Channel disintegration: a hidden key to customer satisfaction and vendor lock-in

The proliferation of marketing channels enables customers to evaluate the advantages and disadvantages of shopping activities across multiple stages of their journey (Liu and Liu, 2024). Omnichannel retailing is defined as “an integrated sales experience that melds the advantage of physical stores with the information-rich experience of online shopping (Verhoef et al., 2015).” Through this lens, omnichannel retailing integrates physical and virtual platforms to create a seamless customer experience (Chang and Geng, 2022). Despite the growing prevalence of omnichannel strategies in marketing practice (Radomska et al., 2024), many retailers continue to implement channel-specific differentiation. For instance, UNIQLO provides app-exclusive discounts that are not available in physical stores, while Zara staggers product releases across online and offline platforms. In the supermarket sector, Costco and Carrefour often assign product categories or promotions exclusively to one channel. Even without explicit strategic intent, discrepancies in pricing, assortments, and promotions remain prevalent across retail environments.

However, empirical research presents mixed findings on how channel disintegration influences customer behaviour. Gasparin et al. (2022) suggest that customers anticipate some inconsistencies across touchpoints and that such differences do not necessarily diminish the shopping experience. Customers understand the unique characteristics of each channel and often exploit them for personal gain (Kim and Han, 2023). Furthermore, Neslin (2022) contends that price consistency across touchpoints does not guarantee satisfaction due to varying customer sensitivities and switching costs.

Achieving seamless channel integration remains a considerable challenge for retailers aiming to enhance customer acquisition and retention (Gao and Huang, 2021). Integration initiatives require substantial investment in technological infrastructure and service platforms (Radomska et al., 2024). Moreover, omnichannel retailing demands strategic alignment across customer journey stages, including uniform pricing, and coherent delivery performance (Sorkun et al., 2020; Towers and Towers, 2022). The complexity of achieving cross-channel alignment in branding, promotion, pricing, and assortment continues to warrant further academic inquiry.

Given that online and offline channels attract distinct consumer segments, firms often tailor their strategies based on segment-specific price elasticity and purchase behaviour (Neslin, 2022). A uniform marketing mix may not be optimal, as consumer needs and expectations differ by segment and touchpoint. Prior research suggests that encounters with unexpected offers or product assortments across channels can elicit surprise (Vanhamme, 2000), which in turn affects information processing and decision-making (Shibly and Chatterjee, 2020). Although surprise is often associated with positive outcomes, it also leads to negative emotions, depending on how the situation is appraised (Ortony et al., 1988). Thus, surprise is considered as a valence-neutral emotional state that plays a fundamental role in shaping consumers' responses to unexpected stimuli. Accordingly, this study posits that customers may experience surprise in response to inconsistent marketing cues. In contrast to prior studies emphasizing full integration, the present research investigates whether disintegration enhances customer retention through the mechanism of surprise. This study examines the following research question.

Shopping motivation plays a pivotal role in consumer behaviour analysis, market segmentation, and marketing strategy formulation (Casteran et al., 2023). It reveals consumers' underlying goals in shopping behaviour (Tena-Monferrer et al., 2022). These goals, influenced by individual characteristics, shape consumers' channel preferences at various stages of the shopping journey (Kim et al., 2024). Accordingly, this study also examines the following question:

This study explores the relationship between channel disintegration and surprise, with anticipation serving as a moderating factor. It further examines the downstream effects of surprise on emotional outcomes, specifically delight, which influence satisfaction and vendor lock-in. In addition, three shopping motivations, namely efficiency shopping, bargain hunting, and variety seeking, are proposed to moderate the relationship between surprise and delight. The findings aim to offer strategic insights for retailers seeking to optimize multichannel marketing efforts to enhance customer retention.

Channel integration reflects the extent to which firms align multiple channels to enhance synergy and deliver seamless shopping (Lin et al., 2023). Customers generally value well-integrated channels that match their preferences. Prior research has highlighted the importance of integration quality (Le and Nguyen-Le, 2021; Lin et al., 2023), channel congruency (Djofack et al., 2023), integration mechanisms (Gao and Huang, 2021), and touchpoint complementarity (Lazaris et al., 2021). These studies emphasize consistency in content and service configuration as critical to customer experience. When integration is achieved, customers benefit from uniform prices and quality across channels, leading to seamless transactions (Gao et al., 2020). Accordingly, minimizing inconsistencies has been deemed vital for sustaining customer relationships (Park and Kim, 2019).

However, inconsistencies can also yield benefits by increasing perceived control, such as access to lower prices (Gasparin et al., 2022). For instance, Gasparin et al. (2022) note that customers often expect variations among touchpoints, which do not necessarily impair experience. Neslin (2022) similarly argues that price uniformity does not inherently enhance satisfaction, given price sensitivity and switching costs. These divergent findings indicate that the effects of channel disintegration merit closer examination.

Zhang et al. (2018) conceptualize channel integration as six dimensions: (1) integrated promotion, (2) integrated product and price, (3) integrated transaction information, (4) integrated information access, (5) integrated order fulfilment, and (6) integrated customer service. The first two relate to the marketing mix, while the latter four involve service and information. Integrated promotion denotes consistent presentation of monetary (e.g. discounts) and non-monetary (e.g. loyalty points) incentives; integrated product and price refer to uniform product availability and pricing. Integrated transaction information enables single-account tracking; integrated information access supports efficient search; integrated fulfilment ensures flexible completion; and integrated customer service unifies communication (Cheah et al., 2020). The overarching goal is to maximize consistency and accessibility across touchpoints.

This study focuses on two dimensions proposed by Zhang et al. (2018), integrated promotion and integrated product and price, for three reasons. First, promotional and pricing elements are directly tied to perceived transaction value (Hua et al., 2025) and thus exert greater influence on decision making than operational dimensions such as fulfilment. Evaluated at the pre-purchase stage, they are also more salient and emotionally charged (Neslin, 2022).

Second, while consistent services across touchpoints enhance convenience, inconsistencies can increase flexibility and autonomy (Gasparin et al., 2022). Shopping is not purely utilitarian; variations in promotions and pricing stimulate deal seeking and heighten hedonic value (Belwal et al., 2025). By enabling cross-channel comparisons, such variations enrich the experience and provide satisfaction from deal-seeking behaviours (Gasparin et al., 2022). Thus, disintegration may serve as a strategic lever rather than a service failure.

Third, algorithmic pricing exposes customers to dynamic promotions and individualized strategies (Grewal et al., 2020). These tactics, tailored to channels or devices, make inconsistencies in promotion and pricing more conspicuous. In contrast, operational elements such as fulfilment usually occur post-purchase and are standardized. Focussing on promotion and product/price therefore captures aspects of disintegration that are both frequent in practice and psychologically consequential.

Surprise is defined as “astonishment, wonder, or amazement” in response to unexpected stimuli (Valenzuela et al., 2010), which often elicit strong emotional reactions (Shibly and Chatterjee, 2020). Although affectively neutral, surprise may produce positive or negative outcomes (Ortony et al., 1988). Emotional assessments shape customer evaluations throughout the decision journey (Tueanrat et al., 2021). Delight, a positive affective state, arises when unexpected benefits or values are perceived (Ariffin and Omar, 2016). Arousal and pleasure are its core components. Accordingly, this study examines the influence of surprise, particularly delight, on satisfaction.

Shopping motivations have been widely studied. Büttner et al. (2013) classify consumers as goal-oriented shoppers, who prioritize efficiency, and experiential shoppers, who seek enjoyment. Aw (2019) extends this by identifying three activity-based motivations: efficiency shopping, bargain hunting, and variety seeking. Following Aw (2019), this study adopts this tripartite classification to examine how shopping motivation moderates the relationship between disintegration-induced surprise and emotional outcomes. Efficiency shopping emphasizes fast, effortless experiences (Wagner and Rudolph, 2010). Bargain hunting involves pursuing discounts and attractive prices (Wagner and Rudolph, 2010). Variety seeking reflects a preference for novelty, avoiding monotony, and maximizing stimulation (Gupta et al., 2020).

The surprise effect arises when individuals encounter stimuli that deviate from established cognitive schemas (Vanhamme, 2000). Congruent stimuli enhance cognitive processing and integration, whereas incongruent inputs elicit surprise (Ozer et al., 2020). Such schema-inconsistent events initiate attribution processes as individuals resolve expectancy–disconfirmation (Gendolla, 1997). In omnichannel retailing, promotional inconsistencies particularly trigger surprise when content is not uniformly integrated.

In an omnichannel environment, customers actively compare product attributes and prices across touchpoints, and event salience intensifies (Grobert et al., 2016). Customers often tolerate inconsistencies, engaging in behaviours such as webrooming to optimize choices (Gasparin et al., 2022). Channel differences help secure better deals or assortments. Marketing strategies may also create surprise through exclusive discounts or channel-specific products.

Surprise often arises from schema discrepancies, when customers encounter unexpected product features or promotional formats. Expectation moderates this effect (Sai Vijay et al., 2025). For example, an advertisement on an unusual platform may heighten attentional salience (Ozer et al., 2020). In omnichannel contexts, customers expect uniform promotions; thus, unexpected offers elicit surprise. When expectations are low, the impact of inconsistency becomes stronger.

Unexpected price variations create uncertainty, prompting affective and cognitive processing (Kovacheva et al., 2022). When customers expect price consistency but find lower prices or broader assortments in certain channels, they experience greater surprise. Discrepancies beyond expectations amplify this effect, with lower anticipation further intensifying it.

Surprise is a valence-neutral state that heightens sensitivity to subsequent stimuli (Kim and Mattila, 2013). When encountering surprise, customers evaluate outcomes to determine favorability (Ariffin and Omar, 2016). Positive outcomes, such as unexpected benefits or lower prices, amplify favourable emotions (Kovacheva et al., 2022). Gupta et al. (2020) also contend that surprise yields hedonic benefits that foster positive emotions.

This study emphasizes the positive side of surprise, as shopping offers psychological benefits rather than risks (Gardner and Rook, 1988). Purchasing restores a sense of control, enhancing positive affect (Rick et al., 2014). Unplanned purchases evoke pleasure, extending from retail therapy, which frames consumption as affective self-regulation (Zulauf and Wagner, 2022). Accordingly, channel disintegration may induce surprise when expectations are violated. When such deviations are appraised positively, surprise becomes a catalyst for delight.

The emotional impact of surprise is not always symmetrical (Kim and Mattila, 2010). Whether it evokes pleasant or unpleasant emotions depends on context (Alden et al., 2000). Customers with strong efficiency-driven motivation perceive shopping as more satisfying when convenient and streamlined (Cocco and Demoulin, 2022). They seek to minimize cognitive and time effort, preferring consistent service across touchpoints (Geng and Ning, 2025). High efficiency motivation reflects a stronger desire for convenience; thus, disintegrated marketing mix strategies may frustrate these customers. Consequently, the positive effect of surprise on delight may be weakened when efficiency motivation is high.

Bargain hunting is a price-driven strategy in which customers compare products across channels to find discounts (Kang, 2018). These consumers view themselves as savvy shoppers who gain satisfaction not only from savings but also from deal discovery (Aw, 2019). Lee et al. (2013) note that price-sensitive shoppers often enjoy dynamic pricing, such as bidding, which mimics a treasure-hunt. Horváth and Adıgüzel (2018) find that sales promotions heighten excitement and engagement among bargain hunters. Motivated by price differences, these customers exploit cross-channel inconsistencies. When comparisons yield financial gain, the surprise may be positively appraised and lead to delight.

Variety-seeking behaviour reflects a preference for novelty, change, and stimulation (Gupta et al., 2020). Variety seekers enjoy exploring unfamiliar products, following trends, and engaging with new experiences (Belwal et al., 2025). Lee et al. (2013) note that novelty adds excitement and a sense of adventure to shopping. In this study, disintegrated experiences meet their desire for change. When consumers encounter surprise through an unexpected promotion or expanded product range, the emotional reaction is more likely to be positive.

Satisfaction is a psychological state shaped by cumulative consumption experiences (Vanhamme, 2000). Emotions strongly influence evaluative judgements (Wu et al., 2015). Positive moods enhance satisfaction, and surprise-induced delight further strengthens it (Tueanrat et al., 2021). Kim and Mattila (2010) contend that affective responses shape cognitive evaluations. Thus, customers delighted by an inconsistent channel marketing mix evaluate the retailer more favourably.

Within-firm lock-in refers to a firm's ability to retain consumers across the search and purchase stages in a multichannel environment (Neslin et al., 2006). It captures continued interaction with the same retailer across channels, even when switching occurs (Maggioni et al., 2020). Because this study focuses on channel disintegration rather than brand-level loyalty, within-firm lock-in is a more relevant measure of cross-channel behaviour. Vendor lock-in reflects a firm's capacity to reduce switching and foster retention (Chiu et al., 2011). Prior research consistently supports the positive link between satisfaction and retention (Le and Nguyen-Le, 2021). Thus, satisfied customers are more likely to repurchase from the same vendor, regardless of channel.

Positive affect serves as heuristic input shaping evaluative judgements and purchase behaviour (Heilman et al., 2002). A good mood narrows decision scope and encourages spending (Xu, 2020). Delighted customers favour the retailer and show reduced switching. Xu (2020) also finds that positive emotions increase favourable online reviews. Thus, customers delighted by the service experience are inclined to remain loyal.

Relevant literature informed the measures, which were adapted to the study's context. Accuracy of item wording in both English and Chinese versions was verified through back translation. All constructs, except delight, were assessed on a 5-point Likert scale (1 = strongly disagree; 5 = strongly agree). Delight was measured with a 5-point semantic differential scale anchored by opposite descriptors. Sex and demographic questions were multiple-choice. The  Appendix lists all measurement items and sources.

Data were collected from customers of UNIQLO and GU in Taiwan, two apparel brands under an advanced omnichannel framework integrating mobile apps, physical stores, and social media. UNIQLO, which entered Taiwan in 2010, is the apparel market leader with 90% brand awareness (Wang and Wan, 2020) and operated 75 stores nationwide by 2023, sustaining growth even during COVID-19 (Cheng, 2023). GU entered in 2014 and has since become the brand's largest overseas market, reaching 23 stores in 2024. Its Taiwan sales quadrupled over the past decade, including a 200% rise in the last three years (Bhasin, 2024). Together, UNIQLO and GU, both under Fast Retailing, hold strong market influence.

Both UNIQLO and GU are well-established omnichannel retailers with advanced integration across multiple channels. While they provide seamless transitions between touchpoints, they also adopt selective differentiation strategies. Inconsistencies remain in product availability, pricing, and transactional procedures. For example, channel-exclusive items, price variations, or returns limited to the original channel. Such practices create perceptible inconsistencies within an otherwise integrated structure, making UNIQLO and GU suitable contexts for investigating intentional channel disintegration in a mature omnichannel environment.

After designing the questionnaire, feedback was obtained from three apparel industry experts and two e-commerce professors. A pilot test with 40 MBA students assessed reliability and validity, and minor adjustments improved clarity. The target respondents were customers who had shopped at UNIQLO or GU within the past six months. Convenience sampling was employed, and data were collected through a two-wave online survey using SurveyCake, distributed via omnichannel shopping and fast fashion forums and fan pages on Facebook and Instagram. To ensure data quality, three screening procedures were applied. First, respondents confirmed interaction with at least two brand touchpoints (e.g. mobile apps, stores) and one purchase within six months. Second, attention check questions (e.g. “Have you purchased from UNIQLO or GU in the past six months?”) identified inattentive responses. Third, incomplete responses or uniform patterns (e.g. straight-lining) were excluded.

The questionnaire emphasized actual shopping experiences rather than hypothetical scenarios. Respondents recalled and reported perceived inconsistencies across brand touchpoints, such as (1) price discrepancies between online and offline channels (e.g. an item discounted online but sold at full price in-store), (2) product availability mismatches (e.g. exclusive items only online or in-store), and (3) promotional variations (e.g. app-exclusive coupons or store-only discounts). Participants completed the survey based on these experiences. To incentivize participation, the study pledged a NT$10 charity donation per valid response and offered NT$800 in cash prize drawings.

The survey ran from September 1, 2023, to January 5, 2024. The first wave measured incongruence, surprise, and emotion. Of 540 invited participants, 521 provided valid responses. Three months later, the same participants were invited to assess satisfaction and vendor lock-in. To verify responses, participants provided purchase records or described their shopping experiences, and responses were matched by email to ensure anonymity. Ultimately, 450 joined the second wave, and after excluding 16 incomplete cases, 434 valid responses remained. Female respondents (65.3%) outnumbered males. Most were under 30 years old (42.1%), followed by ages 31–40 (21.5%). Additionally, 62.2% held a bachelor's degree, and 19.6% a graduate degree.

This study used partial least squares structural equation modelling (PLS-SEM) to test the hypotheses. PLS-SEM was selected as it estimates complex models with moderator variables (Becker et al., 2012). After preliminary screening, measurement reliability and validity were assessed. As shown in Table 1, three items with low standardized loadings (<0.50) were removed: one from inconsistent promotion (IP2), one from vendor lock-in (IA1), and one from variety seeking (VS2). The remaining items adequately captured their constructs, and all retained items loaded significantly, with Cronbach's α above 0.70 (Hair et al., 2017).

Content validity was ensured by aligning items with prior literature. Convergent validity was supported when each construct's composite reliability (CR) exceeded 0.7 and its average variance extracted (AVE) exceeded 0.5. Discriminant validity was examined using the Fornell–Larcker criterion (Fornell and Larcker, 1981) and the heterotrait–monotrait ratio (HTMT). As shown in Table 2, each construct's square root of AVE was higher than its correlations with other constructs, and HTMT values remained below the 0.8 threshold.

Figure 1 illustrates the indirect influence of channel disintegration on vendor lock-in through surprise, emotional response, and satisfaction. Inconsistent product and price positively affected surprise (β = 0.12, p < 0.05), but inconsistent promotion did not (β = 0.02, p > 0.05); thus, H₂ was supported, but H1 was not. Surprise positively influenced delight (β = 0.34, p < 0.001), which in turn enhanced satisfaction (β = 0.22, p < 0.001). Satisfaction significantly affected vendor lock-in (β = 0.31, p < 0.001), whereas delight's direct effect on lock-in was nonsignificant (β = 0.12, p > 0.05). Hence, H5, H9, and H10 were supported, but H11 was not.

Moderating effects were tested using the product-indicator approach in SmartPLS 4, which constructs latent interaction terms by multiplying indicators of the independent and moderator variables (Hair et al., 2017). Results showed that absence of anticipation moderated the effect of inconsistent product and price (β = 0.10, p < 0.05), supporting H4 but not H3. Bargain hunting (β = 0.12, p < 0.05) and variety seeking (β = 0.18, p < 0.001) strengthened delight's effect on satisfaction, supporting H7 and H8. Efficiency shopping did not moderate the surprise–satisfaction link (β = 0.05, p > 0.05), so H6 was not supported. The model explained 23% of variance in surprise, 31% in delight, 15% in satisfaction, and 12% in vendor lock-in.

To further examine the significant results of H4, H7, and H8, a multi-group analysis (MGA) was conducted. MGA compares structural path coefficients across subgroups defined by moderator levels (Sarstedt et al., 2011). Following Sparks et al. (1998), participants were classified into high (one standard deviation above the mean) and low (one below) groups based on absence of anticipation and shopping motivation (bargain hunting and variety seeking). This approach tested whether higher levels of these moderators strengthened the relationships.

As shown in Table 3, the link between inconsistent product/price and surprise was stronger for high absence of anticipation than for low (β_high = 0.554, p < 0.001; β_low = 0.049, p > 0.05; p < 0.05). For shopping motivation, the effect of surprise on delight was stronger for high vs. low bargain hunting (β_high = 0.575, p < 0.001; β_low = −0.281, p > 0.05; p < 0.001). Similarly, surprise more strongly influenced delight under high variety seeking than low (β_high = 0.475, p < 0.001; β_low = −0.257, p > 0.05; p < 0.001).

Inconsistent pricing and product assortment significantly impact surprise. Customers encountering a broader product range or lower prices across channels experience positive or negative surprise. This supports prior findings by Gasparin et al. (2022), who argue that certain inconsistencies enhance customer experience through perceived control and novelty. However, inconsistent promotions did not significantly affect surprise in this study. One possible explanation is that promotional discrepancies may be perceived as unfair or manipulative rather than rewarding. According to fairness perception theory, consumers evaluate whether they are treated equitably, particularly in contexts involving pricing or service delivery (Xia et al., 2004). For instance, when a brand offers a web-exclusive flash sale with substantial discounts but no equivalent offer in-store, customers who shop offline may perceive the tacit as coercive. Such experiences are more likely to evoke distrust than surprise, suggesting that not all types of inconsistency trigger surprise.

Additionally, the absence of anticipation did not moderate the relationship between inconsistent promotion and surprise. As Hayes (2018) suggests, moderation typically operates by conditioning an existing effect; when the baseline relationship is weak, the potential for moderation is inherently constrained. As promotional inconsistency had no significant effect on surprise, the absence of anticipation offered limited explanatory value.

This study also confirms the link between surprise and delight. Surprise, initially neutral, can evolve into a positive emotion (Kim and Mattila, 2013). The degree of delight depends on shopping motivations. Cost-oriented customers perceive surprise positively, as it enables them to secure better deals. Inconsistent channels encourage price comparisons, increasing perceived control. Variety seekers, who pursue novelty and self-expression (Horváth and Adıgüzel, 2018), also respond positively, as channel disintegration provides the variety they desire. This aligns with Grobert et al. (2016), who argue that surprise's emotional impact depends on perceived benefits. Contrary to expectations, efficiency shopping did not moderate the surprise–delight relationship. Efficiency-oriented shoppers may view surprises as inconvenient, preferring seamless experiences.

Finally, this study found that positive emotions enhance satisfaction, contributing to vendor lock-in. These findings align with Wu et al. (2015), who highlight emotions' role in service evaluations. Satisfaction is a key determinant of long-term customer behaviour. Thus, pricing and product inconsistencies can strengthen vendor lock-in by mediating surprise's influence on emotions.

This study differs from existing research in two significant ways. First, while previous research emphasizes integrated marketing mix elements in creating a seamless customer journey (Lin et al., 2023), Gasparin et al. (2022) challenge this fully integrative view. They propose that selective inconsistencies may improve customer experiences under certain conditions. However, their framework lacks empirical validation. This study demonstrates how specific forms of channel disintegration, namely promotional and product and price inconsistencies, elicit emotional responses. By introducing surprise as a mediator, this study explains how such inconsistencies trigger vendor lock-in. This study responds to Gasparin et al.’s (2022) recommendation: “a number of studies have concentrated mainly on dedication factors, such as user satisfaction and affective commitment, but the dedication perspective alone is not able to fully capture users' post-adoption decision-making processes” (p. 9).

Second, this study advances understanding of the valence of surprise, which is often treated as neutral but under-theorized in terms of its emotional outcomes (Grobert et al., 2016). While prior research has typically examined the direct effect of surprise on emotion (e.g. Gupta et al., 2020), this study shows that shopping motivation moderates the surprise-emotion relationship. Specifically, the moderating roles of efficiency-oriented shopping, bargain hunting, and variety seeking are examined to account for contextual differences in emotional responses to disintegrated experiences. This study offers insight into the contextual and motivational determinants of surprise valence, addressing Grobert et al.’s (2016) assertion that “whether surprise can be positive, neutral, or negative is unclear” (p. 240).

Omnichannel retailing integrates multiple channels to enhance the customer experience, yet maintaining consistency in pricing, product assortment, and transactions remains a challenge. When inconsistencies are perceived as opportunities rather than errors, they enhance customers' sense of control. Retailers are advised to adopt a strategic disintegration approach. Rather than striving for full standardization, firms can purposefully vary channel offerings based on behavioural insights. For example, “invisible” assortment strategies, such as restricting certain products to logged-in members, cultivate perceived exclusivity while avoiding overt inconsistencies. Furthermore, disintegration should be context-sensitive. Real-world examples include ZARA's online-exclusive product launches or Nike's differentiated availability across mobile apps and stores. Managers should monitor customer responses using behavioural analytics to fine-tune the degree of inconsistency without compromising brand trust.

Surprise functions as a high-impact marketing stimulus. As a valence-neutral, short-lived emotional state, it presents both opportunities and risks. To convert surprise into loyalty, firms must design it to align with customer expectations and context. Artificial intelligence (AI) technologies offer promising tools for this purpose. For instance, computer vision generates personalized recommendations based on image inputs, while AI-driven dynamic pricing and gamified content, deliver micro-surprises at various stages of the customer journey.

For bargain-oriented and variety-seeking customers, surprise is often welcomed, especially when linked to unexpected promotions or novel offerings. Retailers should regard shopping motivation as a core segmentation variable in designing disintegration-based strategies. Rather than applying uniform tactics, firms can utilize customer data, such as membership status, to classify customers based on behavioural traits. This enables the delivery of personalized, context-aware omnichannel experiences. A segmentation-based approach ensures operational efficiency, enhancing overall customer satisfaction.

While the empirical findings offer valuable implications for multichannel strategy development, several limitations should be acknowledged, each of which opens avenues for future research. First, although UNIQLO and GU were appropriate representatives of a mature omnichannel retail environment, the sample was limited to Taiwanese consumers, with a demographic skew toward female respondents (65.3%), limiting generalizability. Future research should include more demographically and culturally diverse samples.

Second, this study focuses on how channel disintegration influences vendor lock-in via the mediating role of surprise but does not explicitly integrate fairness-related frameworks. Constructs such as fairness and equity provide insight into consumers' evaluations of disintegrated experiences (Gao et al., 2020). Future studies could examine perceived fairness as a mediator or moderator. Third, this study conceptualizes surprise primarily as a positive emotion, which may not capture the full range of reactions. Disintegrated experiences may also evoke frustration. Future research should adopt experimental designs to reflect the full spectrum of emotions.

Fourth, the generalizability of the surprise-based mechanism depends on product type and purchase context. Effects are likely more pronounced in hedonic or high-involvement purchases, whereas for utilitarian products convenience may dominate. Future research should examine how product characteristics moderate outcomes. Fifth, this study is confined to business-to-consumer settings. In business-to-business environments, where purchases are rational, contract-driven, and involve multiple stakeholders, surprise may be seen as disruptive. Future research should explore how surprise operates in these contexts. Finally, although the study identifies a significant pathway from surprise to vendor lock-in, the explained variance is relatively low (R² = 12%), suggesting omitted variables. Future research could incorporate affective and cognitive antecedents of retention, as well as contextual factors such as market saturation.