Introduction

Brand management is regarded as a priority across all types of organizations (Keller and Lehmann 2006) and brand engagement has been recognized as a significant contributor to product differentiation and profitability (Brodie et al 2013; Reichheld et al 2000; Kumar and Pansari 2016). There has been escalating interest in customer–brand engagement both within the industry and in academic circles (Harmeling et al 2017; Pansari and Kumar 2017), though the structure of engagement (Schmitt 2012), as well as the drivers of engagement, has been elucidated (Żyminkowska 2018). As curtailing customer defection has positive implications for company profits (Reichheld 1996), the importance of retaining customers for a long lifespan has been emphasized (Dowling 2002; Jain and Singh 2002).

However, many prior studies have treated customer–brand engagement as a state at a particular instant, with insufficient attention being paid to the continuity implicit in long-term customer engagement. This study defines long-term customer engagement as the continued embodiment of several critical characteristics of customer–brand engagement over numerous years. Key features of customer–brand engagement extend beyond mere purchase and consumption (Dessart et al 2016), and it should generate attitudinal loyalty (Bowden 2009; Dwivedi 2015; France et al 2016; Leckie et al 2016; Solem 2016; Thakur 2016). Maintaining customer engagement is crucial for all brands and products, regardless of their inherent attributes.

Whereas it has been commonly observed that utilitarian products typically generate less engagement compared to hedonic ones. Numerous academic investigations have conducted thorough the analyses of customer motivations, suggesting that the dichotomy between utilitarian and hedonic brands fosters distinct consumer behavior patterns (Li et al 2020; Schau et al 2009). Utilitarian brands, often chosen out of habit, are more inclined to lead to consumer disengagement (Fournier 1998; Hess and Story 2005; Keller 2001; Keller and Lehmann 2006; Verhoef and Langerak 2002). Previous research has utilized the hedonic utilitarian characteristics of brands as a foundation for understanding overarching consumer engagement. In a similar vein, Żyminkowska (2018) has employed the hedonic and utilitarian aspects of customer value as a basis for identifying the driver of customer engagement. However, Żyminkowska (2018) has primarily depicted engagement in a static context, thus inadequately capturing its dynamic aspects. Consequently, the methods through which engagement can be maintained are not comprehensively understood.

This study aims to delineate the drivers for maintaining brand engagement in utilitarian versus luxury brands and provide a comparative analysis with the existing literature on the genesis of brand engagement. To achieve this objective, the study harnesses the case study methodology as outlined by Yin (2003), opting for a representative sample of consumers who consistently engage with both brand categories within the Japanese context, to identify the exploratory drivers behind such enduring engagement.

Japan exhibits the world’s highest proportion of an aged population at 28.9%, alongside an exceptionally low fertility rate of 1.3% in 2020 (OECD 2021). Demographic transitions such as aging and reduced birth rates directly contribute to urban shrinkage, a phenomenon observable in over 60% of Japanese cities (Wang et al 2020). Foot and Gomez (2014) postulated that industries such as clothing and footwear are poised to face significant challenges due to an aging population in future. In several sectors, the customer base may contract as a result of population decline and aging. Under such circumstances, the significance of customer retention is anticipated to supersede the need for new customer acquisition. The urban shrinkage, as indicated by declining population, has also been reported in numerous European cities (Rink et al 2014). Therefore, the implications derived from this study may prove beneficial in similar demographic contexts across various nations.

We adopted semi-structured, critical incident interviews to decipher how and why customer engagement was preserved. To fully comprehend the drivers, it is crucial to grasp the context of the customer’s narratives. We moved beyond simply focusing on frequently used words and applied a co-occurrence network analysis (Higuchi 2016) to examine the network of words. Utilizing the measure of betweenness centrality, we facilitated a contextual understanding based on keywords. Words with high betweenness centrality are crucial, as the loss of connections to these words would decontextualize their meaning (Freeman 1978; Mehra et al 2001).

This study contributes by identifying the drivers for maintaining engagement. Unlike previous studies that primarily focused on the antecedents of engagement for utilitarian and hedonic brands, our study identified the presence of utilitarian elements within hedonic brands and hedonic elements within utilitarian brands. This finding offers practical implications, suggesting that providing incentives opposite to the brand’s usual positioning could be effective for customers who have maintained engagement over a period.

Literature review

The engagement concept is the dynamism that characterizes consumer–brand interactions (Brodie et al 2013) and engagement is one of the most critical issues in marketing (France et al 2016; Nyadzayo et al 2020). Engagement is a loyalty-building process (Bowden 2009) based on the customer’s cognitive, emotional, and behavioral investment in a brand interaction (Brodie et al 2011, 2013; Hollebeek 2011a, 2011b, 2013).

Customer attitudinal loyalty is built through customer–brand engagement (Dwivedi 2015; Leckie et al 2016; Solem 2016; Thakur 2016), thus customer engagement predicts customer loyalty (Hollebeek et al 2016; Ramaswamy 2009; Sawhney et al 2005). In services marketing research, service quality and brand experience have been identified as predictors of customer engagement (Prentice 2014; Prentice and Loureiro 2018; Prentice et al 2019).

Engagement and loyalty reciprocally influence each other, making differentiation challenging. However, high engagement can be considered an embodiment of what Dick and Basu (1994) term “true loyalty,” encompassing both the attitudinal and cognitive dimensions of loyalty. Furthermore, the relationship between brand engagement and brand loyalty has also been shown to be mediated by attachment to the brand (Kumar and Nayak 2019). We understand that the engagement structure and its levels have multifaceted elements (Schmitt 2012). Moreover, we know the antecedents of engagement (Prentice 2014; Prentice and Loureiro 2018; Prentice et al 2019) and its outcomes (Dwivedi 2015; Leckie et al 2016; Solem 2016; Thakur 2016).

However, engagement is neither static, nor an eternal condition. Therefore, recognizing what factors maintain consumer engagement is essential in brand management and companies need to involve consumers in the experience (Siebert et al 2020). While attempting to maintain a state of engagement, the intention of generating and maintaining engagement will likely vary among companies, just as consumer’s preferences and continued use of a service will alter (Montazemi and Qahri-Saremi 2015). Customers enjoy both instrumental (utilitarian) and experiential (hedonic) benefits from consumption (Batra and Ahtola 1991).

Previous studies have examined different aspects of the consumer–brand relationship by distinguishing brands based on variations in consumer attitudes. Two key dimensions commonly used to differentiate brands are brand features (utilitarian vs. hedonistic) and market positioning (luxury vs. mass market) (Chaudhuri and Holbrook 2002; Chen et al 2017; Kronrod and Danziger 2013; Vock 2022).

Hedonic brands are characterized by playfulness, enjoyment, and excitement (Arnold and Reynolds 2003; Babin et al 1994; Hirschman and Holbrook 1982; Nenkov and Scott 2014). Conversely, utilitarian consumption focuses on efficient and effective completion of specific tasks (Chen et al 2017; Childers et al 2001; Mathwick et al 2001; Mehta et al 2004; Strahilevitz and Myers 1998; Voss et al 2003). Utilitarian brands are often chosen out of inertia and typically lack a lasting consumer–brand connection compared to hedonic brands (Fournier 1998; Hess and Story 2005; Keller 2001; Keller and Lehmann 2006; Verhoef and Langerak 2002).

Although both hedonic and utilitarian values positively influence brand loyalty, the impact of hedonic values is more significant (Lee et al 2021). Shimul et al (2019) conducted an in-depth investigation of luxury brand attachment measures, highlighting emotional connection as a crucial measure. However, the process of fostering this emotional connection has not been thoroughly explored.

Within the limited body of research on the drivers of engagement, Żyminkowska (2018) quantitatively examined customer engagement arising from communication, complaints, and collaboration, emphasizing the multifaceted impact of both hedonistic and utilitarian values. Moreover, Hepola et al (2020) argued that brand love can develop for both utilitarian and hedonistic brands. However, they noted key differences in the drivers of service continuity between the two, whereby attitudes and satisfaction hold significance for utilitarian brands, while engagement plays a crucial role for hedonistic brands.

Existing research has demonstrated the significance of customer value, participation, and emotional connection in fostering engagement, irrespective of whether the brand is hedonic or utilitarian. However, there remains a lack of comprehensive research on the dynamic nature of engagement and its maintaining factors. Addressing this research gap, this study investigated the following research questions (RQs):

  • RQ 1: Why and how is customer engagement maintained in utilitarian and hedonistic brands?

  • RQ 2: What are the differences in maintaining customer engagement between utilitarian and hedonistic brands?

Method

This study chose an exploratory research approach (Corbin and Strauss 2014) to identify the drivers of long-term customer–brand engagement. We examined cases of five brands (Eisenhardt 1989), under the case study approach; consequently, this study did not have the “need for control over the behavioral project” as seen in the experimental research, but is required to “focus on the present event” and asked why and how (Yin 2003, 5–7). To avoid biases such as imperfect correlations between intentions and behavior (Bagozzi and Dholakia 1999), the study employed a representative sample of customers who have engaged with a specific brand for many years in the actual situation.

Sample and procedure

The initial starting point was an introduction to a well-known luxury brand with a global presence. The brand managers obtained the customer’s approval before participating and the interviews began with the consent of both the brand managers and customers.

After obtaining two loyal customer referrals for the first brand (the brand name was kept confidential), snowball sampling led to three long-term, engaged customers for different luxury brands (Tiffany, Van Cleef & Arpels and CÉLINE). With the customer’s cooperation, five interviews were conducted.

Next, in the category of utilitarian brands, the researchers studied the case of the Japanese retailer MUJIRUSHI (MUJI). MUJI is a specialty retailer that primarily sells private-label apparel, household goods, and food products. As of August 31, 2021, MUJI operates 497 stores in Japan and 571 in 32 foreign countries. Despite MUJI offering practical products, it has enthusiastic fans called MUJIRER (Mujirushi + er); some of whom have published books about their experiences with the brand and have even appeared on TV programs.

The researchers conducted five in-depth interviews with enthusiastic users. The respondents were women in their 30 s and 40 s. The interviews were conducted between 2016 and 2021. The duration of the interviews ranged from 60 to 120 min (average was 90 min) for the hedonistic brands and from 30 to 60 min (average was 47 min) for the utilitarian brands. The interviews were conducted in non-store locations, such as coffee shops and hotel lounges. In each survey, the process from initial purchase to strong engagement was explored through semi-structured interviews below.

We asked four questions for the semi-structured interview, which are as follows:

  1. 1.

    Why did they initially become interested in that brand: the genesis of interest may be underpinned by either hedonic or utilitarian elements (Chen et al 2017; Vock 2022).

  2. 2.

    What motivated their repeated purchases: the importance lies in attitudinal and cognitive loyalty (Dick and Basu 1994).

  3. 3.

    Why and how did that brand become special to you: engagement is a loyalty-building process (Bowden 2009).

  4. 4.

    What significance does the brand, with which a long-term relationship has been built, hold for you: engagement is mediated by attachment to the brand (Kumar and Nayak 2019).

Data analysis

To eliminate subjective bias and identify key contexts out of the data gathered from interviews, we conducted a co-occurrence network analysis, focusing on betweenness centrality, using an open-source language software KH-Coder (Higuchi 2016). Co-occurrence network analysis identifies not only words that frequently appear in the data but also those that frequently co-occur within the same sentence. In the context of co-occurrence network analysis, the co-occurrence coefficient measures the degree of simultaneous occurrence of two elements, say A and B, within an entity, such as a sentence. The co-occurrence coefficient can be computed using the following equation:

$$ {\text{A}} \cap {\text{B}}/{\text{A}} \cup {\text{B}} = {\text{A}} \cap {\text{B}}/\left( {{\text{A}} + {\text{B}} - {\text{A}} \cap {\text{B}}} \right) $$

In this formula, “A ∩ B” denotes the intersection of element A and B, that is, instances where A and B occur together within an entity. On the other hand, “A ∪ B” represents the union of element A and B, encompassing all instances where either A, B or both are present.

Thus, the left-hand side of the equation quantifies the ratio of co-occurrence of A and B to all instances of either A, B or both. The right-hand side, by summing the total instances of A and B and subtracting the co-occurrences (to avoid double counting), effectively gives the same measure. This equation hence provides the mathematical representation of the co-occurrence coefficient in network analysis, essential for examining linguistic patterns or relationships between words in text analytics.

Co-occurrence relations in KH-coder (Higuchi 2016) can be observed via a network diagram, where extracted words that appear in the same paragraph are identified as nodes and the lines connecting these nodes as links (Figs. 1 and 2). Words connected by a line are in a co-occurrence relationship, used within the same context. The importance of this diagram lies not in the placement of the extracted words, but in the fact that these words are connected by lines.

Fig. 1
figure 1

Result of co-occurrence network analysis: high betweenness centrality words and key contexts of long-term engaged customers in hedonic brands. Source: Authors

Full size image
Fig. 2
figure 2

Result of co-occurrence network analysis: high betweenness centrality words and key contexts of long-term engaged customers in utilitarian brand. Source: Authors

Full size image

Understanding the core context requires pinpointing words exerting significant centripetal force among the extracted vocabulary. To address this task, we deployed an analytical approach rooted in betweenness centrality (Freeman 1978). High betweenness centrality represents a word’s pivotal role in structuring the network’s context. Nodes exhibiting high betweenness centrality can bridge the so-called “structural holes,” which represent areas devoid of connections; thus, reaping intermediary benefits from value gaps spawned by missing structural links (Burt 1992; Obstfeld 2005). These nodes demonstrate a position of superiority (Zaheer and Bell 2005). Accordingly, identifying words with high betweenness centrality signifies the extent to which a word mediates others and therefore bears considerable importance (Mehra et al 2001).

In the context of network analysis, betweenness centrality quantifies a node’s or word’s influence by assessing the number of shortest paths among all other nodes that transit through the node in question. This is typically represented as:

$$ {\text{CB}}\left( {\text{v}} \right) = \sum {\left[ {{\text{s}} \ne {\text{v}} \ne {\text{t}} \in {\text{V}}} \right]\left( {\sigma {\text{st}}\left( {\text{v}} \right)/\sigma {\text{st}}} \right)} $$

Herein:

  • CB(v) designates the betweenness centrality of node v,

  • V refers to the set of all nodes in the network,

  • σst tabulates all shortest paths extending from node s to node t, and

  • σst(v) represents the paths passing through node v.

The computation of a node’s (v) betweenness centrality entails summing fractions of all shortest paths spanning each pair of vertices that traverse node v. This process captures a node’s role as a ‘bridge’ within the network, allocating higher values to nodes that link distinct clusters. Words with high betweenness centrality are illustrated through darker shading in KH-coder (Higuchi 2016) as depicted through Figs. 1 and 2.

All interviews were recorded and transcribed verbatim with the consent of the interviewees and were subsequently analyzed.

Results

As a result of our analysis, we identified the drivers that emerge in the process of maintaining engagement, comparing these based on the elements of hedonic and utilitarian brands. By identifying words with high betweenness centrality, we recognized important points common among the narratives of multiple customers.

In the analysis, we applied the same conditions for both brands: minimum word frequency was set at five or above and only those with the strength of co-occurrence relationships exceeding 0.2 were extracted for comparison. Furthermore, based on the keywords in the context, we referred back to the original data, conceptualized common contexts among several customers (Corbin and Strauss 2014), and identified whether these motivations were rooted in the characteristics of hedonism (Arnold and Reynolds 2003; Babin et al 1994; Hirschman and Holbrook 1982; Nenkov and Scott 2014) or in the characteristics of utilitarianism (Chen et al 2017; Childers et al 2001; Mathwick et al 2001; Mehta et al 2004; Strahilevitz and Myers 1998; Voss et al 2003) (see Table 1).

Table 1 Comparison of the customers’ key contexts in maintaining their engagement with both brands
Full size table

Key contexts concerning long-term customer engagement in hedonic brands

After analysis, from a total of 21,960 words, 4962 words were extracted. The circle size symbolizes the word frequency and words co-occurring in the same context are portrayed in the same sub-graph. Nodes (words) with high betweenness centrality are highlighted with a darker color (Fig. 1).

The analysis yielded a total of 14 sub-graphs. Among these, one central sub-graph containing words with high betweenness centrality was further decomposed into four parts (A–D) for context exploration. We also conceptualized the key contexts for the remaining ten sub-graphs (E–N) that did not contain words with high betweenness centrality (Fig. 1 and Table 1). Three sub-graphs: “Celine and shoe,” “many and customer,” and “year and V (brand name)” were deemed as merely descriptive of the phenomena, and thus, were not included in the targets.

Key contexts of long-term customer engagement in utilitarian brands

After analysis, out of 8123 words, 13,013 were extracted. The size of each circle indicates the frequency of each word and words that co-occur within the same context are depicted in the same sub-graph. Nodes (words) with high betweenness centrality are denoted by a darker color (Fig. 2).

Upon analysis, a total of 13 sub-graphs were identified. A primary sub-graph encompassing words with high betweenness centrality was further subdivided into four categories (A–D) to explore their contexts. Additionally, the key contexts for each of the 11 sub-graphs (E–O) that did not contain words with high betweenness centrality were also conceptualized. The “See and look” sub-graph was deemed as merely descriptive of the phenomena and thus was not included in the targets.

Drivers of maintaining customer–brand engagement

The motivations at the heart of each context, either hedonic or utilitarian, were tabulated to provide a comparative perspective. Upon simple comparison of the hedonic and utilitarian elements, it was found that for luxury brands, the contexts that relied on hedonic elements constituted 57% (8 out of 14). For MUJI, the contexts relying on utilitarian elements accounted for 33% (5 out of 15). The continuation of engagement did not necessarily align with the main characteristics of the product. Specifically, in sub-graphs with key words featuring high betweenness centrality, the opposite elements were markedly prominent. For luxury brands, the utilitarian elements were dominant, accounting for 75% (3 out of 4), while for MUJI, the hedonic elements were entirely dominant, accounting for 100% (4 out of 4).

Discussion and conclusion

Theoretical contributions

Traditional research categorizes customer–brand engagement into two primary groups: hedonism and utilitarianism, each corresponding to varying consumer desires (Chen et al 2017; Kronrod and Danziger 2013). Furthermore, it has been posited that maintaining engagement with utilitarian brands presents more challenges compared to hedonistic brands (Fournier 1998; Hess and Story 2005; Keller 2001; Keller and Lehmann 2006; Verhoef and Langerak 2002).

This study seeks to contribute fresh perspectives to the literature by offering a dynamic viewpoint on engagement, identifying factors that maintain engagement with brands of differing characteristics. Findings from this research indicate that drivers maintaining engagement are not exclusively linked to the brand’s nature. Engagement with hedonistic brands tend to sustain through an appreciation of the additional value provided by the brand. This supports the previously proposed perspective that hedonistic values hold substantial importance. Conversely, in situations where engagement persists, the utilitarian aspect of acquiring the product was identified as a significant factor.

Within utilitarian brands, hedonistic elements that contributed to personal growth and goal attainment through the product or brand were valued, thus deepening engagement. These findings suggest that factors exceeding the defined characteristics of the product play a crucial role in maintaining engagement.

For hedonistic brands, the brand name’s betweenness centrality was found to be low, and centripetal force could not be confirmed. In contrast, utilitarian brands demonstrated centripetal force within their brand names. These results augment the observations of Hepola et al (2020), which posited that ‘brand love’ can occur within both hedonistic and utilitarian brands. Furthermore, this study not only confirms the potential persistence of such ‘brand love’ but also elucidates the factors that facilitate its sustainability.

Managerial implications

The practical implications derived from this study lie in demonstrating the factors responsible for deepening engagement even for utilitarian brands. The process of enhancing and maintaining engagement does not solely rely on means that means involving some gain in the form of discounts. It is desirable to create opportunities associated with needs related to self-affirmation, such as achieving goals, in the customer’s product usage process. Interviewees utilized social media to showcase how effectively they were utilizing the product. Utilitarian brands played an indispensable role in elevating its customers onto the stage, even if it was as supporting actors, as evidenced by the high betweenness centrality of the brand name.

In hedonistic brands, the customers we surveyed were top-tier customers who resumed engagement after a period of inactivity spanning over a decade. This highlights a concern that contemporary businesses rely too heavily on recent transactional data. Parameters such as purchase frequency, recency, financial contribution (Dursun and Caber 2016; Miglautsch 2000), variety (Smaili and Hachimi 2023), social context (Taşabat et al 2023), and future customer potential (Kanchanapoom and Chongwatpol 2023) provide valid information for analysis. The dynamic nature of engagement and the decline in potential customer bases in certain countries further accentuates the importance of retaining customer data even during the periods of dormancy.

Furthermore, all interview participants initially expressed hesitancy in entering luxury brand stores. This suggests that digital communication could alleviate the psychological barriers associated with visiting brick-and-mortar stores (Hollebeek and Macky 2019; Baker et al 2018; Martín-Consuegra et al 2019). However, at the stage of maintaining engagement, the establishment of trust with sales personnel becomes a driving force for continued brand patronage: “I can make cost-effective purchases because I have a discerning eye.” In addition to proximity data, recognizing customer potential, incorporating traditional analog elements can provide a competitive advantage. Such strategies significantly stimulate customer engagement and foster brand differentiation within the market. The balance between digital and traditional customer recognition emphasizes the evolving nature of customer–brand relationships in the modern marketplace and the ongoing relevance of human touchpoints during such interactions.

Limitations and future research

Although this study has provided valuable suggestions for maintaining engagement, it is important to acknowledge its limitations. First, we surveyed customers with high actual concentration because of the gap between their behavioral intentions and actions (Bagozzi and Dholakia 1999), using five cases to reach a state of theoretical saturation (Eisenhardt 1989). However, according to Hennink and Kaiser (2022), in-depth interviews should ideally involve a minimum of nine interviewees, with an ideal sample size of 17 interviewees. Therefore, the number of participants in this study may not be sufficient to generalize the results. To achieve generalizability, further investigation with a larger sample size and cases that are more diverse is needed. Furthermore, since only one utility brand was examined, it is necessary to explore a broader range of utility levels to gain a deeper understanding of the drivers of maintained engagement. Lastly, given the differences in perception and thinking between East Asians and Westerners (Nisbett and Masuda 2003), conducting a study on maintained engagement among Westerners based on these findings could potentially provide more accurate generalizations.