions, provide seamless and pleasing experience, which eventually develops positive attitude towards the ad, brand and develop purchase decisions.

Proposition: Processing dimensions such as advertising value, structure, user interface and perceived value with moderation of social influences significantly include the consumer responses in terms of interactivity, involvement, attitude towards advertisement, attitude towards brand and purchase intention.

LBA is an emerging area of research, which has an attractive marketing platform. This form of advertisements is diverse and ubiquitous as the features of mobile phones, smart phones, tablets and computers are divergent. LBA uses location-tracking technologies, which includes Global Positioning System and Global Navigation Satellite System to understand the real time locations of individuals so that ads can be delivered based on the geographical place or site of mobile phones. Thus, the ads become individualized with the instant, personalized and location specific features that can cater to the specific needs of the ad recipients (Bruner & Kumar, 2007; Dhar & Varshney, 2011). The individualized and personalized advertisements are likely to attract high ‘click through’ rates. Hence, LBA becomes more appealing and persuasive which can lead to success.

LBA facilitates the delivery of the messages to the specific users where the effects would be high and positive. This from of advertising started for static media platforms such as billboards and websites but has eventually moved to mobile phones. Mobile advertisements provide value for different companies by providing their users with individual, location based, dynamic real time content. To optimize this content, location based technology is used to individualize marketing communication for users. Thus, LBA is an effective channel that helps marketers to reach out to users and engage them through innovative approaches which persuade increased purchase of products and services. The clicks by the users indicate that location based advertising has the potential to illustrate the user interest and can encourage them to consider offers related to the advertised products and services. However, these clicks do not ensure they are interested in the products and services, Thus, future studies have to investigate the decision making process with respect to location based advertising.

Proposition: Location based advertising would significantly affect the buying decisions.

Further, contextual factors (Zhang and Katona 2012) would increase the complexity of optimal advertising behavior, which would elevate issues for the marketers. For instance, Google maps offers location sensitive version of the sponsored search advertising. This helps in ensuring that the relevant ads appear on the headlines in the displayed map. Interestingly, users’ willingness to pay reduces with the increase in the distance to the stores. This provokes marketers to develop refined mechanisms, which are related to contextual factors that could help users as well as marketers. Therefore, future studies can investigate moderating effects of contextual factors, which could positively affect the outcomes of mobile advertising with optimum amount and placement of mobile spending within the overall marketing budget.

Proposition: Contextual factors such as location, weather and social interact can affect the intention to buy the products among the users and budget allocation by the marketers for advertising.

3.12.1. Future steps

The definitions concerning the term mobile marketing have, quite surprisingly, not evolved much. The American Marketing Association (2019) uses the Mobile Marketing Association’s definition in its dictionary, which defines mobile marketing as comprising “advertising, apps, messaging, mCommerce and CRM on all mobile devices including smart phones and tablets.” As the definition implies, mobile marketing now covers almost all marketing-related communication in different formats (e.g., text and video) accessed via all mobile devices.

Several papers have proposed a research agenda for mobile marketing and advertising. Grewal et al. (2016) provided an overarching framework for future research, addressing and proposing research directions related to environmental context and technological context, consumer-related variables, market factors, advertising (e.g., the goals and elements of advertising elements such as media), and metrics. Their framework offers an in-depth analysis of the central factors related to understanding, in particular, the advertising side of mobile media.

Lamberton and Stephen (2016) discussed the research evolution of mobile marketing from 2000 to 2015 and proposed an agenda for theory development in the field. Their key proposition was that mobile marketing researchers should “focus on understanding the marketing value of aspects of mobile technology that allow marketers and/or consumers to do things that cannot be done with nonmobile technology (e.g., geo-located ad targeting; making use of sensors in mobile devices that measure ambient contextual attributes, or even user biometrics, in the case of wearable devices)” (p. 165).

From a more practical perspective, the paper by Bakopoulos, Baranello, and Briggs (2017) proposed several key future research avenues for mobile marketing researchers. They identified how prior research has not sufficiently addressed the issue of how much of the total advertising budget should be allocated to mobile and which advertising formats, targeting strategies, and tactics brands should use. In addition, their study proposed three key research questions dealing with 1) consumers’ path to purchase and the role of mobile advertising in it, 2) how to effectively communicate a message on mobile, and 3) how to target effectively using mobile advertising.

As the world has already turned “mobile” in the sense that mobile internet use has bypassed desktop internet use in many markets, research on understanding the various aspects of marketing and advertising on mobile devices can be expected to continue to flourish in the coming year. Hereafter the following two propositions are formulated in order to help guide future research related to mobile marketing.

3.12.2. Research propositions

The use of mobile marketing can bring many benefits to both retailers and consumers. According to Ström, Vendel, and Bredican (2014), the key value that mobile marketing can provide for retailers is enhanced brand communication and improved service interactions. Mobile marketing can also be very useful in post purchase stage. However, to date there is not much research confirming the various possible positive effects mobile marketing might have on customer relationships. Thus, we propose researchers look at these effects within future studies and propose the following:

Proposition: Mobile marketing communication will have a positive significant effect on brand communication, increasing for example brand awareness, brand attitude, customer satisfaction, customer retention, and positive word-of-mouth.

According to Lamberton and Stephen (2016), one of the key features of mobile marketing is to leverage the specific features mobile that can bring additional value to both consumers and marketers. The mobile specific features such as location-based ad targeting and wearable devices, can bring the time and context variables in advertising targeting to a completely new level. Targeting effectively using mobile advertising is of key interest in this debate (Bakopoulos, Baranello, and Briggs, 2017). Thus, researchers are encouraged to provide information relating to these mobile specific elements and study how they might create additional value to both the seller and the buyer. On this basis, we propose that:

Proposition: The more mobile marketing builds on mobile specific features, such as location-based ad targeting and use of mobile marketing in real-time service interactions and post purchase, the more positive the brand communication effects will be.

3.13.1. Data and algorithmic issues

Data is one of the largest creations of humankind. Around 2012, this fact has become popular with the advent of the concept of Big Data. Since then, many claim that ‘data is the new oil’ while others advocate for the absolute necessity to ascertain data validity and veracity before deriving any value from it. Amidst them marketers who succeeded to reshape the digital media environment in order to apply new practices: programmatic advertising, customer retargeting, e-WOM and influencer marketing, to name a few. But at what cost? The overwhelming ocean of data surrounding us is riddled with ‘digital pollutants’ that deteriorate data quality. Bias in data can appear in different ways, from data leaks and privacy breaches to spam, fraud reviews, fake social media identities and fake news.

As explained by Fulgoni & Lipsman (2017), metrics could be at the same time part of the problem and the solution. The search for higher levels of impressions, reach, frequency and demographics has been exacerbated under the pressure of media planning and campaign efficiency measurements, and helped incentivize the audience and engagement systems. As a result, massive clickbait and inorganic (paid rather than earned) if not fraudulent content creation on digital platforms have become prevalent and could negatively affect customer loyalty and their trust toward brands. In addition, brands could lose visibility on their targeted audiences and have to systematically clean the data used. Research from the computer science and social network analysis fields have recently provided valuable insights on how to navigate through polluted data streams, especially concerning the effects produced by fraudulent reviews (Ananthakrishnan et al., 2015), fake social network accounts identification (Sarna & Bhatia, 2018) and fake news detection (Zhou, Zafarani, Shu, & Liu, 2019). In marketing and Management Information Systems, research in this area is still scarce, with a few exceptions concerning mainly the side effects of e-WOM. Tuzovic (2010) has used frustration theory to investigate how negative emotions expressed verbally and non-verbally are reflected in negative e-WOM and their dysfunctional impacts on the brands. Liu & Karahanna (2017) have identified the ‘swaying’ effects of e-WOM and explained how online reviews could be more or less helpful during the consumer preferences construction in a decision-making context. Kim, Naylor, Sivadas, & Sugumaran (2016) have studied incentivized (inorganic) e-WOM recommendations and concluded that manipulating reviews will have impacts not only on the audience but also on the communicator (the producer of the manipulated recommendations) depending on the valence of the recommendation (positive or negative).

Bias in data appears in different forms that needs to be investigated and cured to assure marketers and digital decision-makers a clean and transparent environment. More research is therefore needed concerning: (1) the metrics used to assess digital and social media efficiency actions, especially studies on real and manipulated impressions distinctions; (2) Invalid traffic identification and the measurement of their effects on data quality could also constitute an important research avenue; (3) Strategy-oriented studies that put under scrutiny the roles, power and responsibilities of big tech gatekeepers (Google, Amazon…), regulatory institutions, brands and other human and non-human stakeholders of the digital eco-system will also be highly valued.

When we think about non-human actors in the digital era, algorithms and AI come immediately to mind. An algorithm is defined as a systematic procedure that produces, in a finite number of steps, the answer to a question or the solution of a problem (Britannica Academic, 1999). Algorithms are the building blocks that make up machine learning and artificial intelligence. Omnipresent in the digital marketing systems, they provide search results, content recommendations and targeted advertisements. They also identify patterns and make predictions to improve user experience. Algorithms have also their drawbacks, when they use biased data input and also when they contribute to produce more bias. As they are conceived as ‘black boxes’ that learn from their own previous processes due to machine learning, algorithmic systems can go beyond the control of those who developed and implemented them. Amazon’s AI-based hiring system has been recently shut down because it appeared that it discriminates against women because it did not derive recommendations based on reasoning, but applied a complex statistical model using data from more than 10 years of incoming past applications and hiring decisions. Because men outnumbered women among technical employees during this period, women’s applications were automatically underestimated (Maedche, Legner, Benlian, et al., 2019).

This example illustrates how algorithmic systems are likely to reinforce existing structures of control and power/technology regimes of truth as defined by the French philosopher Michel Foucault (Avgerou & McGrath, 2007). New terms such as algorithmic literacy and algorithmic transparency have entered the public discourse and voices are raised in support of informing and educating public about them. Education here does not only mean to learn how to code but also to be more conscious about the societal impacts of these systems.

A multidisciplinary body of research, including computer science, management science, ethics and philosophy is investigating this topic. Mittelstadt, Allo, Taddeo, Wachter, & Floridi (2016) developed a prescriptive map to organize the debate about the ethical implications of what they call the algorithmic mediation of our society. Bozdag (2013) explored the algorithmic gatekeeping bias that is affecting the filtering protocols of recommendation systems. Baeza-Yates (2016) have stressed the importance of user context to avoid these biases in recommendation and personalization systems. These systems use predictive algorithms to provide consumers with personalized product and information offerings, building on their previous behaviors and actions, with the immediate effect of lowering the decision effort, but with the risk of exposing consumers to narrower content over time. This phenomenon named the ‘filter bubble’ and also known as the ‘echo chamber’ has been explored by Nguyen, Hui, Harper, Terveen, & Konstan (2014) in the context of the movie industry. More research on this topic is certainly needed in different consumer contexts and industries.

3.13.2. User misbehavior issues

Research on consumer misbehavior is not new and has focused on a psychological perspective. It has been defined by Fullerton & Punj (2004) as: “behavioral acts by consumers, which violate the generally accepted norms of conduct in consumption situations and thus disrupt the consumption order representing the dark, negative side of the consumer” (p.1239). Consumer misbehavior could appear in different ways in which consumers can cause harm to brands and other consumers, such as destroying and vandalizing brands property; theft, fraud and shoplifting; abuse, intimidation and bullying other consumers and brands’ personnel. Whereas these issues can be observed in offline and online settings, most of the studies addressing these are conducted within an offline context.

In the online context, the cyber-psychology discipline has extensively focused on technology-related users misbehavior, starting with studies on dependence and addiction (Young, 1998), then on various aspects including cyberbullying, privacy concerns, jealousy, exhibitionism, voyeurism. These studies were conducted especially within social media users’ communities (Kefi & Perez, 2018). Researchers and practitioners have started to notice that the use of digital and mobile devices can be a double-edged sword, i.e. it may have positive and negative effects at the same time, or a Janus Face as concluded by Mäntymäki & Islam (2016) in their study of networking on Facebook. Islam, Mäntymäki & Kefi (2019) have studied the undesirable effects of regret experienced by users on social media especially on brand engagement. The idea of an equivocal behavior of consumers online has more generally been illustrated by Jarvenpaa & Lang (2005) who identified 8 paradoxical situations that result from human–technology interaction: (1) Empowerment/Enslavement; (2) Independence/Dependence; (3) Fulfills Needs/Creates Needs; (4) Competence/Incompetence; (5) Planning/Improvisation; (6) Engaging/Disengaging; (7) Public/Private; and (8) Illusion/Disillusion. We argue here that adapting this model to investigate consumer-brand interactions in a multichannel (online and offline) context could be an interesting research avenue. In effect, specific research on the (intended and non-intended) consequences of these behavioral issues within digital and social marketing is still at its early stage. This has been discussed by Scheinbaum (2017) and Zolfagharian & Yazdanparast (2017) who have used a qualitative approach to identify the facets of technology- related dark-side consumer behavior and their effects at the individual, organizational and societal levels.

We consider in this paper that there are three areas of investigation of user misbehavioral issues implications on digital marketing. The first concerns the user dependence on technology and includes different facets: over-engagement, over-visibility and user presence quality. The second concerns the user-technology-brand triangle and its duos of paradoxical positionings: trust/privacy, authenticity/calculation and destruction/preservation. Finally, the third addresses the cognitive and emotional user limitations and include: information overload, preference construction schemas and out-sourcing decision-making patterns to human and non-human influencers and prescribers.

3.13.3. Conclusion

The two perspectives discussed in this paper related to data/algorithmic and user- misbehavior issues are in fact intimately connected. Both are related to user-technology interactions that may have double-sided effects: positive and negative. We believe that to substantiate the positive side, we have to put within our grasp and constrain the negative one. Research in this area requires large-scale quantitative studies, data-driven and new experimentation protocols. More qualitative oriented studies are also needed to explore newly experienced behaviors and phenomena that emerge from human-digital interactions and their implications for digital marketing decision-makers and consumer wellbeing.

3.14.1. Research propositions

There continues to be a critical need for data literacy. As of 2019, Facebook’s ad disclaimer dictates that the sponsoring entity needs to be disclosed on all ads regarding social issues, politics, or elections. A positive step forward would be to expand the scope of the ad disclosures to all advertisements and to all platforms; the expansion would mean that advertisers are not only responsible for declaring who is sponsoring the ad, but also for providing a description of the microtargeting practices in order to help the public understand why they are seeing the particular ads.

Future research should endeavor to understand what influences individuals’ “marketing comfort” (Jacobson et al., 2020). There is a critical need for scholarly intervention to understand what impacts individuals’ attitudes and perspectives towards the use of their social media data. We need interdisciplinary and intersectional perspectives to develop best practices that afford every person the ability to develop the skills to successfully navigate the digital world. Rather than delegate this important task to grade school education, data literacy needs to be a life-long commitment that is not only taught, but reinforced, by various stakeholders, such as higher education, workplaces, and government. Considering that this area is in need of further research and investigation, this paper makes the following research proposition:

Proposition: The public’s comfort with social media marketing may be affected by data literacy.

As the public—and governments—gain data literacy, there may be increasing pressure on organisations to engage in more transparent and ethical data practices. Across various industries, third parties would approach this task in different ways. For example, researchers may make a statement in their publications about the ethics of their data collection and use; social media marketers may disclose how they obtained and used the public’s social media data.

Ethical data practices need to be taken seriously by third parties. In Europe, the GDPR dictates how businesses can collect and use the public’s personal data; companies in countries outside Europe may elect to become GDPR compliant in an effort to boost their ethical organizational practices. As third parties increasingly gain access to more data and more sensitive data from unknowing individuals, third parties may develop self-regulation social media practices in an attempt to sidestep government regulations. As such, the following proposition is offered:

Proposition: Social media data transparency will play an increasingly important role in ethical organizational practices.