Impact of omnichannel integration on Millennials’ purchase intention for fashion retailer

Abstract The COVID-19 pandemic has hastened the expansion of omnichannel purchasing globally including in Thailand. This quantitative study employs the stimulus-organism-response (S-O-R) framework to investigate the impact of channel integration on increasing customer satisfaction or reducing perceived risk, which leads to increased purchase intention for fashion products. Using an online questionnaire, data were collected from 400 Thai Millennials (aged 22–40 years) interested in buying luxury fashion products. The data were analysed using structural equation modelling. The results showed that integrated product and price, integrated promotion, and integrated information access influenced customer-perceived risk reduction, and integrated product and price, integrated promotion, integrated transaction information, integrated information access, and integrated customer service influenced customer satisfaction. In addition, perceived risk is related to customer satisfaction and purchase intention. These findings provide both theoretical and managerial implications for omnichannel retailers in developing their marketing strategies.


Introduction
COVID-19 has rapidly spread worldwide, affecting not only societies but also marketing ecology (Chen & Chi, 2021). Digital tools such as mobile applications, social media platforms, and e-commerce have been developed and become important marketing channels to facilitate consumers during the pandemic (M. Zhang et al., 2018). Omnichannel marketing is an online retail strategy that has gained impact from this revolution, especially in the fashion industry (Lorenzo-Romero et al., 2020). It has shifted from a single-channel to multi-channel, cross-channel, and then to omnichannel integration consequent to technological development (Truong, 2020). The

PUBLIC INTEREST STATEMENT
After the COVID-19 pandemic, fashion retailers are increasingly selling through the online channel, making their stores transform into omnichannel stores where sellers can reach customers on any channel. The omnichannel allows information to be accessible, whether it is information about goods, prices, promotions, services, or payments, and customers can access such information as well. However, when people want to buy something online, they tend to be concerned about their privacy or transaction data or whether the product information may not be correct. People who have a high perceived risk will not be satisfied and have an intent to purchase a product. It is crucial for businesses to find a suitable integrative marketing strategy to reduce consumer concerns, increase their satisfaction, and lead them to have the intention of buying a product. Therefore, this study was conducted to identify the critical factors of omnichannel that lead to increased Thai customer satisfaction and their purchase intention for fashion products.
omnichannel-in other words, channel integration-not only provides the consumer with a seamless experience across all channels (Lee, 2020), but also increases customers' touchpoints for retailers (Truong, 2020). To illustrate, starting with searching for information on a search engine such as Google, users usually click on links based on the results showing on the first page. Then, if the information on the site suits their queries, they tend to take some action or conversion or make a payment on that page. However, because people perceive a risk regarding online shopping, this may increase their privacy concerns (Cheah et al., 2022) and reduce their satisfaction (Lee, 2020). Therefore, it is crucial for businesses to find a suitable integrative marketing strategy for sales channel management to increase customer purchase intention and sales, especially during and after a crisis.
In Thailand, according to Euromonitor International (2021),because of the COVID-19 lockdown situation in which physical stores were forced to close, many fashion retailers expanded their stores to online channels. Similarly, customer behaviour also changed during this crisis (Izmirli et al., 2021). Thai people tend to purchase products online and use cashless methods such as QR codes or mobile banking. This finding is supported by data from DataReportal (2021), which reveals that more 80% of the Thai population visit shopping apps and 83.6% spend their money purchasing products online. Moreover, Thai consumers, especially Millennials aged between 22 and 40 years, normally search for product information on online sites during both online and offline purchasing decision processes (Euromonitor International, 2019). According to Kim et al. (2017), people tend to search for product information offline and then conduct transactions online at a lower price. In contrast, some people might search for information online and be concerned about uncertain online transactions, leading them to decide to buy products at physical stores instead (Juaneda-Ayensa et al., 2016). Thus, it is necessary for retailer businesses to integrate all channels, which is also known as the omnichannel, dual channel, channel integration, or integrated retail mix, to satisfy customers' needs (Kim et al., 2017) and reduce their concerns (Chen & Chi, 2021).
Most prior research in this area has been conducted from the retailer's perspective. For instance, while research has focused on omnichannel management (Cai & Lo, 2020) or omnichannel supply networks (Izmirli et al., 2021), studies on the factors that influence purchase intention through omnichannel marketing are lacking. Moreover, studies related to omnichannel consumer purchasing behaviour and the perception of channel integration in the omnichannel fashion retail environment are lacking. Piotrowicz and Cuthbertson (2014) state that even when trends change, the channel integration environment remains a huge challenge for retailers because they lack information regarding digital marketing channels or customer views across distributions. affects people's internal state, also called organism (O), which is a factor that motivates various personal responses (R) that could be positive or negative behaviours such as acceptance or avoidance. Zhu et al. (2020) confirmed that the stimulus influences an individual's attitude, while an individual's specific behaviour reflects the external response. This S-O-R framework has been broadly utilised in previous research on the connection between retailers and consumer shopping behaviour (Chen & Chi, 2021;Lee, 2020;M. Zhang et al., 2018). For instance, M. Zhang et al. (2018) found that channel integration, as a stimulus variable, affects Chinese consumer empowerment, leading to increased customer satisfaction and trust, similar to patronage intention. Lee (2020) found that some omnichannel characteristics directly impact customer satisfaction, which is considered an organism and strongly affects the store revisit intention of Korean consumers. Furthermore, in 2021, Chen and Chi utilised the S-O-R framework to investigate how channel integration increases consumer behavioural responses in the US with the moderating effect of perceived COVID-19 vulnerability.
Therefore, similar to previous studies, this research considers six characteristics of channel integration as stimulus factors. Customer satisfaction is considered an organism. However, to extend the existing theoretical model, this study adds perceived risk in the organism stage. This is because people feel uncertain when they perceive a new channel (Herhausen et al., 2015). In addition, as previous studies focused on intention to revisit omnichannel (Lee, 2020) or patronage intention (M. Zhang et al., 2018) for response factors, in this study, the purchase intention of Millennials is considered to determine the response factors in the S-O-R framework. Verhoef et al. (2015) define an omnichannel as "the synergetic management of the numerous available channels and customer touchpoints, in such a way that the customer experience across channels and the performance over channels are optimised". According to Lee (2020), omnichannel retailing is a retail concept in which all channels are combined or integrated to increase customers' experiences and firm performance. All channels in the omnichannel environment include physical stores, telephones, online shops, and mobile stores (Beck & Rygl, 2015).

Omnichannel integration
The omnichannel environment has developed from multi-channel retailing (Juaneda-Ayensa et al., 2016), where distribution channels are available on both online platforms and physical stores (Li et al., 2018). However, significant differences have been observed between these channels. As supported by Hsia et al. (2020), the omnichannel environment is different from a multi-channel one. In multichannel retail, each channel is not necessarily connected, whereas omnichannel methods provide a more comprehensive experience through all channels. For instance, the multi-channel retailer treats customers separately between each distribution and is served by different staff who do not work collaboratively (Piotrowicz & Cuthbertson, 2014). On the other hand, the omnichannel environment helps retailers display all customers' touchpoints and increases customers' procurement by utilising various channels (Kazancoglu & Aydin, 2018). Furthermore, the amount of consumer data collected at every touchpoint increases as additional channels are provided (Hossain et al., 2017). Digitalisation, social media, big data, and other emerging technologies such as artificial intelligence, virtual reality, and augmented reality are transforming retail business models, rendering omnichannel retail a popular strategy (Cai & Lo, 2020).
In this context, fashion companies combine online and offline marketing channels (Lorenzo-Romero et al., 2020). Although Cai and Lo (2020) argue that the omnichannel strategy cannot be applied to every business, fashion retailers can benefit from it (Lorenzo-Romero et al., 2020). By adopting an omnichannel strategy, fashion companies can interact with their clients through direct and indirect channels (Lorenzo-Romero et al., 2020). These authors provide an example in which firms can connect directly with customers via physical stores where customers can try on outfits. On the other hand, fashion businesses could involve indirect interaction with clients by using Internet channels where they can search for product information. In this environment, customers can purchase products from each channel. In addition, Li et al. (2018) and Gao and Yang (2016) confirm that omnichannel fashion retailing can provide customers with seamless experiences in which they are allowed to create action across channels with no differences in the areas of product and pricing, promotion, transaction information, information access, order fulfilment, and customer service.

Integrated product and price (IPP)
The growth of mobile devices has enabled people to easily access product and price information (Fulgoni, 2014). In an omnichannel environment, the same product and price information is provided on all channels (M. Zhang et al., 2018). Fulgoni (2014) argued that omni-retailers should provide lower-price online channels to increase customers' online purchase intentions. Similarly, in Korea, Lee (2020) found that the same product price on both online and offline channels increased consumer dissatisfaction. However, integrating product and price information can benefit customers when they need to obtain accurate product quality or price information, which can increase customer satisfaction during the shopping experience (M. Zhang et al., 2018). Chen and Chi (2021) demonstrated that well-integrated products and prices among sales channels successfully decreased consumers' perceived risk in the US. Moreover, it is crucial that retailers deduct a product and price data mismatch between platforms, because providing dissimilar product and price information across all channels may increase consumers' concerns (Piotrowicz & Cuthbertson, 2014). Therefore, this study assumes that if customers perceive integrated product and price information across channels in the fashion retail environment, their perceived risks will be decreased and their satisfaction increased.
H1a: Integrated product and price has a positive influence on perceived risks.
H1b: Integrated product and price has a positive influence on customer satisfaction.

Integrated promotion (IP)
Some previous research described integrated promotion as promotion on one channel that promotes another (J. Zhang et al., 2010). In this context, integrated promotion means that a promotion campaign's message should be consistent across all places (M. Zhang et al., 2018). For example, some retailers send both emails and catalogues to their clients to demonstrate channel choices (Lee, 2020). This means that if the company promotes products online, it should also promote them in physical stores. In addition, customers can obtain and use sales promotions for every channel (Koo, 2020). M. Zhang et al. (2018) noted that integrated promotion has the ability to increase user satisfaction, as they are aware of the connection between each channel. Furthermore, promotion integration can reduce customer concerns and uncertainty by increasing retail trustworthiness (Li et al., 2018). For example, sometimes information about promotions may be mismatched between channels, which could be concerning to the consumer. Thus, the research hypotheses are as follows: H2a: Integrated promotion has a positive influence on perceived risks.
H2b: Integrated promotion has a positive influence on customer satisfaction.

Integrated transaction information (ITI)
Omnichannel shoppers can move freely across platforms during the transaction process (Kazancoglu & Aydin, 2018). M. Zhang et al. (2018) contended that integrated transaction information means that retailers' clients can track or manage their purchase records on every channel. Retailers can examine accurate records of customers' previous usage behaviours to acquire relevant insights and use them to predict shopping behaviour intention (Kliestik, Kovalova et al., 2022). This tends to reduce a client's uncertainty and risk perception (Li et al., 2018). Customers can also decide which channels are suitable for their transactions based on their needs (M. Zhang et al., 2018). Lee (2020) argued that in Korea, the integration of transaction information does not have a positive influence on customer satisfaction. However, Wang et al. (2021) confirmed that customer satisfaction could be improved when the uncertainty of perceived transaction information on web stores in dual-channel retailers decreases. Consequently, we assume the following: H3a: Integrated transaction information has a positive influence on perceived risks.
H3b: Integrated transaction information has a positive influence on consumer satisfaction.

Integrated information access (IIA)
Channel integration allows customers to easily switch to other companies' channels to access the information provided via various channels (M. Zhang et al., 2018). An example of information access integration is a situation in which people find products available via kiosks in a local store. Similarly, they can search online for available stock instead of making wasted trips to the physical store if no product is available (Oh et al., 2012). Allowed and convenient access to information increases customer satisfaction during their shopping activities (Lee, 2020). In addition, in research by Ortlinghaus et al. (2019), risk perception regarding product availability tends to decrease if the product actually exists. Goraya et al. (2022) also reveal that consumers' perceived risk of nonavailability will be solved through a cross-check on both channels. Hence, the following is hypothesised for the purposes of this study: H4a: Integrated information access has a positive influence on perceived risks.
H4b: Integrated information access has a positive influence on consumer satisfaction.

Integrated order fulfilment (IOF)
Customer order fulfilment has become increasingly crucial in omnichannel retailing, as it requires a large amount of effort to link logistical networks and product flows (Wollenburg et al., 2018). Wollenburg et al. (2018) stated that while additional fulfilment options are expensive, they enable better customer service through cross-channel substitutes and faster deliveries. Consumers can order, pay delivery, return, or exchange a product regardless of where they purchase it (M. Zhang et al., 2018). They can select the channel they prefer to fulfil their order; for example, some customers may order products online and pick up or pay at a local store (Oh et al., 2012). Nevertheless, clients may be concerned and dissatisfied if their orders are incorrect (Lee, 2020). However, the feature of channel integration could reduce concerns regarding incorrect orders. Lee (2020) revealed that customers' order fulfilment process in an omnichannel environment is a multiplicity in which retailers must provide methods for them to check the accuracy of their orders, which increases satisfaction. The omnichannel retailer may send the client an email or a message to confirm that the correct product has been purchased after the transaction. Therefore, the following is hypothesised for the purposes of this study:

Integrated customer service (ICS)
Customers obtain standard and consistent services across channels in an omnichannel environment (M. Zhang et al., 2018). This environment enables answering customer questions instantaneously via an online platform such as a chatbot, live chat, or direct message. These platforms also enable recording data on a client's requirements on the cloud, which can be used for both online and offline services (Pantano & Viassone, 2015). This integration could increase customer satisfaction if the service is of a high quality and can fulfil their needs (Lee, 2020). Moreover, high-quality service integration may lower customer risk perception or uncertainty due to the alternative channel provided for any support (Herhausen et al., 2015). Thus, the following is proposed: H6a: Integrated customer service has a positive influence on perceived risks.
H6b: Integrated customer service has a positive influence on consumer satisfaction.

Perceived risk (PR)
The uncertainty that occurs during a purchase decision is the meaning of risk in the buying state (Ortlinghaus et al., 2019). Perceived risk can also be explained in several dimensions based on previous research (Dai et al., 2014). The most important dimensions in the online retailing situation are risks related to merchandise performance, convenience, privacy, financial impacts, psychological issues, or product delivery (Ortlinghaus et al., 2019).
Prior research by Herhausen et al. (2015) found that as a new retail function, omnichannel shopping is perceived as a risk, especially in terms of performance or finances. However, Li et al. (2018) argued that channel integration tends to prevent data misunderstanding, which leads consumers to have more confidence and reduces their risk during shopping. Quach et al. (2022) agreed that retailers adopting an omnichannel strategy could reduce customers' perceived risk, because service and product information consistency and transaction information integration are provided, which could also increase customer satisfaction. Thus, it is proposed that: H7: Perceived risk has a positive impact on customer satisfaction. H8: Perceived risk has a positive impact on purchase intention.

Customer satisfaction (CS)
According to Anderson and Srinivasan (2003), consumer satisfaction is a continuous assessment of the element of surprise that comes with buying and using a product or service. This situation occurs when the shopper experience is fulfilled at least as well as expected (Oliver, 2014). Customer satisfaction is not a new concept; it has been mentioned in several studies in terms of its effect on shopping intentions (Hsu et al., 2011). It also provides an opportunity to increase company sales, profits, and market share (Ghotbabadi et al., 2016).
Customer satisfaction plays a crucial role in enhancing purchase intention (Lee, 2020). Prior studies found that customers need to evaluate received information and other technology features before buying. Here, channel integration is important, as the consistency of provided information increases customers' satisfaction in the omnichannel environment (Ghotbabadi et al., 2016). Furthermore, the strong connection between channels in the omnichannel environment is a key factor in product purchase intention (Park & Park, 2016). Similar research related to fashion omnichannel retailing in China showed that Chinese people are willing to spend more time and energy in an omnichannel store because their satisfaction has increased through information integration (Gao & Yang, 2016). Thus, if Thai consumers are satisfied with omnichannel integration, they intend to purchase fashion products. Therefore: H9: Customer satisfaction has a positive impact on purchase intention.

Millennials' purchase intention
2.5.1. The Millennial generation Millennials, or Generation Y, are born between 1984and 1999(Bento et al., 2018. This generation is normally familiar with technology, whether it is a mobile, application, or new technological development, and is likely to spend money on fashion (Baykal, 2020). Baykal (2020) suggested that the valuable information provided across channel integration is an important strategy for retailers to satisfy customers. Further, consistency across channel integration could meet Millennials' demands in terms of accessibility and an interactive purchasing experience (Parment, 2013).

Purchase intention (PI)
In the S-O-R framework, purchase intention is considered a response: Here, a response is the outcome behaviour influenced by external stimuli (Pantano & Viassone, 2015).
Previous research on customer purchase behaviour showed that especially during the COVID-19 pandemic, Millennials purchase tailored products through online channels using voice bots and algorithms to avoid social involvement (Nica et al., 2022). In the online environment, customer data gathered from social media can be used to determine a brand's customer sentiment and to shape positive customer behaviour (Kliestik, Kovalova et al., 2022). Similarly, Andronie et al. (2021) found that positive attitudes developed based on perceived risk and trust consequences regarding online shopping result in more frequent purchases and value perception. Moreover, augmented reality personalisation, which is provided by the omnichannel retailer, delivers customer value (Hopkins, 2022). Furthermore, artificial intelligence in algorithms reduces perceived risk, which is related to increasing users' intention and social value creation (Hopkins, 2022;Kliestik, Zvarikova et al., 2022;Nica et al., 2022).
In Thailand, while online shopping has increased significantly because of the COVID-19 pandemic, some Thai Millennials still want to visit physical stores (Manakitsomboon, 2021). In addition, some of the younger generations have changed their attitude towards fashion products to demonstrate more sustainable consumption during COVID-19 (Vătămănescu et al., 2021). Parallel research conducted on omnichannel shopping in Turkey found that the non-specific channel shopping experience provided by the omnichannel retailer influences customers' willingness to purchase in that environment (Kazancoglu & Aydin, 2018). On the other hand, in online food delivery services (Rowland, 2022), online ordering decreases consumers' concerns regarding the COVID-19 crisis, leading them to order more frequently.
However, there is a gap in research on the omnichannel environment in Thailand, specifically how omnichannel retailing affects customer purchase intention. Hence, this study examines this issue to help retailers in Thailand in determining the most important channel to us, one with the ability to satisfy customers, reduce their perceived risk, and increase their purchase intention (Figure 1).

Sampling and data collection
Snowball sampling was used in this study, as this technique allows the researcher to classify a primary sample that can provide a research questionnaire to others (Atkinson & Flint, 2004). The questions were designed based on M. Zhang et al. (2018), Li et al. (2018), and Lee (2020), and Shi et al. (2020) using a seven-point Likert scale (1 = strongly disagree to 7 = strongly agree; Table 1). The 35 questions included items related to six characteristics of an omnichannel. The results of the reliability tests ranged from 0.867 to 0.935. Furthermore, Cronbach's alpha was 0.985 > 0.70 (Table 1), indicating high reliability.

Ethical considerations
The study was approved by the Ethics Committee of Human Research, Walailak University (WUEC-21-267-01) on 17 September 2021. Respondents provided their informed consent to participate in the survey. They could withdraw from the project at any time without any consequences.

Demographic statistics
The primary data in this research were collected using SurveyMonkey-a survey-based platformfrom 400 Thai Millennials aged between 22 and 40 years, who have experience purchasing fashion products on both online and offline channels of the same brand. The results showed that 76.50% of the participants were female and 23.50% were male. Most respondents were aged 22-30 years (83.50%), followed by those aged 31-40 years (16.50%). The largest group of respondents (72%) had a bachelor's degree. Furthermore, most respondents occasionally shop for fashion products at fashion omnichannel retailers (29.75%).

Data analysis
A three-stage data analysis was performed using SPSS 17.0 and AMOS 21.0. In the first stage, the descriptive statistics were calculated of the variables of the impact of omnichannel integration on Millennials' purchase intention for fashion retailer. Second, a confirmatory factor analysis (CFA) was conducted to verify the proposed model. Finally, structural equation modelling (SEM) was employed to test the relationship between the channel integration variables, consumers' perceived risk, consumer satisfaction, and purchase intention for fashion products.

Descriptive statistics
Using the mean and standard deviation analysis method, the results of the analysis indicate that the overall average mostly agrees: The mean ranged from 5.6 to 5.97, as illustrated in Table A1 in Appendix A. Furthermore, the descriptive data distribution analysis shown in table A2 provides that the data has a normal distribution and is considered acceptable. Furthermore, the correlation coefficient ranges from 0.702 to 0.854, which is less than 0.90, and the correlation is significant at the 0.001 level. Thus, the construct validity in this paper was ensured and did not have multicollinearity, and the data revealed to explain the result of table A3. A consistent product description is provided on both online and offline channels.
A consistent product price is provided on both online and offline channels.
A consistent product category classification is provided on both online and offline channels.

Integrated Promotion (IP)
M. Zhang et al. (2018), Lee (2020) 0.929 Promotions on online channels are same as at the physical store.
I can find advertisements for the products on both the online and offline stores.
The retailer's website provides the address and contact information of the physical store.
I can find the same brand name, logo, and slogan of the retailer on their website, social media channels, and at the physical store.
I can redeem the retailer's vouchers in the physical store or on their website.

Integrated Transaction Information (ITI)
M. Zhang  Post-purchase services support for products purchased is provided at the retailer's physical store and website.
I can access the service assistant through a real-time chat program through the retailer's website and social media platforms. Perception of channel integration reduces my concern that this retailer may have misrepresented this product in the website description.
Perception of channel integration reduces my concern that this retailer will deliver the product as promised in a timely manner.
Perception of channel integration reduces my concern that this retailer has fully disclosed all product defects.
Perception of channel integration reduces my concern of a high degree of uncertainty regarding the quality of this retailer.
My concern that there's a high risk is decreased when making omnichannel purchase decisions.

Confirmatory factor analysis (CFA)
According to Harrington (2009) The CFA of the proposed model strongly suggests that each set of items represents a single underlying construct and provides evidence of discriminant validity. In addition, the AVE scores for all constructs were greater than 0.50, ranging from 0.665 to 0.765, and the composite reliability (CR) values for all constructs in the model were more critical than the threshold value of 0.70 (Fornell & Larcker, 1981). Finally, the composite dependability was greater than 0.60, ranging from 0.871 to 0.940, indicating good validity (Table 2).
The omnichannel retailer has met my expectations.
The omnichannel retailer has offered me exactly what I need.
The omnichannel retailer provided me with good experiences. Shi et al. (2020) 0.867 I have an intent to purchase the products(s) from this retailer.

Purchase Intention (PI)
I have an intent to recommend this retailer to my friends.
I have an intent to make another purchase from this retailer if I need the products I will buy.
Overall 0.985 a positive influence on PR), H4b (IIA has a positive influence on CS), H6b (ICS has a positive influence on CS), H7 (PR has a positive impact on CS), H8 (PR has a positive impact on PI), and H9 (CS has a positive impact on PI). However, the other hypotheses were rejected (H3a, H5a, H5b, and H6a) at a significance level of 0.05.

Discussion
During the COVID-19 pandemic, consumer behaviour in Thailand changed dramatically. Most people typically seek product information on both offline and online channels (Euromonitor International, 2019). Consumers are increasingly using several channels throughout their shopping process (Kim et al., 2017). Therefore, it is crucial for omnichannel fashion retailers to provide consumers with an integrated retail mix across channels, allowing them to finish their shopping effortlessly as if shopping on a single one of the retailer's channels (Kim et al., 2017). This study applied the S-O-R framework to determine what factors related to channel integration perception increase customer satisfaction or decrease perceived risk, and thus increase customer purchase intention for fashion products.
The results indicate that of the six characteristics of retail channel integration, three influenced customer perceived risk reduction, and five impacted customer satisfaction. In addition, perceived risk was related to customer satisfaction and purchase intention.
First, consistent with the results of Chen and Chi (2021), this study found that the product and price information provided on both online and offline channels reduces customer-perceived risk, as this practice decreases the uncertainty of a data mismatch among sales channels. Moreover, while Lee (2020) found that people in Korea are dissatisfied when a product is offered at the same price on omnichannels, the current study confirmed that integrated products and prices improved the satisfaction of Millennials in Thailand.
Second, integrated promotion seems to have the highest impact on consumer-perceived risk reduction and satisfaction. These results support the findings of J. Zhang et al. (2010)  * Significant at the 0.05 level credibility tended to decrease when they found consistent promotional messages and brand identity across channels. Furthermore, the ability to use promotional vouchers in dual channels led to satisfaction.
Third, integrated information access has a positive effect on perceived customer risk and satisfaction. One possible reason could be that Thai Millennials search for fashion product information on both online and offline channels. For the omnichannel fashion retailer, consumers' satisfaction will improve when they can search online and shop at the physical store should they want more product information on inventory status or reviews, for example. This finding is supported by Lee (2020). Furthermore, as Ortlinghaus et al. (2019) suggested, the level risk perception may be reduced if the purchaser finds information regarding the availability status of the desired fashion product.
Similarly, integrated customer service has a positive influence on increasing the satisfaction of Millennials in Thailand with fashion omnichannel retailers. These results substantiate those of Herhausen et al. (2015) and Lee (2020). The findings imply that Thai Millennials seem satisfied with real-time online customer service, even if buying a fashion product in-store.
However, in this study, integrated customer service did not have a positive influence on perceived risk. Several studies showed that customer service may decrease perceived risk (Ghotbabadi et al., 2016;Herhausen et al., 2015;Sun, 2014). This could be because while customer service from the omnichannel retailer creates a seamless experience, the level of risk perception is only reduced by high-quality service (Sun, 2014). Moreover, the ability to order fashion products online and then pick them up in-store for seamless order fulfilment did not play a significant role in this research. Rather, the research results demonstrated that integrated order fulfilment had no impact on either reducing customer perceived risk or improving customer satisfaction. Furthermore, although M. Zhang et al. (2018) discovered a relationship between integrated transaction information and perceived customer risk, integrated transaction information did not play a significant role in this study. Unexpectedly, this study also found that integrated transaction information had a negative effect on consumer satisfaction. The complication of accessing transaction information by allowing customers to check their purchase history across online and offline channels seems to reduce satisfaction levels, supporting the findings of Rangaswamy and Van Bruggen (2005). Therefore, Thai consumer satisfaction decreases when information on fashion product transactions can be accessed solely through the transaction channel.

Conclusion
In summary, a well-integrated omnichannel fashion retailing environment for product and price, promotion, and information access will reduce the risk perceived by Millennials in Thailand, especially in terms of mismatched retail mixed data across all channels. Furthermore, integrated product and price, promotion, transaction information, information access, and customer service will increase their satisfaction and product purchases.

Theoretical implications
Theoretically, the research framework used in this research was adapted from the S-O-R framework developed by Mehrabian and Russell (1974). The authors aimed to investigate the impact of the external environment on people's internal state and the resulting behavioural responses. As an extension, this study has made efforts to apply the S-O-R framework in the context of an omnichannel fashion retailer in Thailand during the COVID-19 pandemic. As such, this study is the first to examine the significant role of the perception of omnichannel characteristics in customer-perceived risk, satisfaction, and consumer fashion product purchase intention. This study was also the first to consider perceived risk as an organism (O). Furthermore, the results indicate that the integration of order fulfilment, transaction information, and customer service is not essential in reducing customer-perceived risk. Similarly, integrated order fulfilment does not much attention to increase satisfaction in this theoretical model when utilised in research in this field in Thailand.

Managerial implications
The results of this study have several managerial implications. The outcomes of the research are expected to help omnichannel fashion retailers understand which characteristics of channel integration are crucial for consumers in the Millennial age group to decrease their risk perception, increase their satisfaction, and increase their purchase intention. This will aid in generating better marketing plans to suit customer needs and improve competitive advantage. The results of this study may also help omnichannel fashion retailers in terms of channel strategy and investment. For example, as a result of the integration of product and price, promotion, and information access, which impacts customer perceived risk, omnichannel fashion retailers in Thailand could invest more in terms of the consistency of information regarding these characteristics on both online and offline channels. This would reduce the concerns of Millennial customers regarding the product they want and willingness to buy that product. This will also reduce costs for fashion retailers, because they would not have to spend much on the transaction information and orderfulfilment process, which is very expensive. Furthermore, fashion retailers could develop integrated marketing strategies for aspects pertaining to product and price, promotion, information access, customer service, and transaction information to improve customer satisfaction.

Limitations and future research
This study has several limitations. First, it focused only on the consumption of fashion products. According to M. Zhang et al. (2018), customers may react differently to channel integration for tangible products, either hedonic or utilitarian. Therefore, it may be worthwhile to investigate the impact of other product types on the relationship between channel integration and consumer purchase intention in the future. Second, although this research was conducted during the COVID-19 pandemic, it did not include the perception of COVID-19 as a moderator, which should be considered in the future. Last, because this research used a qualitative approach, future research could adopt a mixed-method research design by including systematic reviews, interviews, or focus groups with omnichannel consumers for more in-depth insights. Perception of channel integration reduces my concern that this retailer will deliver the product as promised in a timely manner.