Targeting the very important buyers VIB: A cluster analysis approach

Abstract The Pareto Principle, known as the 80/20 rule, predicts that most sales generate from a minority of buyers. Further, consumer theory stipulates certain shoppers have a preference to consume experience versus tangible goods. Some people value the consumption of experience due to lifestyle and innate factors. Following a cluster analysis approach, this paper collected survey data from 700 consumers (adults that belonged to a consumer panel) to examine these propositions. The results support the propositions and show that the e-market comprises three segments: VIB of Experience, VIB of Material, and Normals. The discriminatory analysis shows that the affluent shoppers of experience diverge from the other segments on psychological and lifestyle factors. In addition, social class helps pinpoint these shoppers. Whereas VIB of Experience primarily belong to the Upper Uppers and the Upper Middles, VIB of Material belong to the Lower Uppers and the generic Middle Class. The results are of value to practitioners that aim to target affluent shoppers of experience versus affluent shoppers of material goods. The results suggest that the affluent e-shoppers are not homogeneous according to their shopping preference and highlight the relevance of the Pareto Principle to segment and target the e-shoppers.


Affluent buyers of experience versus material products
Retail e-commerce is growing. This growth has been facilitated by the COVID pandemic (Watson & Popescu, 2021), reflecting the need to study this evolving market and shed light on its participators. The Pareto Principle predicts that about 80% of sales generates from 20% of shoppers (Westerby & Nortun, 2021). In other words, few shoppers should account for most online sales. Researchers have examined the factors leading to the adoption of e-purchasing, postulating that consumers allocate resources in compliance with their pattern of Internet use. Maignan and Lukas (1997) were the first to require that online behavior be studied according to consumer's views of the Web-as an information source, communication medium, place of consumption, and social system. Bourdeau et al. (2002) highlighted that Internet use is bounded by five values: social, utilitarian, hedonic, learning, and purchasing. Aljukhadar and Senecal (2011) employed the latter theory and found that consumers form three online segments: basic communicators, lurking shoppers, and social thrivers.
Studying the affluent shoppers and analyzing their online behavior is a relevant research field (Kooti et al., 2016). While the Pareto Principle indicates that few consumers (termed Very Important Buyers or VIB) comprise the affluent e-shoppers, how can managers pinpoint these VIB to target them? In other words, what are their characteristics and consumption preference? Product type (i.e., what the consumer is shopping for) helps explain the heterogeneity in the e-shopping behavior (Mallapragada et al., 2016). The literature yet lacks in studies that examine the heterogeneity of shoppers with respect to the products they buy and the amounts they spend online. The current research addresses such gap. Specifically, it aims to address two research questions: RQ1. Do most online sales generate from a minority of consumers? RQ2. Are the affluent e-shoppers homogeneous in their preference to conduct experience versus material purchases?
In effect, research has addressed the question of why consumers use the Internet to purchase. Yet far less research shows how consumers use the online medium to purchase and what do they purchase. Need exists to fathom how the online consumers differ in the purchases they perform. Kim (2018) suggests that people value certain experiential consumption due to innate traits. If, as consumer theory suggests, some shoppers assign high value to experience, they would be more likely to shop for experience goods if they possessed the financial mean. On the same shopping channel, people with divergent product preference and purchase power transact. With the prominence of the Internet and the COVID pandemic, the choice of the online channel to perform mundane purchases is becoming binding. Consumers' heterogeneity according to an array of factors-particularly their product preference and the amounts they spend online-come to be reminiscent.
Whereas the Pareto Principle predicts that most sales are generated by a small percentage of buyers, the literature suggests that practitioners do not apply this straightforward principle. Westerby and Nortun (2021) indicated that marketing executives forgo the Pareto Principle because it is difficult to pinpoint the affluent shoppers, or the small core. Scholars suggest that buyers diverge according to their preference for experience consumption (Kim, 2018). Intangibility is a product characteristic that forms a source of perceived difficulty during information gathering and decision-making (Murray, 1991). Scholars have classified products into tangibles and services, underscoring the particularities of the design and distribution of services (Fitzsimmons & Fitzsimmons, 2004). Van Boven and Gilovich (2003) concluded that the experience-material classification of products, though inexact, is widely acknowledged and applied. The current research builds on this theory by contending that buyers differ in their consumption preference: Some are inclined to conduct experience versus material purchases.
To minimize the discrepancy between academia and practice, this research follows a consumer-revealed segmentation approach (Allred et al., 2006). Using data collected from a consumer panel, the research sheds light on the segments comprising the affluent shoppers. Consumer-revealed segmentation was used to organically detect striking shopper groups and provide insight on each segment motives, characteristics, and needs. It helps the firm attains a strategic advantage over competitors by identifying the segments' unique attitudes and needs, thus transform strategic opportunities into tactical policies. As Hansen (2005) indicate, a crucial benefit derived from conducting segmentation analysis is showing why the segments are different. Firms that follow a resource-based approach benefit from the revealed segments for they allow for an optimal allocation of marketing expenditures. The research goal, captured in the two research questions, is to underscore the segments comprising the affluent e-shoppers and pinpoint the differences between them. This paper comprises the following section. The next section provides an overview of the overarching empirical studies leading to the research questions. Details about the data collection, sample, and analysis are provided in the method section. The results and the discussion are then provided, followed by their implications and directions for future research.

Theory development
A limited number of segmentation studies has been conducted on e-purchase, particularly in the context of product type, highlighting the need of such studies. Online shoppers have been classified into heavy, moderate, and non-shoppers (Forsythe & Shi, 2003). Swinyard and Smith (2003) studied the e-shopper versus non-shopper and identified four segments (shopping lovers, adventuresome explorers, suspicious learners, and business users). These results were replicated by Brengman et al. (2005) and Allred et al. (2006), who verified their 2003 study by classifying e-consumers into holiday shoppers versus non-shoppers. Most of the studies in this domain were based on consumer-revealed segmentation, which allows the segments to be identified based on natural associations observed during data analysis via cluster analysis techniques (Wedel & Kamakura, 2000). Rohm and Swaminathan (2004) proposed four archetypes based on motivation-convenience shoppers, variety seekers, balanced buyers, and store-oriented shoppers. Barnes et al. (2007) studied e-consumers according to psychographic, cultural, and purchase behavior factors, finding three archetypes: Risk-averse doubters, open-minded online shoppers, and reserved information-seekers. Their results suggest that psychological factors are key to discriminate among the archetypes. Jayawardhena et al. (2007) studied e-consumers according to orientation and identified five segments (active, price sensitive, discerning, loyal, and convenience shoppers). Liu et al. (2015) segmented e-shoppers in six categories: economical, active-star, direct, high-loyalty, risk-averse, and credibility-first purchasers. Harris et al. (2017) segmented shoppers who had purchased groceries online and offline based on channel perceived advantage using a sample of 871 participants. They conclude that channel choice is shaped by the perceived disadvantage of the other channel.
The relevance of studying heavy buyers (i.e., affluent shoppers) and reveal their preferences stems from the notion that the marketing-mix strategy of the firm differently affects them. Chiou and Pan (2009) for instance, showed that price and value had stronger impact on satisfaction for typical buyers, while trust had a stronger impact for the heavy buyers.
The theory suggests that the VIB diverge in their preference to purchase experiences. Research indeed has focused on cost and tangibility to categorize products. McDougall and Snetsinger (1990) defined tangibility as the degree to which a product can be envisioned and provide a transparent, specific image prior to purchase. Experience goods are generally intangible while shopping items such as electronics, furniture, and household equipment are tangibles. Whereas scholars have emphasized the need to differentiate between firms based on how intangible the products they offer are (Aljukhadar & Senecal, 2015;Laroche et al., 2005), scholars are yet to differentiate between consumers based on this criterion.
Consumers might be heterogeneous in their shopping orientation. Van Boven and Gilovich (2003) indicated that experiential purchases are made with the intention of acquiring a life experience (an event that one lives through), while material purchases are made with the intention of acquiring a tangible object (kept in one's possession). Their theory calls for considering consumer heterogeneity in shopping preference-experiential versus material.
Some consumers perceive experience goods to be cool, convenient, and indispensable due to innate factors. For one, experience consumption entails high level of social interactions. Consumers emphasizing social interactions will value this type of goods. The need to distinguish the design and delivery of services from tangibles has been stressed, and the complexity of managing services has been highlighted (Fitzsimmons & Fitzsimmons, 2004). Product intangibility is a source of perceived difficulty during information search (Murray, 1991). Because services "intangibility gives consumers a fuzzier and less accurate cognitive representation" and because of the lack of precise and obvious properties available for the consumer (Laroche et al., 2005, p. 253), some will value material purchase. For instance, consumers with concrete thinking style, high need for cognition, or less need for social interaction will value material purchase.
Introducing the "need for touch" concept in an earlier work, Peck et al. (2013) indicate that people are heterogeneous in their need to touch a product to examine and purchase it. Following this reasoning, consumers would be heterogeneous in their disposition to conduct material vs experience purchase on the electronic medium. Online retailers offering material goods are advised to focus on providing shoppers with quality information, whereas those offering services are advised to focus on the e-store's aesthetics, interactive, and personalization features (Aljukhadar & Senecal, 2015). Interactivity and responsiveness are linked to service shopping (Ball et al., 2006). Laroche et al. (2005) stressed the inherent lack of precise and clear qualities in experience shopping. Such lacking would appeal to certain shopperswho consider experience consumption cool and indispensable-and to a lesser degree to other shoppers. Pascual-Miguel et al. (2015) found that product type (digital versus non-digital goods) shapes the relation between perceived risk and e-purchase intent. Fang et al. (2016) studied e-shopper heterogeneity according to task-oriented versus experiential shopping. Vijayasarathy (2003) examined the effect of tangibility and cost on the relation between shopping orientation and e-shopping intent, supporting a direct effect of tangibility and concluding that intent differs according to tangibility because of innate factors (normative beliefs). Individual differences explain online shopping behavior (McElroy et al., 2007).
Earlier studies showed that e-shoppers were more likely to be younger males with high income and Internet experience (Sin & Tse, 2002;Swinyard & Smith, 2003). Nowadays, these results should be scrutinized given the increasing participation of females and the less educated. Research highlighted the significance of elderly shoppers. Sorce et al. (2005) noted that whereas older shoppers search online for fewer products compared to younger shoppers, they spend as much as younger shoppers do. Pandey et al. (2015) suggest that lifestyle and social class help explain shopping orientation. Kooti et al. (2016) found that e-shoppers from wealthy neighborhoods tend to purchase more expensive items, and they do that more frequently. Kumar et al. (2018) characterized affluent shoppers according to spending, visit frequency, and profitability. Their results suggest that while affluent shoppers have higher technical expertise and Internet service adoption, they have lower deal sensitivity. Morisada et al. (2019) showed that behavioral and lifestyle factors indicate the affluent shoppers (i.e., shopper profitability). They found that profitable segments are more likely to comprise consumers that are innovators, brand-conscious, and loyal who regularly use mobile devices. As such, income, education, and other determinants of social class and lifestyle should discriminate the heavy shoppers more strongly than age.
Based on this backdrop, we advance two research questions: RQ1: Does most online sales generate from a minority of consumers?
RQ2: Are the affluent e-shoppers homogeneous in their preference to conduct experience versus material purchases? If not, what are the factors discriminating between them?

Data collection
To address the research questions, we conduct a study on online shoppers gathering data on their shopping preference as well as psychological and demographic factors. The pattern of online product shopping can be obtained using objective or subjective measures. Objective measures that track consumer navigation over an extended period can provide unbiased assessment. Several factors yet favored a self-reported measure.
Firstly, the current research addresses the e-consumer shopping pattern, which constitutes the products bought and amounts spent over a relatively long period. Tracking consumer shopping for an extended period is challenging. Secondly, log files and clickstream data cannot provide a complete picture of shopping pattern. The consumer shops online using different peripherals, platforms, and networks. It is conceivable for a consumer to shop from home, work, or any lieu with Internet connection. In addition, the consumer at times delegates the shopping task to an acquaintance or relative (e.g., because the relative has a credit card, is more Internet-savvy, has a loyalty program membership, located in another town where a certain offer is available). The use of log files and clickstream data also assumes that one consumer uses the IP address. Several individuals such as other tenants, neighbors, and guests might use the same IP address when purchasing. Besides, the use of log files and clickstream data requires informing the participants beforehand that their activities will be monitored, resulting in measurement bias (consumers with privacy and security concerns will participate less, consumers will change their behavior, etc.). The Hawthorne effect, or observer effect, is well documented. Objective measures such as log files are not reliable in reflecting the e-shopping pattern because of the noise accompanying data. The consumer might use the shopping cart but check out using a different IP address, which results in labelling the activity as non-purchase. The consumer also might return the item offline or purchase the item with the intent of reselling it. All these factors favored the subjective measure (self-reported) for the study. The collection of the research survey was approved by the Research Ethics Board of HEC Montreal (https:// www.hec.ca/en/research/research-ethics-board/research-ethics-board.html no. 943-423-013).

Sample
Participants were recruited using the consumer panel of Leger, a market research company headquartered in Canada. The panel, characterizing the sampling frame, comprised about 170,000 members. Panel members, who already provided personal and contact information, had a chance to win monthly monetary prizes-their chance to win increases when participating in the study. Using an online panel warrants that the participants are conversant Internet users.
The email list of the randomly selected entries was used to contact the members and invite them to participate. The final sample was composed of 700 adult consumers that adequately responded to the questionnaire. The sample distribution was satisfactory. Age was distributed as follows: 35% 34 y/old or younger, 43% 35 to 54 y/old, 22% 55 y/old or older. Educational level distribution was: 2.9% primary education, 27.9% secondary school, 12.9% professional training, 42.7% undergraduate diploma, and 13.7% graduate studies. Income (US dollars) was distributed as follows: 21.3% earned less than 35 K, 17.1% earned between 35 and 50 K, 24.1% earned between 51 and 80 K, 22.1% earned between 81 and 110 K, and 15.3% earned 111 K or more. Females comprised 43% of sample.
Because the lieu of study (Canada) comprises English-and French-speakers, the participant was offered the option to respond to the English or French version of questionnaire (the French questionnaire was professionally translated from the English one and contained the same items). Almost half participants chose to respond to the French version of the questionnaire.

Measure
To reflect the pattern of online product shopping, participants were instructed to specify the amount spent on online purchases during the last three months. They were also instructed to specify how much of this amount was spent on each product category. The categories were: (1) electronics (electronics, computers, and their accessories), (2) tickets of musical and cultural events, (3) items of entertainment IOE (games, toys, music and video including DVDs), (4) books (books, magazines, and journals), (5) travel and tourism, (6) fashion (clothes, jewelry, and accessories), and (7) household (furniture and related items). Thus, our segmentation base is an observable product-specific base (Wedel & Kamakura, 2000). Susceptibility to interpersonal influence (seven points: totally disagree/totally agree) was measured using Bearden, Netemeyer, and Teel's 12-item scale (1989) after adapting it to a retail context. These authors developed the construct of susceptibility to interpersonal influence as a trait that reflects a person's need to identify with or enhance own image in the eyes of others in a consumption context. That is, a consumer high on this trait would seek to enhance own image through the acquisition of certain products and brands that conform to others' expectations. Exploratory factor analysis showed that the scale has two dimensions, like Bearden et al. (1989) findings. The first dimension, termed normative interpersonal influence, comprised the following items (It is important that others like the retailers I shop from; I generally shop from retailers that I think others will approve of; If other people can see me shopping, I often shop from retailers they expect me; I rarely shop from the latest fashionable retailers until I am sure my friends approve of them; I like to know what retailers make good impressions on others; I achieve a sense of belonging by shopping from the same retailers as others; If I want to be like someone, I often try to shop from the same retailers as them; I often identify with other people by shopping from the same retailers as them; To make sure I shop from the right retailer, I often observe what retailers others shop from; α =95). The second dimension, termed informational interpersonal influence, comprised the following items (If I have little experience with a retailer, I often ask my friends about this retailer; I often consult other people to help choose the best retailer for a given product class; I frequently gather information from friends or family about a retailer before I shop; α =90).
English language skill (three points: Beginner/Intermediate/Expert; α =96) was measured using three items (My skill level in speaking English is . . .; my skill level in reading English is . . .; my skill level in writing English is). Familiarity with US online shopping (seven points: totally disagree/totally agree) was measured using two items (I am familiar with buying on US websites; I am familiar with searching for products on US websites; Pearson correlation coefficient r =85).
Participants were also instructed to specify the amount of money spent on each of the following e-tailer type: (a) Websites with retail outlets (e.g., stores) or headquarter in the respondent's province, (b) Websites with retail outlets (e.g., stores) or headquarter in another Canadian province, (c) US websites without retail outlets (e.g., stores) and no headquarter in Canada, and (d) Websites that are non-Canadian and non-US. On average, the amount spent on each e-tailer type was as follows: (a) $306.14; SD = 654.97, (b) $162.56; SD = 562.99, (c) $207.60; SD = 604.65, and (d) $74.68; SD = 253.34. The questionnaire concluded with items measuring the demographic factors.

Analysis and results
A consumer e-shopping pattern (the amounts spent on each of the seven product categories) constituted the segmentation base. Table 1 reports the correlations between these amounts. The table shows that the amounts are not highly correlated, indicating that e-shopping pattern is idiosyncratic. Most of the correlations are insignificant. In addition, the significant correlations are weak (0.223 was the highest r). The books category significantly correlated with almost all categories. This result implies that buying books, magazines, and journals online is a shopping behavior widely shared. It further implies that purchasing this product category is key to perform additional purchasing of other categories. Spending on fashion correlates with spending on tourism (p =001). Spending on tourism correlates with spending on electronics (p =002). Whereas the correlation table does not allow to address the research questions (consumers heterogeneity in their online purchasing amounts and in their preference to conduct experience vs material purchases), this is achieved in the analysis reported next.
To explore the segments, a nonoverlapping descriptive method was used (Wedel & Kamakura, 2000). That is, based on e-shopping pattern, each consumer was assigned to a single homogeneous group. We performed a two-step cluster analysis, available in SPSS, to segment the observations. This analysis, adequate for larger samples, resulted in favoring three segments (Table 2 for AIC). The three segments provided a parsimonious solution with indicative and significant mean differences for the segmentation base variable (Table 3 for centroids of total purchasing amount and the amount spent on each of the seven product categories).
Then, we performed discriminatory analysis to test the variables that diverged across the three segments. Chi-Square and ANOVA tests show, as expected, symptomatic differences in key factors across the segments. The results (summarized in Table 4) shows that the segments differ according to the total amount spent online, income, and educational level. In addition, susceptibility to interpersonal influence (the informational dimension) and familiarity with US online shopping vary. English language skill marginally varies (p = .074). Further, the amounts spent according to e-tailer type differ across the segments.
The discriminatory analysis revealed anticipated differences between the segments (Table 4). First, the total amount spent on e-purchases differs (F = 280.16, p = .000): VIB of Material ($ 2707.83) and VIB of Experience ($2084.76) spent remarkably higher amounts compared to Normals ($328.00). While the VIB do not form a consumer majority (they form 22% of consumers), they generate a significant portion of sales (about two-thirds). The results show that affluent consumers are heterogeneous in their shopping preference-to conduct experience versus material purchases.
The segments differ according to psychological factors. For consumer susceptibility to interpersonal influence, one dimension (normative interpersonal influence) did not differ across the segments (F = 0.467, p = .627). However, the second dimension (informational interpersonal influence) differs (F = 4.363, p = .013): VIB of Experience scored the highest (3.82) while VIB of Material scored the lowest (2.98) on this dimension (Table 4). Because social influence differently affects consumer groups (Bearden et al., 1989), susceptibility to interpersonal influence discriminates among the segments (VIB of Experience scored highly on this trait). Experience consumption entail greater levels of personal interactions (Fitzsimmons & Fitzsimmons, 2004). This finding suggests that VIB of Experience and VIB of Material have divergent shopping preference because of psychological underpinnings. It also suggests that VIB of Experience allocate a higher value to the social aspect.
Because consumers get involved in purchasing from a retailer group due to habit (Lwin et al., 2016), familiarity is expected to discriminate among the segments (VIB should score higher than English language skill marginally differed across the segments (χ 2 = 6.79, p = .074): VIB of Experience comprised more consumers with high skill (72%) compared to Normals (60.8%) and VIB of Material (63.4%). Because many retailers have e-stores with English interface, and fluently speaking the retailer language facilitates shopping, English language skill should discriminate among the segments (VIB should score higher than other consumers do). This notion received marginal support.
Theory predicts that consumer e-purchase pattern is affected by e-tailer type (Frasquet et al., 2015). The e-tailer type should discriminate among the archetypes. The amount spent according to e-tailer type differs across the archetypes in a manner like that of total amount spent online (i.e., for each e-tailer type, VIB of Material spent more than VIB of Experience, and both spent more than the Normals; Table 4). Two differences are noteworthy. The Normals were more likely to shop from e-tailers located within their area of residence (47% of total amount). Secondly, VIB of Material were the most likely to shop from e-tailers located overseas (15% of total purchases).
Consumer demographic profile deviates across the segments. The deviation was significant for income (χ 2 = 34.40, p = .000) and educational level (χ 2 = 40.63, p = .000). Compared to VIB of Experience, VIB of Material had lower educational levels (49.6% vs 26.8% hold undergraduate degree, respectively). Moreover, compared to other segments, VIB of Experience were the most likely to be university graduates (76.1% of them).
In terms of income, the VIB belonged to higher income brackets compared to Normals. A notable difference between the VIBs is that more VIB of Material (39.0%) belonged to the 81-110 K income bracket, while more VIB of Experience earned over 110 K per year. This income gap suggests that VIB of Experience enjoy the utmost purchasing power, even when compared to VIB of Material. Age marginally differed across the segments (χ 2 = 9.80, p = .065), with more youth belonging to Normals. Gender did not differ (χ 2 = 2.20, p = .33).

Discussion
Researching the adoption of Pareto Principle by marketing practitioners, Westerby and Nortun (2021) emphasized that empirical attempts to pinpoint the affluent shoppers are needed for effective targeting. They indicated that such targeting is beneficial because it helps increase sales for all segments-the affluent shoppers and the typical shoppers. Our results respond to this call and support the notion that two segments of consumers generate most sales and

Notes-Numbers in cells are the amounts spent on a product category in US dollars (standard deviation in brackets)
-A value discriminating a segment is in italics -IOE refers to items of entertainment (games, toys, music and video including DVDs) Notes-Only the factors whose distribution significantly differs across the archetypes are reported in the table -For e-tailer type, percentage of total amount spent is reported in brackets dominate the online market: those who shop for experience and those who shop for material goods. The third archetype, the Normals, do not spend large amounts online.
In effect, the analysis revealed three homogeneous segments: The Normals (consumers that buy various products online yet spend small amounts), VIB of Experience (consumers that spend large amounts shopping for experience, i.e., travel and tourism as well as tickets of musical and cultural events), and VIB of Material (consumers that spend large amounts shopping for material products such as electronics, household equipment, clothes, jewelry and accessories). A post-doc discriminatory analysis showed that the segments have divergent demographic and psychological profiles and suggested that lifestyle is an underpinning aspect.
Scholars noted that consumers are heterogeneous according to their shopping frequency, which determines their shopping behavior. Arce-Urriza et al. (2017) showed that shopping frequency determines consumer preference for marketing mix, with frequent shoppers affected more by promotions. Our results reveal three segments of online shoppers. The Normals constituted the largest segment of consumers (78.0%) who shop for various products online yet spend remarkably low amounts (average of e-purchases was $ 328) compared to VIB. While the Normals form the biggest purchasing group, they do not generate many sales (they generate about one-third of sales). Alternatively, VIB of Experience constituted the second biggest segment of consumers (16.1%) with average e-purchases of $ 2085. They disproportionately shop for travel and tourism products as well as for tickets of musical and cultural events. They hence have a preference to shop for experience more than other consumers do. Finally, VIB of Material comprised 5.9% of consumers with average e-purchases of $ 2708. Members of this segment spend the highest amount online, disproportionately shopping for material goods such as electronics, household equipment, clothes, jewelry and accessories. Dahana et al. (2019) suggest that lifestyle helps explain the heterogeneity between affluent shoppers because lifestyle associates with customer lifetime value. These researchers indicate that a research gap exists because no research used shopping data such as purchase amount to link it with observed lifestyle. Our findings help mitigate this gap and show that some e-shoppers have a disposition to conduct experience versus material purchasing. Our findings further suggest that such disposition is driven by lifestyle.
The discriminatory analysis results suggest that the underpinning factor differentiating between the VIBs is lifestyle, governed by social class factors such as income and educational level (Sobel, 2013). No difference was found between the VIBs according to age. The differences were however notable in income, educational levels, and English language fluency. According to social class segments (Armstrong & Kotler, 2016), the profile of VIB of Experience corresponds to the Upper Uppers (a segment within the upper class) and the Upper Middles (a segment within the middle class), whereas the profile of VIB of Material corresponds to the Lower Uppers (a segment within the upper class) and the generic Middle Class.

Theoretical implications
This research has contributions to theory. Firstly, it offers an empirical examination for the Pareto Principle, showing its relevance to distinguish and target the affluent e-shoppers. Secondly, it contributes to the stream contemplating consumer segments (Barnes et al., 2007;Jayawardhena et al., 2007) by highlighting the existence of homogenous and eloquent archetypes of e-shoppers. It extends the theory proposed in Aljukhadar and Senecal (2011) and suggests that consumers can be reliably segmented according to their Internet use and purchase pattern. Thirdly, it elaborates the theory on experience versus material consumption (Van Boven & Gilovich, 2003). Kim (2018) showed that people value experience consumption due to an innate trait-the desire for exclusivity. This research extends this theory by showing that consumers are inclined to conduct experience purchase due to innate and lifestyle factors. Experience consumption delineates the heterogeneity between influential buyers. Using purchase type (experience vs material) to segment the affluent shoppers is warranted. This finding extends the notion originally suggested by Wedel and Kamakura (2000), who evaluated several segmentation bases and concluded that usage pattern satisfactorily meets the criteria required to obtain an effective segmentation. Our results show that shopping pattern offers a meaningful segmentation base.
The results challenge the notion that consumers in general prefer to buy tangibles (over services) online because uncertainty is low-a notion initially proposed by Liang and Huang (1998). Affluent shoppers comprised two groups: those inclined to purchase experience (VIB of Experience) and those inclined to purchase tangibles (VIB of Material). Shopping pattern appears to dissociate from uncertainty level and to associate with a disposition driven by lifestyle factors. The results offer insight to the experience recommendation (which advises consumers to buy experience rather than material goods to become happier). Given that a considerable segment of affluent shoppers (VIB of material) tends to purchase material due to lifestyle-driven disposition, advising those shoppers to purchase experience might backfire. The lifestyle-driven disposition to consume material helps fathom the results that challenge the experience recommendation, e.g., experience purchases led to discontent (Nicolao et al., 2009).

Practical implications
Significant practical implications emerge from the findings. Firstly, they invite practitioners to recognize consumers' heterogeneity in conducting online purchases. Practitioners should view the online market as comprised of homogeneous submarkets. The results show the online market is comprised of three submarkets including shoppers who diverge not only in the type of product sought but also in the amounts spent. The segments show idiosyncratic criteria that foretell lifestyle. Westerby and Nortun (2021) found that marketing executives do not target the small core, or affluent shoppers, because they believe that the other shoppers, termed the trivial many, are sufficient to sustain their businesses. Our results encourage executives to reconsider such strategy. Executives should acknowledge that a significant portion of online sales originates from a small segment (22% of consumers, i.e., VIB of Experience and VIB of Material) and target these shoppers accordingly. Firms can integrate the VIB notion in their CRM and social media strategies. They can pursue an effective targeting strategy by segmenting shoppers according to their shopping disposition and purchasing power. The analysis (Table 4) offers a starting point. Executives can also use social class to identify their target market. The results suggest that VIB of Experience belong primarily to the Upper Uppers (a segment within the upper class) and the Upper Middles (a segment within the middle class). On the other hand, VIB of Material belong to the Lower Uppers (a segment within the upper class) and generic Middle Class. Executives can predict a consumer latent segment based on purchase history, sociodemographic, and lifestyle factors-which can be gauged by consumer social media profile and posts (Hu et al., 2017).
To improve convergence, firms should adapt their integrated marketing communications according to the consumer segment. Ads that fit context are more effective. Firms can follow a resource-based approach by targeting their apt online segment for an optimal allocation of marketing expenditure. They can focus on the segment for which they can maximize value. For instance, a vendor issuing tickets of cultural and musical events should principally target VIB of Experience, who are fluent in English, university graduates, belong to the high-and utmostincome brackets, familiar with US e-tailers, and highly susceptible to interpersonal influence. Alternatively, a vendor of electronics, a dealer of home decoration items, or a retailer of fashion items should primarily target VIB of Material, who belong to a high-income bracket, received a secondary education or professional training, and are more willing to shop from foreign, remote vendors.
The store interface can be configured to suit the need of the target segment. Upon interacting with a VIB of Material, the store's recommendation agent should prioritize a variety of tangibles. The recommendation agent however should give priority to experience goods upon interacting with a VIB of Experience. The recommendation agent should display discounted, end-of-season products and services using local language upon observing a shopper belonging to Normals. The store interface can thus be optimized-or morphed-according to shopper segment.

Limitations and future research
The results show two distinct types of influential buyers and highlight the relevance of Pareto Principle based on the input of adults that subscribed to a consumer panel. The research has limitations. Data come from a self-reported survey; the results hence reflected purchase behavior bestowing on participant memory and collaboration. Future work should validate the results using objective measures of purchase behavior. Each of the revealed segments might comprise sub-segments (Allred et al., 2006); nonetheless, the sample size did not allow investigating the sub-segments. The sample comprised participants belonging to a large consumer panel in a developed Western country (Canada). Work should inspect the findings in developing countries (emerging and subsistence economies such as Brazil, India, and African countries) to show the role of economic development, and in Eastern countries (e.g., Japan, Malaysia, Arabian Gulf) to show the role of culture. Future work should also elaborate the findings. Besides lifestyle, the factors that discriminate VIB of Experience and VIB of material should be examined. In addition, the factors and events that lead consumers to switch segments (move from VIB of Experience to VIB of material and vice versa) should be highlighted. The results suggest that a remarkable change to consumer lifestyle due to social mobility or analogous reason would result in such a switch. These topics delineate future research prospects.