Abstract
The future of businesses is not only orchestrated by emerging trends and megatrends but also to a greater extent struck by surprising events. With the accelerating presence of high-tech innovations and smart technologies, business activities are exposed to an increasingly volatile, uncertain, complex, and ambiguous environment. Any unexpected changes in the environment could lead to the malfunction or even collapse of a company, a business, or even an industry. Discerning the seeds of change and anticipating the potential disruptions in the external environment is a precondition for avoiding potential risks and threats. The present study contributes to the conceptual and methodological discussion of disruptive transformation by identifying, analyzing, and interpreting the potential disruptive factors in the external business environment. The venture is undertaken through the lens of probing into the dynamism of China’s e-commerce industry, considering its growing influence both domestically and abroad. Any unexpected disruptions would result in a significant ripple effect on interrelated businesses, industries, and even economies. The research adopts a combined mode of qualitative and quantitative methods in the form of horizon scanning and a Consumer Delphi study. Horizon scanning collects futures signals indicating the seeds of change (i.e., weak signals) and potential disruptions (i.e., wild cards), whereas Delphi study solicits the evaluations on the degree of likelihood and impact of the collected factors from an expert panel. As a result, twenty-seven potential disruptive factors are discovered, categorized, estimated, and discussed.
Research Highlights
The growing momentum and global influence of China’s e-commerce industry were depicted.
Literature review as a method for horizon scanning served well for disruptive factor identification.
Twenty-seven potential disruptive factors were discovered, categorized, and discussed.
Consumer Delphi delivered perspectives on the likelihood and impact of the disruptive factors.
Introduction
With the increasing presence of high-tech innovations and smart technologies, business activities are exposed to a volatile, uncertain, complex, and ambiguous environment. Any unexpected disruptions in the environment could lead to the malfunction or even collapse of a company, a business, or even an industry. According to Malaska and Holstius (2009), knowledge of the future is a precondition for avoiding unexpected risks and threats, making good decisions and influencing the course of events in a desirable direction. Individuals and organizations should anticipate and prepare proactively rather than respond passively (Godet and Roubelat 1996). With respect to the growing influence of China’s e-commerce industry both domestically and abroad and the interconnectivity among online retailers, payment systems, logistics, and customers, any unexpected disruption in the environment will result in a significant ripple effect on many interrelated businesses, industries, and even economies. The objective of the research article is to raise the awareness of corporate executives, retailers, and consumers of the possible disruptions that could impede the growth and trigger the regression of China’s e-commerce industry in the future, and meanwhile demonstrate a novel approach in identifying and analyzing potential disruptive factors, namely, consumer Delphi study. The research is established with the aim to address the following questions:
To fulfill the objective of the research with constructive insights, a combination of qualitative and quantitative methods is utilized in the form of horizon scanning, social, technological, economic, ecological, political, and value (STEEPV) analysis, and Delphi study. Horizon scanning through literature review collects futures signals and driving forces indicating the signs of disruptions and organized in the form of a STEEPV table, whereas Delphi study solicits the quantitative evaluations and qualitative reasonings on the degree of the likelihood and the impact of the collected factors from an expert panel. The time horizon of the study is stretched into the year of 2030, considering the rapid expansion of online commerce activities and the dynamism invigorated with the increasing presence of high-tech innovations and smart technologies. A time span of fifteen years laden with a rich amount of possibilities incurred by business digitalization is capable of ushering in a new era distinctive from today.
The Case of E-commerce in China
Traditionally, China supplied the world with manufactured goods featured with cheaper prices and lower product quality. However, with the emergence and continuous influence of local B2C e-tailers, such as Alibaba’s T-mall and JD.com, propelled by the maturation of the Internet, logistic and payment infrastructure, increasing ownership of mobile devices, the impact of social media, and the growing trust in online transactions, China has transformed into the world’s largest e-commerce market, amounting to US$630 billion of total retail sales in 2015 (K. W. Wang et al. 2016). According to Statistics (2017), the number of China’s Internet users has gone up to more than one billion in 2017 and is estimated to reach 1.14 billion in 2022.
To date, as the disposable income of China’s middle and upper-middle class grows, online consumers have been obsessed with the purchase of overseas goods, leading to an explosive emergence of cross-border e-commerce. Eager for the latest models, better prices, and higher product quality, e-consumers are increasingly willing to purchase products directly from foreign retailers and suppliers (Xia 2017). Online shoppers on cross-border e-commerce are feeling to be protected from counterfeit goods that often pass for offshore brands in the second-tier cities and rural areas (Xia 2017). Seeing the opportunities and the size of the market, an increasing number of offshore brands has joined the force to compete for the e-commerce market in China.
It is projected that China’s e-commerce industry will continue to rise (eMarketer 2016). However, while online retailers and suppliers are focusing on boosting revenues and maximizing profits, the potential risks and threats stemming from the dramatic growth of technological innovations and cultural change, to name but a few, tend to be overlooked. Most importantly, these risks might jeopardize the continuous development of the industry and even affect the profitability of the offshore brands if no proper precautions are being made. Thus, the current research aims to make an exploration of the possible future threats and meanwhile bridge the research gap in the academia.
Theoretical Framework
Advancement of technology is correlated with business model reinvention (Bouwman et al. 2012). Anticipating potential threats and risks in the operational environment contributes to the construction of robust business models. It is claimed that companies with an above-average life span are extremely sensitive to the external environment and capable of adjusting themselves to the changes (De Geus 1997). Companies that reacted quickly to unexpected changes in the environment used to survive and succeed (Barker 1993). However, the traditional managerial model does not fit the dynamics of today’s business realm. In the fast-changing environment with the winner-takes-all market dynamics, managers not only operate the day-to-day running of the business but also need to be aware of the incremental and radical changes in the external environment to avoid unexpected risks and threats.
Corporate foresight, a business-oriented futures research activity, enables corporate executives to anticipate and act uncertainties in the environment. According to Rohrbeck (2015), corporate foresight is a participatory process of identifying, observing, and interpreting the factors in the external environment that induce changes, triggering appropriate corporate responses, and mobilizing joint efforts to steer toward a desirable future. Carrying out corporate foresight helps firms gain insights into changes in the environment, contribute to the reduction of uncertainty, facilitate organizational learning, and shape the future through influencing other players (Rohrbeck 2013). In corporate foresight, wild cards and weak signals are two critical factors, residing outside the familiar range of a company, embedded with the potential to chart the course of events to an unexpected direction (Mendonça et al. 2004; Wilenius 2008).
Wild Cards
Definitions and expectedness
In futures studies, wild cards refer to rare, surprising, and unexpected events (Heinonen 2013). In contrast with trends and megatrends (i.e., continuities), they represent unexpectedness, disruptions, structural breaks, unprecedented developments, and bifurcations (Mendonça et al. 2004). The terms of black swan and the unknown unknowns, conveying the same message of wild cards, are often used interchangeably with the concept in different contexts. In fact, the concept of wild cards did not gain the attention of futures researchers until in the 1980s when an increasing amount of occurrence of unexpected events with large-scale impact took place, such as the oil crises in 1974 and 1979 (Heinonen 2013). Apart from that, events such as 9/11 in 2001, global financial crisis in 2008, Fukushima nuclear accident in 2011, and the Russian intervention in Crimea in 2014 are all of relevance to wild cards that somehow changed the rules of the game.
Over the course of time, academic scholars and business practitioners from different disciplines perceive wild cards in a slightly different fashion, which often causes confusion and denotes minor disagreement in perception. From the terminological point of view, Ansoff (1975) developed the term strategic surprises to indicate unfamiliar, sudden, and urgent changes that either endanger the profitability of the company or result in a lot of fleeting opportunities. Wack (1985) described the oil crisis that caught most of the western countries and oil suppliers off guard in 1973 as one of the rapids, which had been used equivalently with disruptions and ruptures. In 1992, BIPE Conseil, Copenhagen Institute for Futures Studies, and Institute for the Future suggested that “a wild card refers to a future development of an event with a relatively low probability in occurrence, but a likely high impact on the conduct of business.” Petersen (1997) and Cornish (2003) depicted wild cards as surprising and startling events. Petersen (1997) emphasized that wild cards may trigger a chain of events much worse than the initial one, and oftentimes leave less time for people to react. Meanwhile, Cornish (2003) pointed out that surprising events were not always negative and could be favorable and invaluable at times. In the same year, Dewar (2003) introduced wild card scenarios and proposed that not all surprises in the future were unpredictable. Some surprises can be anticipated and even avoided through scenario planning (Dewar 2003). Mendonça et al. (2004) stated that a wild card could be described as an incident whose occurrence appeared to be improbable, but which would breed immediate and far-reaching consequences for organization stakeholders if it were to happen. They further asserted that wild cards were sudden incidents that would make a nonlinear turning point in the development of a trend or a social system (see also Kuosa 2010). Taleb (2007) used black swan as a metaphor for wild card and further explicated the concept as surprising, unusual, improbable events or outliers that carried massive impacts, and were explainable only in hindsight.
The definitions of the wild cards presented by the aforementioned futurists seem to be analogous to one another. In fact, minor differences can still be spotted. In light of the diversity in lexicon, we can see there is a clear disagreement on the duration of wild cards. In other words, the terms of incidents, events, surprises, and rapids communicate the abruptness and shorter duration, which implies that wild cards tend to catch people off guard leaving little time for people to prepare for. On the other hand, “future development,” “trend changes,” and “wild card scenarios” refer to a longer duration or a gradual process in maturation, indicating that surprises can be predicted, anticipated, and even evaded if proper measures are taken beforehand (see Hiltunen 2006).
According to Heinonen (2013), the future exists in the present, that is, the seeds of the future are sown in the ground of today (Heinonen 2013). The barrier to understand the wild cards relates to our blindness or arrogance that prompts us to believe a given event to be very improbable or even impossible (Steinmüller 2007). The poor system in communication and information flows that is deeply rooted in the organizations oftentimes causes failure in early detection (Turoff et al. 2013). In this article, we presume that wild cards do not emerge abruptly but experience a brewing period before maturation, and with the detection of weak signals, they are identifiable. Admittedly, wild cards can also be positive (i.e., a rapid progress in biotechnology prolonging life expectancy), imaginable, and anticipated (Steinmüller 2007).
Dimensions and analysis procedure
Wild cards can be classified from different perspectives. According to Steinmüller (2007), four aspects need to be clarified in wild card management, namely, topic, reach, plausibility, and time scale, described as follows:
As can be seen, wild card forecasting is a systemic, multidimensional, challenging task. Organizations may set up their own policies, procedures, and measurement scales in wild card forecast, depending on the nature of the project, trade of the organization, information source, and strategic importance. However, according to Petersen et al. (2009), there is a general procedure to follow for wild card management, which can be broken down into four major steps, detailed as follows:
Weak Signals
Definition
The history of weak signals can be traced back to the 1970s when Igor Ansoff, a mathematician and a business manager, introduced the idea of weak signal management as an alternative to strategic planning. He emphasized the importance of weak signals, and proposed that the study of strong signals, namely, trends and megatrends, was not sufficient enough to make strategic plans, because the trend- and mega-trend-postulated strategic planning was not capable of coping with future surprises (Martinet 2010; see also Wack 1985). According to Ansoff (1975, 1980), weak signals are random, disconnected pieces of information that initially appear to be background noises but constitute a significant part of a larger pattern when inspected through different lens or by connecting them with other pieces of information. They are seen as the nutshell of unstructured leads on the external environment and are hidden among the “noises” of the sense-making patterns (Heupel and von Juterzenka 2015). Weak signals represent the first indications of paradigm shifts, futures trends, and discontinuities (Koivisto, Kulmala, and Gotcheva 2016). They are the symptoms of a possible future change (Holopainen and Toivonen 2012) or the early signs of a possible but not confirmed turn of events (Saritas and Smith 2011).
One of the most straightforward and well-structured definitions of weak signals is developed by Sirkka Heinonen, a professor at Finland Futures Research Centre. Heinonen (2013) proposes weak signals are the first indications of a possibly emerging trend, and the first expression of change, the impact of which could change the course of events to a different direction. Heinonen and Hiltunen (2012) further state that weak signals point to not only rising trends, or megatrends, but also wild cards. Weak signals are synonymous with the expressions of early indicators, early warning signs, emerging issues, germs (Masini and Vasquez 2000), and seeds of change.
Criteria and sources
It has been widely admitted that scanning for weak signals is a critical process for wild card management (Mendonça et al. 2004). However, the question lies in what can be counted as a weak signal and where to spot them. Of all the four factors in environmental scanning—trends, megatrends, weak signals, and wild cards—weak signals are the most obscure, mysterious, and secretive actors, yet carrying significant weight in foresight processes and findings. There is a three-layer filtering system that tends to lead to the failure or unfulfilling results in weak signal administration, namely, surveillance filter, mental filter, and power filter (Ansoff and McDonnell 1984). Surveillance filter failure constantly takes place because there are always gaps in the surveillance system of organizations. Some signals pass without being detected. Mental filter failure refers to the incapability of making sense of a given signal. When a signal is successfully caught on the surveillance system, it will be discounted simply because it does not conform to the established frames of reference or the mental models of the organization (Mendonça et al. 2012). Power filter connects with decision-making (Ansoff and McDonnell 1984). Even though a weak signal is detected and analyzed, managers or executives choose not to use it, because in reference to their professional experiences, the given signal does not serve the best interest of the company (Holopainen and Toivonen 2012). Based on recent literature on sustainability transitions, a fourth filter, the institutional filter could be added to the list. It covers regulations and policies, routinized practices, and information systems within and around the organization that constitute a regime that tends to maintain certain technological or mental lock-in, and thus maintain the state as business as usual (Geels and Schot 2007; Schönach et al. 2017).
The failure of the second and third filters can be overcome provided that people act cautiously and meticulously by taking the scanned signals into serious deliberation, road-mapping their development paths and connecting them with the strategic management. However, the surveillance filter is the one that oftentimes produces complexity and confusions. According to Hiltunen (2010), if one of the following reactions to a piece of futures information takes place, the information should be counted as a weak signal:
Your colleagues laugh about it,
Your colleagues oppose it,
People wonder about it,
No one has heard about it, and
No one talks about it as it is a taboo.
If several points of this list apply, a weak signal is definitely found. If, following Heinonen’s (2013) definition, we consider the beginning of trend changes as weak signals, one more feature could be added:
Your colleagues regard a small trend change as meaningless, random noise
Having discussed the criteria of weak signals, it begs the question what sources are the best spot to look for them? To look for weak signals and anticipate the futures of change, it is of necessity to understand the dynamics and the logic of change. According to Choo (2015), information life cycle of an emerging issue consists of six stages: idea creation, elite awareness, popular awareness, government awareness, procedural routinization, and record keeping, as can be seen in Figure 1. The first two stages of idea creation and elite awareness are regarded as the exact places for weak signal identification, because they are the phases that ideas are being exposed to the public for the first time (Hiltunen 2010).
Information life cycle of weak signals.
Furthermore, Choo (1995) categorizes the sources of information into three broad divisions: human sources (i.e., internal and external sources), textual sources (i.e., published sources and internal documents), and online sources (i.e., online databases and blogs). According to the findings from a survey targeted to futurists (Hiltunen 2008), human sources such as scientists/researchers, futurists, and colleagues are considered as the major sources for weak signal scanning, followed by textual sources of academic/scientific journals, reports of research institutes and popular science and magazines, and online sources of organizational webpages, electronic journals, discussion groups, and blogs. Meanwhile, Day and Shoemaker (2005) underline that online platforms are as well good sources as published articles for scanning future signs of changes. In the present article, a combination of sources has been utilized for weak signal scanning with the electronic ones being stressed.
To sum up, wild cards bearing the feature of discontinuity, disruption, and bifurcation, be it positive or negative in influence, are capable of altering one’s expectation and charting the course of events to an unexpected direction. Weak signals in this article are regarded in a narrow sense as the first signs of future surprising changes; they herald the development paths of the wild cards, and thus carry disruptive features. Considering the properties of both concepts, we are inclined to merge them into one generic concept in the present article and termed it as disruptive factor. The concept originates from technology foresight, where disruptive technologies and disruptive innovations are frequently studied, but we are also interested in social, cultural, economic, and ecological factors. Recent discussion has emphasized wider disruptive transformation (e.g., Cagnin et al. 2013), which is here understood as being driven by more specific disruptive factors. A detailed illustration of the key concepts of the study related to the subject of the e-commerce in China can be seen in Figure 2.
Illustration of the key concepts of this study.
Materials and Method
Horizon Scanning with the STEEPV Framework
Sutherland and Woodroof (2009) state that horizon scanning is a systematic approach to identify and improve preparedness to potential threats and opportunities. Its objective is to discover the changes emerging outside the established mental models of the organizations (Schultz 2015). There are three approaches to gather information via horizon scanning, namely, literature searches, expert workshops, and open fora (Sutherland and Woodroof 2009). For literature review, researchers tap into a series of published, peer-reviewed evidence, which leads to a high degree of data quality. The method is considered as one of the efficient ways to retrieve data from a range of sources. This approach, however, sometimes leads to a backward looking predicament. Expert workshops refers to a group of experts who are organized to deliberate on a specific topic with the outcome strictly dependent on the expertise of the participants. Open fora takes the form of crowdsourcing, which allows researchers to collect information from a broad spectrum of people and utilize the wisdom of crowds, but the quality can least be assured. The present study adopts the approach of literature review for its efficiency and quality. The drawback of backward looking featured in literature review can be mitigated through restricting the evidence to be the ones published after the year of 2010. A more philosophical question is whether the specialized academic publications are the most relevant source for searching weak signals and wild cards, since in this late-modern era experts’ and lay-people’s knowledge become increasingly mixed and contextual (Bogner and Menz 2009, 3–5; Giddens 1991; Varho and Huutoniemi 2014). We, therefore, looked for weak signals and wild cards also from nonacademic published sources, however, not from social media.
The fundamental principle of horizon scanning, according to Sutherland and Woodroof (2009) is to broaden the horizon and avoid tunnel visions in foresight activities. Therefore, collecting data in a systematic and broad perspective is essential to the quality of the outcome. STEEPV framework offers structure to a learning process (Loveridge 2002) and stands for six major themes, namely, social, technological, economic, ecological, political, and value factors. In the study, STEEPV framework will be utilized to classify the potential disruptive factors collected pertinent to China’s e-commerce industry. As a result, twenty-seven factors have been spotted from human, textual, and online sources. A detailed overview can be seen in the section “Research Findings.”
From Expert Delphi to Consumer Delphi
Delphi method, originated in the 1950s, was invented by Helmer, Dalkey, and Rescher from the U.S. military think tank known as RAND Corporation (see Dalkey and Helmer 1962; Helmer and Rescher 1958). Delphi studies aim to predict and/or explore alternative futures, their probability of occurrence and desirability drawing on the expertise of the participants (Bell 1997). The method explores the diversity of expert viewpoints through a structured communication process (Linstone & Turoff, 1975). Delphi study is both a qualitative and quantitative method characterized by the iterative and anonymous process in data collection with questionnaires, and the feedback to the respondents.
In Delphi study, a critical factor that influences the validity and reliability of the findings is the selection of experts. According to Varho and Huutoniemi (2014), expertise is both cognitively and socially determined. Expertise in the cognitive sense refers to the acquired knowledge and skills possessed by an expert on the content of an inquiry. It includes both propositional knowledge and tacit knowledge, which are acquired through education, research, experience, or any other form of cognitive refinement, such as practical know-how. On the other hand, expertise in the social property relates to the possession of an expert status in the eyes of others. It is obtained mainly through the demonstration of aptitude and competence in a given domain. Oftentimes, it is the educational degrees, higher professions, and leading management positions that determine the status of being an expert. However, one should be aware that the acquisition of expert status does not always reflect the possession of the real cognitive knowledge, and so is the opposite (Varho and Huutoniemi 2014.)
Delphi method is traditionally used in studies on expert views of the future. However, in this study, we intend to try the method for a consumer panel. This is relevant from three perspectives.
First, in the times of easy Internet searches, an overflow of information, mistrust of technical know-how, and “alternative facts,” public demand for the social status of expertise is going down. Bogner and Menz (2009, 5) describe the deterioration of social expertise as “laicization”:
The “laicization” of the expert changes the relationship of trust between lay people and experts: expertise must be increasingly stage-managed to gain acceptance.
Even if we would still believe in the traditional scientific values of objective knowledge and that established experts tend to “win” lay-people in estimating future events, lay-people are in fact increasingly affected by their peers and the wisdom (however illusory) of the crowds. So it is interesting to know what lay-people think about the future, be they technically plausible or not.
Second, lay-people might lack social expertise, but they, as routinized e-commerce consumers, could have a lot of practical know-how that established and specialized experts do not possess (Beck 1992). A recent, interesting Delphi study by Morgan et al. (2016) reports that clinicians’ and researchers’ views of coping strategies out of moderate anxiety were strikingly different from the views of lay-people, who actually had suffered from anxiety and were at the time of the study working as patient advocacies in nongovernmental organizations (NGOs). This way, consumer Delphi brings in a new viewpoint, which could be helpful in making conclusions about the social plausibility of weak signals and wild cards.
Third, consumer studies, on the other hand, seem to mostly use standardized surveys, qualitative one-round interviews, or focus groups. Although structured questionnaires are used, Delphi study is essentially following the principles of qualitative studies, as the expert panel is not selected through random sampling. Delphi technique adds a second round of inquiry to data collection and provides feedback to the consumers with the answers of their peers. In this regard, Delphi study is very close to a focus group discussion except that face-to-face contact between panelists is omitted. This prevents the emergence of the bandwagon effect and the dominance of single strong opinion leaders. In Table 1, the consumer Delphi is compared with standard surveys, focus groups, and qualitative interviews. Using the Delphi technique in a study of consumer views of the future has novelty value.
Comparing Consumer Delphi with Individual Interviews, Focus Groups, and Surveys.
The purpose of this study is not to demonstrate that consumer opinion would win expert views in terms of accuracy of prediction but to extend futures research methods to the field of consumer studies to advance the field.
Expertise Matrix in a Consumer Panel
The expertise matrix was developed by Kuusi et al. (2006) and presented in an article regarding the future of the Finnish health care system. The objective to create expertise matrix in Delphi studies is to ensure that representative experts from all the pertinent domains of the subject under study are invited, the viewpoints of whom are included and taken into consideration. As Kuusi et al. (2006) points out, different fields have their own knowledge and expertise to offer in resolving a given issue. To date, Delphi studies carried out in Finland tend to use the expertise matrix regularly to ensure a good coverage of sources and saturation of materials. That is to say, no novel or fresh viewpoints would pop up if an additional respondent is added to the panel (Varho and Tapio 2013).
For the present research, there are three spheres to be of relevance to determining the diversity of the expert panel: cognitive expertise in terms of years of e-shopping experience, social expertise indicated by educational level, and sociodemographic background reflected in gender, age, and income level. In the first round, thirty-two experts participated in the research, showing a good enough diversity of expertise for a consumer Delphi study, young adults under the age of thirty being emphasized (Table 2). The saturation criterion was met on the first round, but the second-round material was a little too scarce, even though Ziglio’s (1996, 14) review suggests that with rather homogeneous panels, ten to fifteen individuals are acceptable.
The Consumer Panel Expertise Matrix.
The research was not developed to achieve statistical representation, but to explore the diversity of viewpoints from a consumer panel. In the research, the family, friends, and relatives of the first author in China were contacted and invited to the survey forming the first tier of the panelists. After completing the questionnaire, the respondents were encouraged to invite their family, friends, and relatives to the study there on. This type of “snowballing” sample is typically used for focus group studies.
The consumer Delphi was conducted in two rounds. In the introduction of the questionnaire, the background and objective of the research were presented and clarified, as well as the instructions to proceed with the subsequent sections. The main section was made of thirty questions based on the observed disruptive factors with one question pointing to one corresponding factor, for which the panelists were invited to evaluate the probability and the impact on a scale of 1 to 5 with 1 being the lowest in value and 5 the highest. Of the thirty questions, three were not considered disruptive factors during closer examination after data gathering. Therefore, twenty-seven disruptive factors will be qualified for further analysis. In addition to the quantitative data, qualitative data were obtained. Two repetitive open questions “Please justify your evaluation” were posted under each main question for the participants to express the logics and rationales behind their quantitative evaluations, respectively, on the degree of probability and the impact of the factor concerned. The questionnaire concluded with the panelists’ personal information regarding cognitive expertise, social expertise, and sociodemographic data. The idea was to ensure the diversity and heterogeneity of the panel. If the saturation level of the panel was not reached after the first round of the Delphi survey, further effort would be made to make contact with the missing panelists. On completing the questionnaire, two second-year master’s degree students majoring in Futures Studies were invited to make a test run to ensure that the layout and the content of the questionnaire were straightforward with no ambiguous or misleading expressions embedded.
At the end, thirty-two responses were generated from the first-round consumer Delphi. After the data were analyzed, nineteen panelists who had expressed their interest in participating in the second-round Delphi were contacted. In the second-round Delphi, the main questions remained the same except that the open questions inquired in the first round were removed, because the focus of the second-round study was to monitor the traces of change or deviations in the opinions of the panel after they studied the feedback report with the frequency bars per question and qualitative arguments expressed by fellow panelists in the first round. This was also the way to avoid respondent fatigue, the common pitfall of opinion surveys. Eventually, thirteen responses were received from the second round.
Research Findings
Disruptive Factors from Horizon Scanning
As mentioned in the section “Materials and Methods,” the concepts of weak signals and wild cards were synthesized into one generic term, in our case, called the “disruptive factors.” The disruptive factors were collected through horizon scanning through textual, online, and human sources. Different futurists categorize information sources differently. It varies by research topic (Hiltunen 2008) and source type (Choo 1995), for instance. In our article, Choo’s categorization by source type is applied to strengthen the breadth of data. According to Choo (1995), textual sources stand for published academic/scientific papers and internal documents/reports, and human sources represent internal communications and personal networks, while online sources indicate online databases. Admittedly, Internet and blogs have gained popularity among the futurists, and considered to be good sources for horizon scanning (Day and Shoemaker 2005; Hiltunen 2008). Meanwhile, focusing on published sources makes data gathering more transparent. Table 3 is a detailed overview of the sources from which data were withdrawn.
Sources of Horizon Scanning.
Note. DHL = Dalsey Hillblom Lynn; IPR = Intellectual Property Rights.
In scanning for disruptive factors, a list of evaluation criteria was developed drawing on the logic of the five questions proposed by Hiltunen:
Is the potential factor directly termed as a weak signal or a wild card in the article?
Does the potential factor bear any disruptive features?
Does the potential factor make you wonder?
Does the potential factor catch you surprised?
If the factor is commonly believed to be unlikely to happen, it should be counted in.
At the end, twenty-seven potential disruptive factors were qualified, and put into the STEEPV framework, as shown in Table 4. Of the twenty-seven collected factors, eleven factors were counted as weak signals and sixteen factors as wild cards in accordance to either the claimed features in the corresponding articles or wild conjectures on the observed weak signals, trends, and megatrends. An overview of the disruptors categorized by weak signals and wild cards, as well as the interpretations and the uncertainties of the factors, is demonstrated in Tables 5 and 6.
Disruptive Factors in the STEEPV Framework.
Note. STEEPV = social, technological, economic, ecological, political, and value; IT = information technology; P2P = peer-to-peer.
Disruptive Factors in the Format of Weak Signals.
Disruptive Factors in the Format of Wild Cards.
Note. IT = information technology.
Findings after the Two-Round Consumer Delphi Survey
In preparing for the research, both versions of the questionnaire were developed using the online survey tool Webropol. In the questionnaire, there were twenty-seven main questions with each to elicit answers from two perspectives: likelihood and impact. To interpret the result in a precise and logical fashion, median values were further classified into three groups of high, moderate, and low (5; 4,5; 4), (3,5; 3; 2,5), and (2; 1,5; 1), respectively. Figures 3 and 4 display the results of Delphi Round 1 and Delphi Round 2, respectively.
Likelihood-impact matrix of the disruptive factors for consumer Delphi Round 1 (on a scale of 1–5).
Likelihood-impact matrix of the disruptive factors for consumer Delphi Round 2 (on a scale of 1–5).
In Figure 3, the disruptive factors are clustered into eight groups with Consumer information leakage in Group 1 viewed to be the highest in both likelihood and the impact according to the expert panel, followed by the ones in Group 2, namely, Alipay monopolizing online payment in China and Introducing new laws to regulate the misconducts of online retailers. In contrast, the factors situated in Group 8 Establishing web stores for senior citizens, E-payment system collapse, and Apple pay monopolizing online payment in China are listed as the lowest in both the likelihood and the impact. It was interesting to find that the consumers on average seemed to view a correlation between impact and likelihood of the factors.
On the second round, the number of groups dropped from eight to six, indicating that the opinions of the panel tended to converge to a relative consensus. According to the graph, Alipay monopolizing online payment in China overtakes Consumer information leakage, topping the chart as the highest in likelihood and impact, with Consumer information leakage downgraded to Group 3 (moderate-likelihood and high-impact) and Introducing new laws to regulate the misconducts of online retailers that appears to be high-likelihood and high-impact in the first round listed as moderate-likelihood and high-impact in Group 4. As for the factors at the lowest value for both aspects, Apple pay monopolizing online payment in China remains unchanged with a new arrival Terrorist attack, making Group 6.
As demonstrated in Figures 3 and 4, Alipay monopolizing online payment in China is projected to be the highest in likelihood and impact in both rounds of the study. If the online system of Alipay were hacked, the marketplace of the e-commerce would become chaotic. In addition to the instance of a massive amount of online payments being postponed, consumer’s online wallets might operate to the advantage of the hackers in a more severe situation. However, the chance of the online systems being hacked remains fairly low in likelihood and impact of the first round, and moderate for the second round, according to the panelists. For the factors classified as being moderate in likelihood and impact, most of them remained unchanged between the two rounds, namely, Sluggish growth of Chinese economy, Introducing new taxes to limit consumption, Battling against counterfeits, Drone delivery service, Wearable devices, P2P lending, and Renaissance in the brick-and-mortar model. Apart from that, the number of factors under this category for the second round aggregates up to eighteen from thirteen in the corresponding section in the first round, suggesting that the opinions of the experts in the second round were influenced by the feedback. After reviewing the evaluations made by the peers, the panelists responded rather carefully to avoid making bold conjectures or extrapolations.
Surprisingly, the category of low-likelihood and high-impact that wild cards represent maintains void for both rounds of the survey. Yet six factors represent the genre of low-likelihood/moderate-impact. They are Internet submarine cable crash, Hackers attack, Delivery men’s strike, Sharing economy, National IT infrastructure collapse, and Military actions between China and neighboring countries, in which Delivery men’s strike appears in the same category from both rounds. These factors can be regarded to have the greatest potential to take the shape of the wild cards. Furthermore, it is rather captivating to recognize that the panelists believe “terrorist attack” to be a low-likelihood and low-impact event for the second round of the study, which moves directly from the moderate-likelihood and moderate-impact group in the first-round Delphi.
To study how stretched the opinions of the panelists are from the averages of the evaluations in both the consumer Delphi rounds, standard deviations were calculated, as demonstrated in Table 7. According to the statistics, distribution of the evaluations on the likelihood and the impact of the twenty-seven disruptive factors from both rounds of the Delphi are rather squeezed than stretched. That is, the range of the opinions is very centered around the averages indicating that the panel reached a relative consensus on the topic.
An Overview of Statistical Indicators of Mean and Standard Deviation on the Quantitative Data in Both Rounds of the Consumer Delphi (Sorted by the Mean Values of Impact).
Note. Sorted by mean values of impact. STEEPV = social, technological, economic, ecological, political, and value. IT = information technology.
In the first round of the Delphi study, the qualitative question “Please justify your answer” was inquired to the panelists for evaluating the likelihood and the impact separately. The qualitative questions were determined to be optional in the questionnaire on the grounds that optional questions represent relaxation, making the respondents feel comfortable to deliver answers when they have points to make. In the present article, the qualitative answers to the factors that were rated high and low in likelihood will be presented, the reason being that the factors listed as high-likelihood help construct the image of the future of the dynamics of the e-commerce industry in China, and the ones graded to be low fit into the category of potential weak signals or wild cards existing outside the peripheral frame of cognition. Such answers as “It will happen,” “It is a natural trend,” and “It is possible,” which are considered to be merely repetitions of the quantitative estimates or tautology, were excluded from the analysis.
In the high-likelihood/high-impact category, “consumer information leakage,” “Alipay monopolizing online payment in China” and “Introducing new laws to regulate the misconducts of online retailers” are demonstrated in Table 8. All three factors are rated high in both likelihood and impact. Alipay is commented to be the most favored and popular online payment system in China, while new laws to regulate online marketplace are strongly anticipated as product misrepresentations and misleading information are scattered around rampantly. From the perspective of the impact, all three factors carry the highest weight. If consumer information cannot be stored securely in the database of the e-commerce operators, it is believed that the profitability and sustainability of the business will be threatened. Besides, if Alipay monopolizes the online payment, the innovation in the industry is predicted to experience a downward trajectory. Furthermore, the online marketplace will be changed significantly if the gap of the regulatory system specific to the e-commerce industry is merged.
Illustrative Qualitative Answers for High-likelihood and High-impact Factors.
Note. “n/a” refers to no corresponding valid data observed.
Toxicity of electronics and Green packaging are listed in the category of high-likelihood/moderate-impact as can be seen in Table 9. As the panelists point out, the harmful effects of electronic devices are taking shape on the users to a serious extent because consumers become increasingly reliant on the electronics in life and work. However, the impact of toxic tech on online shopping is not phenomenal. As said, people would take preventive measures to reduce the negative effects. Online shopping remains as is. Regarding Green packaging, it can be seen that environmental degradation in China has raised the awareness of the public, as having product packed in an environmental friendly manner is expected.
Illustrative Qualitative Answers for High-likelihood and Moderate-impact Factors.
Note. “n/a” refers to no corresponding valid data observed.
As demonstrated in Table 10, Internet submarine cable crash, Hackers attack, Delivery men’s strike, and Sharing economy are categorized under low-likelihood/moderate-impact. It is predicted that delivery men calling for strike action is rarely possible because there is an oversupply of workforce in China. If one refuses to work, someone else will step in right away. Another perspective has it that the landscape of the logistic industry will change, but it is not attained through strikes but gradual refinement in legislation. Sharing economy, being least possible to occur, is stated to be against the values rooted deep in the Chinese culture, as the concept of “sharing” is not worshipped. People tend to appreciate “owning” or “privatizing.” Concerning the impact, the crash of the undersea cable will not exert huge influence on the Internet connection. As stated, most of the cities in China rely principally on onshore cables. When it comes to Hackers attack, according to the panelists, the rare occasion is predicted to wield considerable influence on the e-commerce industry, as consumers will feel insecure and reluctant to shop from the web stores that have been hacked. Meanwhile, Hackers attack will influence the online shoppers who tend to make an aggregate of orders at a time. It is enunciated that the ones who shop less will be affected less.
Illustrative Qualitative Answers for Low-likelihood and Moderate-impact Factors.
Note. “n/a” refers to no corresponding valid data observed.
In terms of Delivery men’s strike, the impact is evaluated to be significant as the credibility and reputation of the logistic firm concerned will be seriously jeopardized that the profitability would be deflated. In fact, according to the qualitative answers of the panelists, it seems that Hackers attack and Delivery men’s strike would yield widespread, severe influence on the e-commerce industry when occurring. Controversially, the ratings on average point to the value of “moderate.” It may relate to the Chinese culture that people tend to avoid extreme responses in survey questions (R. Wang et al. 2008).
Establishing web stores for senior citizens, E-payment system collapse, and Apple pay monopolizing online payment in China are rated as the events with being low in probability and impact, detailed in Table 11. As people age, physical mobility decreases. Senior citizens might be the group of people whose demand for online shopping is more justified and logical than that of the youth. However, according to the opinions of the panelists, it is least possible to create an ad hoc web store for the senior, as the elderly are characterized by least proficiency in computer techniques and being the perfect subjects lured into online traps. As some panelists point out, senior citizens are niche market for online shopping but the last target group to cater for at the moment. In terms of online payment system breakdown, it is believed to be least possible in occurrence. If it occurs, it is predicted to lapse only for a limited period of time before resuming to work again. Meanwhile, the impact is deemed to be minimal since consumers could choose to pay at a later time, as articulated by one panelist. According to the evaluations made by the experts, Apple pay is considered to be unlikely to monopolize the online payment industry in China, mainly because of the existence of the well penetrated, mighty rival “Alipay.” Besides, assimilation into the market of China tends to be costly and time-consuming for Apple pay, indicated by one respondent.
Illustrative Qualitative Answers for Low-Likelihood and Low-Impact Factors.
Note. “n/a” refers to no corresponding valid data observed.
Discussion
Main Research Findings
With this article, we aim to contribute to the global discourse on the radical upsurge and expansion of China’s e-commerce industry by adhering to a combined mode of horizon scanning, consumer Delphi, and the theoretical discussion on disruptive transformation. Twenty-seven disruptive factors were discovered through literature review, representing weak signals and wild cards. As a result of both rounds of Delphi study, six factors are estimated to be low probability in occurrence and moderate in impact, namely, Internet submarine cable crash, Hackers attack, Delivery men’s strike, Sharing economy, National IT infrastructure collapse, and Military actions between China and neighboring countries, and four factors Establishing web stores for senior citizens, E-payment system collapse, Apple pay monopolizing online payment in China, and Terrorist attack are classified as low in both likelihood and impact.
According to the research findings on the cultural characteristics in responding to Likert-scale questions in surveys, the Chinese, believing in interpersonal harmony and less emphasis on individual opinions, demonstrate a greater preference for midpoints and less preference for extreme values (R. Wang et al. 2008). Therefore, we believe the factors Internet submarine cable crash, Hackers attack, Delivery men’s strike, and Sharing economy, rated low in likelihood and moderate in impact, are the result of the cultural paradigm concerned, as they demonstrate a stark contrast between quantitative ratings and qualitative arguments in the first round of the Delphi. Owing to the cultural norms, the four factors are evaluated to be mild in impact but in fact qualified for the category of wild cards or surprising events by 2030, and so are the two factors National IT infrastructure collapse and Military actions between China and neighboring countries that are scored relatively high in the second round.
Furthermore, Establishing web stores for senior citizens, E-payment system collapse, Apple pay monopolizing online payment in China, and Terrorist attack rated low in both probability and impact might be potential weak signals. According to Mendonça et al. (2012), the fact that signals are identified in environmental scanning but typically disconnected from strategic management is due to the established mental models or the frames of reference of the evaluators concerned. In our case, consumers holding a strong belief in the current physical infrastructure of online transactions, e-payment systems, and national security might override the chance of the occurrence and the impact of the relevant factors.
In addition, the research outcome reveals that some of the disruptive factor candidates are perceived to be constant or stable in the minds of the consumers, the strike of submarine cable crash, hacker’s attack, delivery men’s strike, e-payment system collapse, and Apple Pay monopolizing China’s online payment system, for instance, delivering a clear message of consumer’s unwavering trust in China’s e-commerce infrastructure. At the same time, the factors that are considered to be volatile and vulnerable such as consumer information leakage, toxicity of electronics, and green packaging help identify potential market opportunities and distinctive services to be further developed to the business players in the e-commerce field. Thus, we believe that Delphi approach with a “nonexpert” or consumer panel help unveil fresh perspectives and offer a set of valuable viewpoints to business operators in complementing the understanding of the future risks and opportunities in the external business environment.
Methodological Reflections
It is of importance to point out that this research is by no means to aim for an inclusive study, and it is relatively impossible to identify unexpected, disruptive factors that fit into all categories of electronic businesses. Admittedly, companies that are operated online are distinctive from one another in terms of product portfolio, target market, supply chain, distribution channels, and competitions. The overarching objective of the project is, however, to raise the awareness of the e-commerce business entities in China of the potential threats over the course of online operations, to provide their overseas business counterparts that are planning for an international business expansion into the online marketplace of China with insight and consumer’s perspectives and concerns, and at the same time to demonstrate a preliminary model in risk management, namely, scanning weak signals and wild cards leading up to the identification of potential wild cards through Delphi approach combined with a consumer panel.
Traditionally, expert panels are composed of professionals, scholars, and managers featured with years of practical experiences and renowned educational degrees in the domain under study. Our study turned to regular e-shoppers due to their growing influence and leverage in determining the success of businesses in the highly networked social dynamics. The exploitation of consumer expert panels may result in biased, incomplete findings skewed to one side of the story. But as the academic literature is filled with expert Delphi studies, complementing them with consumer Delphi findings can bring about new viewpoints and perspectives.
The issues dealt within this study are complex and difficult, and comprehending them requires imagination and critical thinking. Based on the results and our impression in general after this exercise, we would conclude that using Delphi technique in a consumer panel brings about more well-argued answers than standardized surveys. The second author has been involved in dozens of Delphi studies, and it seems here that a consumer panel is more likely to end up in consensus than an expert panel, where many experts see the variety of future options more clearly than lay-people and might have stronger opinions. The consensus-building nature might thus work even more effectively in a consumer Delphi. However, the main benefit for Delphi as opposed to one-round surveys or interviewing individuals is the communication and the ability to learn from each other. The traditional goal of consensus has been contested within expert Delphi studies (Steinert 2009; Tapio et al. 2017). Should consensus then be regarded as a virtue or a vice in consumer studies? We would answer that it is neither of these—it is merely a neutral research result. In the future, researches such as the present study is highly suggested to be complemented with an expert Delphi panel. Comparing the findings from the expert panel with the one from the consumer panel will bring out new perspectives and insights.
Conclusion
The emergence and the dramatic development of the e-commerce industry in China are phenomenal and carry a significant weight in contributing to national economic growth. With the increasing presence of high-tech innovations and smart technologies, the global landscape of the online business will continue to transform, evolve, and influence. In the history of China, risk management can be dated back as far as the Warring States Period two thousand years ago. As indicated in an ancient Chinese maxim, one should always prepare for dangers in the times of peace. We believe that anticipating or being sensible of future disruptions, wild cards, and the weak signals foreshadowing them will equip e-commerce operators with ambidextrous capability in coping with the roadblocks along the journey.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
Author Biographies
.