Feeling Confident and Smart with Webrooming: Understanding the Consumer's Path to Satisfaction

The multichannel marketing literature consistently shows that consumers who use multiple channels in their purchase journeys are more satisfied, loyal, and can be more profitable, than single-channel consumers. However, there is little research investigating how specific channel combinations affect the customer experience. Recognizing that webrooming (research products online, purchase offline) is the prevalent form of cross-channel shopping, this paper examines its influence on the consumer's search process satisfaction. The results of three studies combining qualitative, survey, and experimental methods show that webrooming leads to more satisfaction than showrooming behaviors. Furthermore, we find that webrooming makes consumers feel more confident and like “smart shoppers.” Both factors subsequently determine satisfaction. Perceptions of money savings also affect search process satisfaction. Importantly, saving time and/or effort during the purchase process (convenience) has no influence on satisfaction with cross-channel shopping. The results are robust across shopping motivations and product categories. Theoretical implications and proposals for effective channel integration are offered.


Introduction
Consumers easily interact with online and offline channels to search for information and buy products. They generally adopt two shopping patterns. With showrooming, they visit physical retailers to search for information and then log on to the Internet to make the purchase. According to eMarketer (2014), 72% of U.S. digital shoppers bought after seeing a product in a store, and 10% of European online users research products offline before buying them online (Google Consumer Barometer 2015). With webrooming, consumers research products online and then make their purchase offline. Webrooming is the dominant cross-channel combination; 78% of U.S. shoppers (eMarketer 2014) and 42% of European online users (Google Consumer Barometer 2015) engage in it. When making large, expensive purchases, consumers may spend over a month researching product information through online and offline sources, and almost two-thirds go through a webrooming shopping experience, whereas 30% prefer to showroom (PushOn 2018).
These shopping patterns afford firms less control of the customer experience and may threaten retailers in the form of free-riding behaviors (Chiu et al. 2011;Lemon and Verhoef 2016). However, past research reports that consumers who use multiple channels purchase more products, spend more and are more satisfied than single-channel consumers (Chatterjee 2010a(Chatterjee , 2010bLee and Kim 2008;Piercy 2012; Van Baal and Dach 2005). Satisfaction is one of the key elements of customer experience management (Lemon and Verhoef 2016) and is the cornerstone for retaining and establishing long-term customer relationships (Wallace, Giese, and Johnson 2004). Yet, however, there are few studies that investigate the specific channel combinations that influence consumer satisfaction. This research aims at filling this gap in the literature.
The objective of this research is to analyze how webrooming (versus showrooming) influences satisfaction with the search experience. We focus on the consumer's feeling of confidence in the adequacy of the product, and his or her feeling of being a smart shopper, as drivers of the influence of webrooming on search process satisfaction. We investigate these relationships in a two-stage methodology, with three studies, combining survey-based data, qualitative-based data and experiment-based data.
We contribute to the literature by confirming that webrooming is the most effective cross-channel combination to increase satisfaction, which, in turn, enhances customer loyalty and determines a firm's long-term survival (e.g. Jang, Prasad, and Ratchford 2017). First, we find that webrooming is undertaken more frequently and is more satisfactory to the consumer than other single-channel and cross-channel behaviors. This may be due to the customer's perception, felt with a high degree of confidence, that he or she has made the right purchase. Second, we demonstrate that webrooming increases satisfaction with the search process (hereafter, search process satisfaction), compared to showrooming, and that confidence and smart shopping feelings explain this effect. The perception that money is saved also contributes to search process satisfaction, but convenience appears to have no influence.
The results are fairly consistent across different shopping motivations and types of product. Thus, companies who want to integrate channels and take advantage of channel synergies should embrace webrooming as the cross-channel behavior most likely to enhance consumer satisfaction. Allowing the customer to feel in control, feel confident and experience smart shopping feelings is the best path to elicit feelings of satisfaction.

Literature Review
Our definition of webrooming is in line with the cross-channel (e.g., Van Baal and Dach 2005) and research shopping (e.g., Verhoef, Neslin, and Vroomen 2007) streams, and follows previous conceptualizations of two-stage decision-making processes (Balasubramanian, Raghunathan, and Mahajan 2005;Peterson, Balasubramanian, and Bronnenberg 1997). Webrooming is a cross-channel process with a decision-making phase divided into two parts. In the first stage, the consumer searches for and finds on the Internet an alternative product that probably best matches his or her needs; in the second stage, the consumer confirms the information at the physical store and makes the purchase.
The previous literature generally adopts an economic perspective to analyze webrooming, in which consumers weigh up the costs and benefits of channel use during the purchase process. Shoppers combine the channels that allow them to minimize their inputs (e.g., time, effort, money, perceived risk) and/or maximize their outputs (e.g., making the right purchase, finding better deals, hedonic value) (Alba et al. 1997; Gensler, Verhoef, and Böhm 2012; Kulkarni, Ratchford, and Kannan 2012; Noble, Griffith, and Weinberger 2005; Pauwels et al. 2011;Ratchford, Lee, and Talukdar 2003;Verhoef, Neslin, and Vroomen 2007).
In this way, motivations, goals and schemas determine consumers' informational needs and lead to different channel usage (e.g., Balasubramanian, Raghunathan, and Mahajan 2005;Burke 2002;Dholakia et al. 2010;Pauwels et al. 2011;Puccinelli et al. 2009). Goal-directed consumers, and those who demand convenience, combine channels to maximize shopping efficiency (Noble, Griffith, and Weinberger 2005). For these consumers, the Internet saves time and effort in searching for product information, and the physical store offers immediate possession of the merchandise (Heitz-Spahn 2013). Important purchases with high implicit risks affect the use of crosschannel shopping (Jang, Prasad, and Ratchford 2017;Piercy 2012;Ratchford, Lee, and Talukdar 2003). Consumers gather objective information about attributes and prices online, which reduces purchase risk; the physical channel thereafter provides them with reassurance (Flavián, Gurrea, and Orús 2016;Singh, Ratchford, and Prasad 2014). Cross-channel shopping may also help price-oriented consumers find better deals (Gensler, Neslin, and Verhoef 2017) and increase the experiential value of shopping for hedonic-oriented consumers (Balasubramanian, Raghunathan, and Mahajan 2005).
The economic approach has been applied to attribute-based decision-making research (Rapp et al. 2015;Verhoef, Neslin, and Vroomen 2007). Consumers' perceptions of channel characteristics at different stages of the purchase process determine cross-channel shopping (e.g., Burke 2002;Frambach, Roest, and Krishnan 2007;Ratchford, Lee, and Talukdar 2003). In this way, Verhoef, Neslin, and Vroomen (2007) classify channel attributes in terms of benefits and costs and compare the Internet, catalogs and physical stores. They find that the Internet is the preferred search channel because it provides the highest convenience and facilitates comparisons. However, the physical store is preferred for purchasing due to its enhanced service quality and low purchase risks. These results have been consistently confirmed in the literature (Chiu et al. 2011;Frambach, Roest, and Krishnan 2007;Gensler, Verhoef, and Böhm 2012;Konuş, Verhoef, and Neslin 2008;Noble, Griffith, and Weinberger 2005).
Less is known about the consequences of cross-channel shopping for customer experience management (Lemon and Verhoef 2016;Paccard 2017). Previous studies consider the influence of channel synergies on consumer behavior. Using multiple channels during the shopping process produces complementarities (Gensler, Verhoef, and Böhm 2012;Singh, Ratchford, and Prasad 2014) and is positively related to consumption (Lee and Kim 2008), attitudes toward retailers (Pantano and Viassone 2015), satisfaction (Herhausen et al. 2015) and loyalty (Piercy 2012;Sopadjieva, Dholakia, and Benjamin 2017). However, previous studies consider crosschannel behavior in general, neglecting the impact of specific channel combinations.
We analyze this research question. Our argument is clear: webrooming is the most frequent cross-channel shopping behavior because it provides consumers with more satisfactory search experiences. We focus on search process satisfaction as a particular outcome of the shopping experience (Burke 2002;Lemon and Verhoef 2016). Fig. 1 shows the research model and the proposed hypotheses. Specifically, we expect that carrying out webrooming behaviors, compared to showrooming behaviors, increases consumer satisfaction with the search experience (hypothesis 1). Furthermore, we expect that confidence (hypotheses 2 and 3) and smart shopping feelings (hypotheses 4 and 5) explain this main effect. The possible influence of contextual factors, in terms of shopping motivations and product characteristics, is also controlled.

Search Process Satisfaction
Search process satisfaction is defined as the satisfaction with the actual information search process (Creyer and Kozup 2003). Consumers experience satisfaction not only with the chosen product but also with the shopping experience (Puccinelli et al. 2009). Previous research notes that satisfaction is a key outcome of cross-channel shopping (Balasubramanian, Raghunathan, and Mahajan 2005;Burke 2002;Piercy 2012;Rapp et al. 2015;Wallace, Giese, and Johnson 2004). Satisfaction with the decision-making process leads to consumption satisfaction and positively influences post-choice behavior (Heitmann, Lehmann, and Herrmann 2007).
The previous literature addresses the overall impact of channel integration on satisfaction (Chatterjee 2010a;Herhausen et al. 2015;Wallace, Giese, and Johnson 2004). For example, Pantano and Viassone (2015) find that using multiple channels increases perceived service quality, leading to satisfaction, in a multichannel retail environment. Other researchers show that multichannel information searches provide enhanced satisfaction with the experience in comparison to single-channel searches (Flavián, Gurrea, and Orús 2016;Lee and Kim 2008;Piercy 2012). However, until recently, no study had analyzed the impact of different channel combinations on the consumer's search process satisfaction. In a recent exception, Jang, Prasad, and Ratchford (2017), examine price satisfaction as the final search outcome in car purchases, and find no differences based on the number of channels used to search for information.
We propose that webrooming has a positive influence on search process satisfaction. Following cognitive fit theory (Vessey 1991), webroomers take advantage of each channel during their search experience to satisfy their purchase needs (Konuş, Verhoef, and Neslin 2008). Compared to showrooming, webrooming offers the reassurance of a physical interaction with the product (Smith and Swinyard 1983;Wright and Lynch 1995), immediate possession (Chatterjee 2010a(Chatterjee , 2010b and sensory engaging experiences (Kacen 2003). Consumers can derive more satisfaction through the use of self-service technologies during search activities than when purchasing (Meuter et al. 2000). Therefore: H1. Webrooming (versus showrooming) has a positive impact on search process satisfaction.

The Role of Confidence in the Webrooming-Satisfaction Relationship
Confidence is a mental state of certainty when for evaluating a product, brand or purchase situation (Petty, Briñol, and Tormala 2002). Consumers not only form a favorable impression of the product but also feel right about it (Heitmann, Lehmann, and Herrmann 2007;Marks and Kamins 1988). In a pre-decision phase, the concept of the strength of the emerging preference is referred to as "confidence in the leading option" and reflects "the confidence that the emerging leader in the choice process will ultimately be one's final decision, even if confronted with additional information" (Chaxel 2016;p. 180). Confidence thus relates to traditional marketing concepts such as attitude strength (Smith and Swinyard 1983) and feelings of certainty in evaluations (Fitzsimons et al. 2002).
Cross-channel consumers are assumed to have a certain degree of involvement with the product and/or the purchase (e.g., Burke 2002;Konuş, Verhoef, and Neslin 2008). They are motivated to choose the best shopping option, which increases the uncertainty associated with the purchase (Piercy 2012;Puccinelli et al. 2009). Cross-channel consumers strive to reduce uncertainty and feel confident that the product is the best match to their needs (Flavián, Gurrea, and Orús 2016). When consumers combine channels, they create individuated information that enhances their perception of being in control and their belief that they are making the right choice (Schul and Mayo 2003).
We propose that webrooming is the best channel combination to evoke feelings of confidence. Webrooming search enhances the consumer's knowledge of and preferences for the product (Keng, Liao, and Yang 2012;McCabe and Nowlis 2003), reduces information asymmetries and enhances control over the purchasing process (Burke 2002;Heitz-Spahn 2013). For consumers who expend effort to find the best alternative, the Internet is superior to other information channels (Alba et al. 1997;Kulkarni, Ratchford, and Kannan 2012;Ratchford, Lee, and Talukdar 2003). Showroomers might seek low prices or convenience (Rapp et al. 2015), in which case confidence may be less of an issue, which it certainly is for webroomers. Although consumers may showroom to make the best purchase decision (Gensler, Neslin, and Verhoef 2017), they have less control over their purchases than webroomers, given that online purchases may result in delayed delivery or a product that does not meet with expectations. Thus: H2. Webrooming (versus showrooming) has a positive impact on confidence in making the right purchase.
We also note the positive relationship between confidence and search process satisfaction (e.g., Balasubramanian, Raghunathan, and Mahajan 2005;Heitmann, Lehmann, and Herrmann 2007;Lee and Kim 2008). The more confident consumers are about their judgments, the more favorably they evaluate the search experience. If webrooming provides consumers with confidence, their search process satisfaction will be enhanced. Formally: H3. Confidence mediates the impact of webrooming (versus showrooming) on search process satisfaction.

The Role of Smart Shopping Feelings in the Webrooming-Satisfaction Relationship
Consumers feel smart because they have invested time and effort in searching and using promotion-related information to achieve price savings (Mano and Elliott 1997). From a utility perspective, smart shopping feelings arise when the consumer pays a lower price than the internal reference price and feels pleasure, pride or "like a winner" (Burton et al. 1998). Recently, smart shopping feelings are shown to be related to outcomes other than paying a low price. Specifically, Atkins and Kim (2012) develop a three-dimensional structure of shopping benefits that leads to smart shopping feelings. They show that consumers feel smart not only because they achieve monetary savings, but also because they achieve time and/or effort savings (convenience), or because they perceive they are making the right purchase.
Smart shopping feelings are likely to occur in cross-channel shopping settings. Using multiple channels helps to affirm personal traits, such as thrift or expertise (Balasubramanian, Raghunathan, and Mahajan 2005;Chiu et al. 2011). Crosschannel consumers may feel smart because they believe that "searching on one channel allows them to make better purchase decisions on another channel due to their own 'smart' search behavior" (Verhoef, Neslin, and Vroomen 2007;p. 132). We propose that webrooming can lead to smart shopping feelings and thus influence the consumer's search process satisfaction. Pauwels et al. (2011) find that informational websites may help consumers to make smarter purchases. The information availability, transparency and convenience of the Internet reduce information asymmetries and empower consumers in their relationship with vendors (Walsh and Mitchell 2010).
Smart shopping feelings arise when consumers perceive themselves as being responsible for the purchase outcome and having control of the causes that generate the outcome (Schindler 1998). Therefore, webroomers, in comparison to showroomers, may feel more in control and responsible for purchase outcomes. Although showroomers may feel smart when they find low prices or save time in the purchase process (Gensler, Neslin, and Verhoef 2017), they are not ultimately responsible for, nor do they have control of, the final outcome of the purchase. Thus: H4. Webrooming (versus showrooming) has a positive impact on smart shopping feelings.
Previous studies establish a positive link between smart shopping feelings and satisfaction (Darke and Dahl 2003;Schindler 1998). Consumers who feel responsible for purchase outcomes have smart shopping feelings and derive positive consequences from the experience, such as purchase satisfaction (Mano and Elliott 1997;Schindler 1998). We propose that smart shopping feelings may account for the effect of webrooming on search process satisfaction: H5. Smart shopping feelings mediate the impact of webrooming (versus showrooming) on search process satisfaction.

Contextual Factors
Channel integration benefits may depend on situational factors (Dholakia et al. 2010). We analyze webrooming as a goal-directed cross-channel behavior where consumers invest time and effort to make the best possible purchase. However, we must take into account that cross-channel consumers may also be driven by other goal-directed motivations (Kang 2018; Noble, Griffith, and Weinberger 2005; Schröder and Zaharia 2008). Our research proposal focuses on the maximization of the output of the experience (best possible purchase), yet consumers may also be motivated to minimize inputs (investment of time, effort and money). Balasubramanian, Raghunathan, and Mahajan (2005) argue that cross-channel consumers are driven by their desire for convenience, efficient information and the best value. Previous studies show that cross-channel consumers seek efficiency in their purchases and try to save time and effort in making the decision (Gensler, Neslin, and Verhoef 2017). Furthermore, consumers use information and purchase channels to buy products at reduced prices (Balasubramanian, Raghunathan, and Mahajan 2005) and the perception that the Internet allows consumers to search and find low prices intensifies this motivation (Schoenbachler and Gordon 2002). Finding low prices is a key reason for online shopping (e.g., Schröder and Zaharia 2008) and showrooming (Kang 2018). For webroomers, previous research shows that they may be motivated to find better deals at stores (Verhoef, Neslin, and Vroomen 2007) but also that price may not be an important variable (Kang 2018). Our exploration of webrooming behavior takes into account these different instrumental motives for cross-channel shopping.
In a similar vein, we note the possible influence of product type on cross-channel behavior (Burke 2002). One key determinant of cross-channel behavior is the consumer's capacity to assess product quality before interacting physically with the product. Consumers may prefer to purchase experience or touch products (e.g., Flavián, Gurrea, and Orús 2016; Heitz-Spahn 2013) in a physical store more so than is the case for search and non-touch products. Furthermore, products differ in the amount of time, money and energy that the consumer spends in the purchase process (Murphy and Enis 1986). This product dimension has also been identified as a determinant of cross-channel behavior (Burke 2002) and smart shopping feelings (Atkins and Kim 2012). The purchase of products that require a high investment of resources (high cost products) are carried out through cross-channel processes to a greater extent than products with low resource investment demand (low cost). Consumers perceive higher risk in the purchase of high cost products than in the purchase of low cost products (Burke 2002). Therefore, the consumer's need to complete the shopping experience in a physical store may be higher for high cost products. Thus, we consider the influence of these product dimensions (search/experience attributes, high/low cost

ONLINE SURVEY
Shopping Episodes Sample: 421 customers from a multichannel retailer database invested in the purchase) in the analysis of webrooming behavior.

Method
The research uses a mixed methodology, combining experimental procedures with survey-based data to explore webrooming in depth and test the hypotheses (see Fig. 2). First, an online survey explores the channel combinations most frequently associated with satisfaction or dissatisfaction outcomes. This requires participants to recall and describe a shopping experience that had different satisfaction levels. Two follow-up analyses offer a comprehensive picture of the webrooming phenomenon and a tentative test of the hypotheses.
Taking into account the results of the first exploratory study, we carried out a lab experiment (ensuring high environmental control) to analyze the impact of webrooming and showrooming behaviors on confidence and satisfaction (hypotheses 1 to 3). Finally, Study 3 consists of an experimental design, involving real consumers, who were asked to evaluate webrooming versus showrooming behaviors in terms of confidence, smart shopping feelings and satisfaction (hypotheses 1 to 5). In all the studies we consider different shopping motivations and/or product characteristics to control for the effects of contextual factors. The results of the analyses involving these contextual factors are reported in the Appendix.
The studies were conducted with samples made up of millennials, since their growth as consumers is parallel to that of the Internet as an information and purchase channel. Millennials are a generational cohort born roughly between 1980 and 2000 (Deloitte 2016) and comprise the largest consumer segment in the US and Western Europe (Pew Research Center 2010). Millennials are characterized by their ubiquitous use of technology. Thus, they represent the most suitable population to test our hypotheses.

Study 1 -Exploratory Research
The exploratory research was carried out in collaboration with a national multichannel retailer. We prescreened participants to ensure they were part of the millennial generation. A total of 421 customers took part in the study (53.9% women, 96.9% had more than 5 years' experience of the Internet and 93% had purchased online in the previous 12 months). Participants had to recall and describe a recent purchase episode (Heitz-Spahn 2013; Meuter et al. 2000). We manipulated the types of experience to obtain different levels of satisfaction. We asked participants to recall a recent purchase experience that made them feel especially satisfied, dissatisfied or just a purchase experience (control condition). They had to describe the product they purchased and give all the information sources used. They were also asked to give the reasons for their (dis)satisfaction.

Analysis and Results
The shopping episodes were classified into six categories: two single-channel episodes, namely, purely online and purely physical; four cross-channel shopping episodes, namely, webrooming, showrooming, printed media (catalogs, magazines, newspapers) and the physical store; and at least two channels (e.g., online, printed and physical channels through which to search for information and then purchase at the physical store). The content analysis also allowed us to refine the data and check the satisfaction manipulation. The final sample was 368 participants (55% female; mean age = 25.47; 96.7% with more than 5 years' experience of Internet use).
Cross-channel shopping dominated single-channel shopping in all conditions (Table 1). For the cross-channel episodes, virtual and physical channels in two-stage processes (i.e., webrooming, showrooming) were more frequently reported than any other channel combination or sequence. Also, we created a dummy variable (1 = webrooming, 0 = otherwise) and submitted it to a chi-squared analysis crossed by the experimental groups (χ 2 (2) = 46.654, p b .001). Webrooming episodes were higher in the satisfaction group than in the other conditions, providing initial support for H1.
We developed two follow-up analyses. First, we developed a qualitative analysis to better understand cross-channel behaviors in terms of motivations and outcomes. Second, we collected additional data to investigate the associations between channel combinations, products and satisfaction. Both followup analyses shed light on the influence of confidence and smart shopping feelings on satisfaction, as well as on the influence of the contextual factors.

Follow-up Analysis 1
We arranged two group interview sessions with 16 participants from the first study. Eight individuals took part in each session. The samples consisted of graduate and postgraduate students and employees with three to five years of work experience. Three experts in marketing, sociology and new technologies conducted the content analysis using ATLAS.ti v6.2 software.
The discussions focused on webrooming and showrooming. The motivations behind the behaviors had common characteristics. First, both webroomers and showroomers seek to reduce the uncertainty related to the purchase situation (e.g., lack of knowledge about the product) or the channel (e.g., lack of physical interaction online, distrust in the offline salesperson's recommendations). Second, they carry out these behaviors to save time and/or effort in the purchase. Showrooming involves using the physical channel to search for product information with convenience (e.g., the sizes in which a particular brand are available) and to take advantage of the assistance that might be offered by sales staff. The convenience motivation was also associated with webrooming. Consumers said that they employed the online channel because it conveniently accelerated the information-gathering process and allowed them to check product availability at the physical store. However, differences were also identified. Paying a low price is a main motivation for showrooming. For webroomers, price does not appear to be such an important factor and the effort exerted during the process is not perceived as a cost. In fact, several participants declared they had paid more than they had planned for the product, but that the surcharge was not a problem, because "the effort is well worthwhile" (participant 4, session 1).
Following this idea, webrooming is carried out to make the best possible purchase. The high level of purchase importance motivates the consumer to exhaustively search the Internet to find the product that may be the best match to his/her needs. Yet this still does not dispel all the uncertainty. The consumer still has to visit the store to verify that the researched information is accurate and confirm that the product is the best choice. Although the participants care about the experience and take an active role (both in webrooming and showrooming), webrooming allows them to control the process and purchase outcomes to a greater extent than showrooming ("if you end up buying online, you do not know when the product will arrive or if it is really the one that you bought; this does not happen at the physical store"; participant 2, session 2).
In sum, convenience is shown to be an important determinant factor for cross-channel shopping, both in webrooming and showrooming. However, webrooming and showrooming appear to have different motivations: the wish to make the right purchase leads the consumer to webroom, whereas the wish to save money drives consumers to showroom. In the next follow-up analysis, we explore the consumers' reasons for their satisfaction with webrooming and showrooming.

Follow-up Analysis 2
Following the same procedure as in the main study, we obtained 140 additional, valid respondents in the satisfactory condition. The final database was 264 respondents. 1 Again, webrooming was the recollected behavior most associated with feelings of satisfaction (n = 116; 43.9% of all the episodes, 60.4% within cross-channel behaviors). Showrooming behaviors were reported by only 30 participants (11.4% of the total sample; 15.6% within cross-channel shopping). The appendix displays crosstabulated information about the types of shopping episodes and product categories. Table 2 displays the reasons for satisfaction depending on the type of purchase episode. The reasons are coded in terms of smart shopping perceptions (Atkins and Kim 2012;Chandon, Wansink, and Laurent 2000): time/effort savings or convenience, feelings of having made the right purchase and money savings. However, the codification process identifies other types of reason, also related to the smart shopper construct: effort appreciation (e.g., "I had been searching for this product for a long time"), product uniqueness (e.g., "The product was customized") and the hedonic component of the experience (e.g., "I really enjoyed the experience at the store") (Darke and Dahl 2003;Mano and Elliott 1997). Each statement could be assigned to several codes.
Most webroomers indicate that finding the best product/good value for money is the main reason for their satisfaction (Table 3), whereas they barely mention convenience and price (less than 20%). Webroomers also derive satisfaction from the effort exerted during the process. For showroomers, price is the most important reason for satisfaction; more than three quarters of them highlight the lower prices found online. However, showroomers cite reasons related to having made the right purchase less than the other cross-channel shoppers. Therefore, webroomers and showroomers seem to purchase similar types of products (see appendix); however, the reasons that lead to satisfaction, which may reflect different shopping motivations, differ.
Altogether, the results of this exploratory research reveal that webrooming is more frequently undertaken and satisfactory than other single-channel and cross-channel behaviors. Webrooming purchases are important for consumers, who care about making the best possible purchase. Webroomers need to be sure or be convinced about the suitability of an alternative, and lack of confidence may be a key reason why they visit the physical store. The following studies examine whether confidence in making the best purchase can be the driver of consumer satisfaction with webrooming.

Study 2
In Study 2 we compare webrooming to showrooming and analyze the differences between a product under consideration in a cross-channel interaction (target) and a product with no previous interaction (rival). If webrooming instills confidence in the product initially considered as the best shopping option, it will be interesting to examine how stable this preference is when confronted with new, competing information (Chaxel 2016). Also, we examine take into account the possible influence of product type by analyzing two categories, clothing and electronics, which differ in their search-experience properties. These product categories are the most cited by the Study 1 participants. They are frequently purchased through cross-channel shopping (Google Consumer Barometer 2015). Previous research categorizes electronics as search goods, and clothing and accessories as experience goods (e.g., Van Baal and Dach 2005). Moreover, the results of a pre-test confirm the manipulation of the type of product and helped in the selection of appropriate stimuli for the study (appendix).
The study context was a simulated purchase in the university store. The design of the experiment was full factorial, with 2 (sequence: webrooming vs. showrooming) × 2 (type of product: college strap bag vs. usb memory stick) between-subjects conditions. A sample of 262 undergraduate students (50% female; mean age = 20.3) was randomly assigned to one of the four conditions (cell sizes ranged from 63 to 69). Previous studies of multichannel consumer behavior use students, who represent a valid sample population (e.g. Keng, Liao, and Yang 2012). We prescreened all participants to ensure that they had had Internet purchase experience.
The experiment was in two parts. In the first (T1), the participants had an initial interaction with a product that they were told had to be considered for purchase. Participants were randomly assigned to either the online or the physical interaction.
After interacting with the product information, the participants responded to 7-point Likert scales regarding confidence (α = 0.88; 74.18% explained variance; Petty, Briñol, and Tormala 2002) and search process satisfaction (α = 0.82; 65.40% explained variance; Heitmann, Lehmann, and Herrmann 2007). In the second part (T2), the participants had either a physical or an online interaction with the same product (target) and a new alternative (rival). The rival product had characteristics similar to those of the target. After a few minutes, they completed the second part of the questionnaire on their confidence levels in both products (target: α = 0.93; 81.91% explained variance; rival: α = 0.85; 70.15% explained variance), and search process satisfaction (α = 0.90; 76.55% explained variance).

Results and Discussion
The correlations between the dependent variables were positive and significant (rs N 0.274) and the results of the . Participants in the webrooming condition reported higher levels of search process satisfaction and confidence in the target product than participants in the showrooming condition (Table 3), supporting H1 and H2. Confidence in the rival product was also significantly higher in webrooming than in showrooming (Table 3). This result is logical and consistent with previous literature regarding the role of direct (vs. indirect) experiences in reinforcing consumers' confidence (McCabe and Nowlis 2003). Regarding the effect of product type, the MANOVA yielded a significant multivariate effect (Wilk's λ = 0.928; F (3,256) = 6.646, p b .001). At the univariate level, we find that (Table 5). No interaction effects were found (ps N 0.203). The direct effects of webrooming on confidence and satisfaction did not depend on the product type. Next, we used the macro PROCESS v3 to examine the mediating effects of confidence. Confidence in the target and the rival products acted as parallel mediators and the product type was included as a covariate. After controlling for product type, 2 confidence in the target mediated the effect of webrooming on search process satisfaction (Table 4); yet, the type of sequence still had a direct influence, revealing partial mediation (H3).
Overall, the analysis confirms that webrooming (vs. showrooming) participants are more confident in the adequacy of the product and satisfied with the search process. In fact, confidence drives the customer experience. However, webrooming (vs. showrooming) still has a direct effect on satisfaction. This complementary mediation suggests that there may be other potential mediator(s) omitted from the model. Furthermore, the relationships between webrooming, confidence and search process satisfaction have been proven in a controlled setting and with a limited number of products. To overcome these limitations, Study 3 used real consumers to test the hypotheses and to analyze smart shopping feelings as a complementary mediator of the effects of webrooming (vs. showrooming) on satisfaction. Furthermore, the study analyzes a wider set of products and shopping motivations. By controlling for these contextual factors, we aim to test the robustness of the relationships across different shopping situations.

Study 3
In Study 3 we examine the influence of webrooming (showrooming) on consumer confidence and smart shopping feelings, and whether these variables together explain why webrooming produces more search process satisfaction than showrooming. Feeling smart is a direct consequence of the transfer of power that the proliferation of marketing channels causes in customer-firm relationships (Verhoef, Kannan, and Inman 2015), and determines customer satisfaction and customer experience (Lemon and Verhoef 2016;Mano and Elliott 1997). As previously stated, smart shopping feelings include benefits related to savings in money, time and/or effort and to the perception that one has made the right purchase (Atkins and Kim2012). These shopping benefits are intertwined with the goal-directed motivations of cross-channel consumers (Balasubramanian, Raghunathan, and Mahajan 2005;Kang 2018;Noble, Griffith, and Weinberger 2005;Schröder and Zaharia 2008). By manipulating shopping motivation types, we control for this contextual factor in the relationships between webrooming, confidence, smart shopping feelings and search process satisfaction.
Specialist retailers were contacted through their trade associations and asked to collaborate in the study. The study pool was formed by 21 multichannel specialist retailers, trading in 6 different product categories (clothing, electronics and home appliances, leisure, sports, home furniture and gourmet food). We obtained a final valid sample of 468 real customers (65% female; mean age = 26.3; 43.2% with university degrees; 99.1% had more than 5 years of Internet experience).
We developed cross-channel shopping situations with a varied set of motivations that could lead to smart shopping feelings (Schindler 1998). Specifically, we developed an experimental design with 2 (webrooming vs. showrooming) × 3 (right purchase vs. time and effort savings vs. money savings) in a between-subjects factorial design. We adapted each condition to the product type sold by each participating retailer; thus, the participants were introduced to a realistic situation. We generated a total of 21 × 2 × 3 = 126 scenarios.
The participants were randomly assigned to one of the six conditions. They read an account of a shopping experience that started with an online search (a visit to a physical store) and ended with a purchase at a physical (online) store. The accounts were manipulated in terms of different shopping motivations, in accordance with the smart shopping literature (Atkins and Kim 2012;Mano and Elliott 1997). We then asked the participants give their opinion as to judge how the consumer would react to the shopping scenario presented. To check the manipulation of the shopping motivation, the participants were asked to answer the three-dimensional smart shopping perceptions scale developed by Atkins and Kim (2012). They also reported smart shopping feelings: (1) Alex feels happy about this purchase; (2) Alex feels it was a smart purchase; (3) Alex feels pride about this purchase (Chandon, Wansink, and Laurent 2000;Mano and Elliott 1997). Confidence and search process satisfaction were measured in the same way as in Study 2.

Scales Validation
The scales were validated in two steps through an Exploratory Factor Analysis (Hair et al. 1998) and a Confirmatory Factor Analysis (Fornell and Larcker 1981). First, we carried out an analysis of reliability and dimensionality (Churchill 1979; Anderson and Gerbing 1988;Hair et al. 1998). Second, we conducted a Confirmatory Factor Analysis (Fornell and Larcker 1981) using Structural Equation Modeling (SEM) with EQS. 6.3 software. The initial factor structure revealed that all the item loadings scored above the recommended benchmark of 0.7 (Henseler et al. 2009), with the exception of one item of the perceived time and effort savings scale. This item was removed. The composite reliabilities were above 0.65 (Jöreskog and Sörbom 1993), supporting the internal consistency of the scales. The Average Variance Extracted (AVE) was higher than 0.5 (Fornell and Larcker 1981), assuring convergent validity. Finally, discriminant validity was supported, since the square root of the AVE was higher than the shared variance among the constructs (Fornell and Larcker 1981). Following the validation of the measurement instruments, we calculated the average values of the items to check the manipulations and test the hypotheses.

Manipulation Check of the Type of Shopping Motivation
Type of shopping motivation has a significant effect on perceptions of time/effort savings (F (2,462) = 31.701, p b .001). The post-hoc Tukey test shows that these perceptions are higher when the consumers seek to reduce the time and effort they expend in shopping (M = 5.24, SD = 1.49) than when they are motivated to make the right purchase (M = 3.91, SD = 1.39; p b .001) and to save money (M = 4.56, SD = 1.48; p b .001). In addition, perceptions of money savings are significantly affected by the shopping motivation (F (2,462) = 90.437, p b .001). These perceptions are higher for participants motivated to save money (M = 5.69, SD = 1.25) than for those motivated to save time/effort (M = 3.87, SD = 1.41; p b .001) and for those with the right purchase motivation (M = 4.12, SD = 1.23; p b .001). However, the best purchase motivation does not directly produce higher perceptions of having made the right purchase than the other motivations (p = .972). Further analysis indicates that these perceptions depend on the type of shopping experience (F (1,462) = 3.794, p b .05). The perception that the right purchase has been made is higher for the right purchase motivation group than for the other motivation groups, but only in the webrooming scenarios (see the appendix for a detailed description of the results of the interaction). With this exception, the manipulation of the type of shopping motivation was successful. Table 5 shows descriptive data for each treatment and the results of the univariate ANOVAs. In support of H1, search process satisfaction is significantly higher in webrooming than in showrooming. Hypotheses H2 and H4 are also supported by the data: confidence and smart shopping feelings are significantly higher among the webrooming than the showrooming participants (Table 5). No other main or interaction effects are significant (ps N 0.071).

Hypotheses Testing
To test hypotheses H3 and H5 a mediation model with parallel mediators was estimated, using the macro PROCESS v3 (Hayes 2017). The model proposes that webrooming has direct effects on confidence, smart shopping feelings and search process satisfaction. In addition, it is posited that confidence and smart shopping feelings mediate the effect of webrooming on satisfaction. Perceptions of time/effort savings, money savings and having made the right purchase are included as covariates. Table 6 displays the results of the mediation model. The model shows an acceptable explanatory power (R 2 = 0.476). The direct effect of webrooming on search process satisfaction disappears when the mediators and covariates are included in Table 5 Descriptive data and analysis of variance -Study 3.

Confidence
Smart shopping feelings Search process satisfaction the model. Confidence and smart shopping feelings significantly influence satisfaction. The results of the bootstrapping for the indirect effects reveal full mediation of both variables. The pairwise comparison between the two indirect effects is not significant, since zero was included in the bootstrap confidence interval (− 0.060, 0.119). Thus, both confidence and smart shopping feelings equally mediate the impact of webrooming on satisfaction. Hypotheses H3 and H5 are supported. In addition, perceptions of time/effort savings do not affect satisfaction, which is consistent with our previous findings. Webroomers value the effort invested in their purchases, and efficiency/convenience are not related to their search process satisfaction. However, perceptions of money savings have a significant positive influence on satisfaction. Overall, the results generally support our research proposal. Webroomers are more satisfied with their search experience than showroomers. Search process satisfaction is explained by confidence in having made the right purchase and smart shopping feelings. The analysis regarding the influence of type of product on the dependent variables and their relationships is detailed in the appendix. The general model is highly consistent across product characteristics.

Summary and Implications
This is one of the first studies to analyze the impact of the use of specific combinations of online and offline channels on the consumer experience (Lemon and Verhoef 2016). Table 7 summarizes the main findings of this research. We contribute to the multichannel marketing literature by showing that webrooming is a more satisfactory cross-channel shopping experience than showrooming. This may be a reason why webrooming is the prevalent form of cross-channel shopping behavior. We consistently find this effect in a series of studies combining quantitative and qualitative techniques.
Our research focuses on the analysis of this effect and explains the different factors that define the consumer's search process satisfaction with webrooming. The results of the exploratory research reveal that webrooming is associated with satisfactory shopping experiences. Webroomers are motivated to make the best possible purchase, and they search for information intensively with the ultimate goal of having confidence in their decisions. Webrooming is the best channel combination for accomplishing this shopping goal, which has a strong influence on consumer satisfaction. The results of three conclusive studies confirm our propositions (Table 7). Confidence is higher following a webrooming experience than following other cross-channel experiences. Confidence, then, explains why webrooming leads to search process satisfaction.
Moreover, webrooming generates smart shopping feelings, which also determine satisfaction. Webroomers weigh the benefits and limitations of each channel to make smart choices. The online information search reduces information asymmetries and empowers consumers during the purchase journey. The physical store provides reassurance to the consumer and immediate possession of the product. Thus, the consumers perceive themselves as being responsible for the purchase outcome and as being in control of the causes that generate it, which, in turn, causes smart shopping feelings. The results of the exploratory study support this notion. 3 Thus, our findings may help researchers to explain the underlying mechanisms of the effects of webrooming on search process satisfaction. Webrooming increases the consumers' confidence that they have made in making the right purchase and their smart shopping feelings; both mental states equally determine satisfaction.
Our findings are fairly consistent across shopping motivations and product categories, but some differences must be noted. First, cross-channel consumers may be motivated to make purchases efficiently. However, saving time and/or effort appears to have no influence on search process satisfaction. This finding is in line with Gensler, Verhoef, and Böhm (2012), who also find that convenience has no significant role in the search and purchase stages of the cross-channel purchase decision-making process. On the contrary, the effort expended may have a positive effect on satisfaction (Table 7). Nevertheless, the relationships between perceived effort, convenience and satisfaction should be investigated in future research.
Second, saving money on the purchase may also be an important motivator of cross-channel shopping and determine satisfaction, which is consistent with previous literature on smart shopping (Darke and Dahl 2003;Schindler 1998). When consumers want to pay a low price for a product, showrooming appears to be the preferred channel combination (Table 7). Third, although product characteristics have significant direct effects on smart shopping feelings and satisfaction (Table 7), the path to search process satisfaction seems not to depend on product type (Burke 2002).

Managerial Implications
Our findings have valuable implications for multichannel marketers. In the omnichannel era (Verhoef, Kannan, and Inman 2015), retailers must learn to integrate channels to offer seamless and unique experiences that retain consumers throughout the entire purchase experience and avoid free-riding behaviors (Chiu et al. 2011;Herhausen et al. 2015). Providing customers with satisfactory experiences may be the key to retaining them and establishing long-lasting relationships with them (Lemon and Verhoef 2016; Wallace, Giese, and Johnson 2004).
Our findings advocate that retailers should shift from design orientation to customer orientation in order to deliver satisfaction with the cross-channel shopping experience. Most practitioners focus on delivering convenience in cross-channel experiences (Parry 2016), which surely helps consumers make efficient purchases. Convenience represents a critical aspect of online shopping and affects satisfaction in online settings (Szymanski and Hise 2000). However, in cross-channel settings we show that providing convenience may not be the best strategy for creating customer satisfaction. In fact, consumers appear to value the time and effort invested in making these purchases. In cross-channel shopping, convenience and efficiency may be hygiene factors, rather than motivators (Herzberg 1966).
Instead, this research stresses the importance of developing strategies to foster consumer confidence during the purchase journey. This can be achieved in a number of ways. First, the search for confidence is most associated with important purchases. Highly involved consumers are driven by the desire to make the best possible choice (Puccinelli et al. 2009). These consumers strive to feel confident about their choices. Thus, cross-channel marketers should try to boost the consumer's involvement or engagement with the product, brand or purchase process situation (e.g., by increasing emotional attachment or appealing to the symbolic meaning of the purchase; Balasubramanian, Raghunathan, and Mahajan 2005) so that he or she will seek to gain confidence in their choices. This will establish the grounds for a more satisfactory experience.
Second, our findings show that cross-channel consumers gain more confidence from a webrooming experience than from a showrooming experience. Retailers should take advantage of the different capabilities offered by the channels to design optimal customer experiences.
In particular, managers should maximize the unique informational power of the Internet to offer the fullest information about their offerings. For example, reading online reviews before making a physical purchase increases consumers' confidence in their choices (Flavián, Gurrea, and Orús 2016). Online channels might be used to promote a visit to the physical store to complete the purchase journey (e.g., by offering special promotions or opportunities for sensory experiences). At the physical store, consumers should be allowed to touch and feel the products; this diagnostic information cannot be obtained elsewhere and increases choice confidence.
Furthermore, smart shopping feelings are likely to arise in cross-channel shopping. Our findings show that webrooming makes consumers feel smarter than does showrooming. Crosschannel consumers may feel responsible for the final outcome of a purchase and need high control of the shopping process. Webrooming is the channel combination that better accommodates these determinants of causal attribution. Thus, companies should follow this trend in their communications strategies by appealing to their customers' intelligence (e.g., by challenging the consumer's mastery in making clever purchases).
Online retailers may also benefit from our findings. Money savings have traditionally been associated with smart shopping • Convenience is a determinant factor for webrooming and showrooming; however, it does not affect satisfaction.

Yes N.A. Yes
• Paying a low price is a main motivation for showrooming and the main reason for showroomers' satisfaction.
• Webroomers also derive satisfaction from the effort exerted during the process. Price does not appear to be an important factor.

Yes N.A. Yes
• Perceptions of money savings have a positive influence on satisfaction.
Yes N.A. Yes • Search process satisfaction and smart shopping feelings are higher for experience products (versus search products), especially when the cost of search and purchase is high.

N.A. Yes Yes
• Contextual factors (shopping motivations, product characteristics) do not affect the relationships between webrooming, confidence, smart shopping feelings, and satisfaction.
feelings (Schindler 1998), and online shopping and showrooming are associated with the search for low prices (Gensler, Neslin, and Verhoef 2017;Rapp et al. 2015). This research shows that paying a low price is the main reason for satisfaction in online purchases and that money saving perceptions lead to search process satisfaction. Furthermore, by increasing the perceived control of the process (e.g., continuous tracking of the order) and the responsibility for the final outcome of the purchase (e.g., reassuring after-sales messages), online retailers may increase the online shoppers' smart shopping feelings and satisfaction. Overall, our results acknowledge the value of information integrity across channels. Offering accurate product information online and good physical interactions in stores may help retailers integrate their channels more efficiently. Most offline retailers have some form of online component that can be integrated into the consumer experience (Frambach, Roest, and Krishnan 2007), in the same way as online retailers are increasingly embracing the offline presence. Facilitating consumers' control of the process and their knowledge acquisition during the purchase journey can have great potential to improve their customer experience in terms of confidence, smart shopping feelings and satisfaction.

Limitations and Further Research
Despite the importance of our findings for marketing research and practice, some limitations need to be acknowledged. First, the experimental nature of the confirmatory analysis should be noted. In studies 2 and 3, consumers were compelled to take part in either a webrooming or showrooming shopping experience. This manipulation was carried out to maintain a balanced number of participants in both scenarios. A more realistic scenario could have been achieved if the participants had self-selected to do webrooming or showrooming; however, the number of webroomers would have greatly surpassed the number of showroomers, as recent reports (PushOn 2018) and previous studies (Gensler, Neslin, and Verhoef 2017) find. Thus, the balance between the conditions required for the experimental design would have been compromised. Nevertheless, future studies should allow participants to choose freely between webrooming and showrooming and analyze their impacts on satisfaction.
Second, we focus only on millennials. The use of a homogeneous group of participants guarantees internal validity, and testing the hypotheses over a wide range of customers ensures a certain degree of external validity. However, it would be interesting to validate these results and proposed relationships with other consumer profiles.
Third, we focus on the linear online-to-offline sequence better to isolate the effects of webrooming. However, future research should also investigate the cross-channel process as a recurrent path (e.g., on-off-on). Mobile technologies allow consumers to use several channels simultaneously at the same stage of the purchase process, turning cross-channel experiences into omnichannel experiences (Verhoef, Kannan, and Inman 2015). Further research should analyze how omnichannel environments affect the generation of confidence, smart shopping feelings and satisfaction.
Fourth, we examine the impact of cross-channel behaviors on the consumer's search process satisfaction. This optimistic view neglects other outcomes that may be worth investigating. Cross-channel behaviors can also lead to search regret, which results from an unsuccessful search that leads to an undesired purchase decision (Reynolds, Folse, and Jones 2006). Search satisfaction and regret might share some common aspects in cross-channel shopping (e.g., effort invested in the search; Puccinelli et al. 2009), but might be caused by different circumstances. For example, the large amount of data that the Internet offers can cause the consumer to suffer information overload, creating confusion and anxiety during the search process (Walsh and Mitchell 2010), which may lead to heightened feelings of regret if consumers do not find what they need (Reynolds, Folse, and Jones 2006). It would be interesting to examine how consumers evaluate the outcome of the purchase journey, depending on whether it results in a satisfactory or dissatisfactory experience.

Funding Details
This work was supported by the Spanish Ministry of Economy and Competitiveness under Grant ECO2016-76768-R; and European Social Fund and the Government of Aragon under Grant S20_17R.