There is no place like home: home satisfaction and customer satisfaction in online grocery retailing

ABSTRACT

This study acknowledges that home is a central and valenced place for most people, and therefore it is assumed that (a) the content in people’s general category for home issues can be activated easily, particularly by various home-related stimuli, and that (b) the affective charge of such content can have a valence-congruent impact on evaluations of home-related stimuli. These assumptions were the basis for two empirical studies carried out in an online grocery shopping context with respect to home delivery. Both studies indicate that home satisfaction had a positive influence on satisfaction with home delivery – and that home satisfaction had an indirect and positive influence of overall satisfaction with the retailer that provides home delivery. These findings, then, emphasize the need to ‘emplace’ the contemporary online shopper: although the shopper faces an abundance of alternatives only a mouse-click away, he or she is still emplaced in an environment when orders are placed and when goods are received. And when this environment is home, it is particularly likely to influence home-related offers.

KEYWORDS:

• Online retailing
• the last mile
• home satisfaction
• customer satisfaction

1. Introduction

As an increasing part of all shopping occurs online, retailers are struggling with last mile deliveries (i.e., the final leg of a supply chain in which products are delivered to customers). Two main options exist today, namely home delivery and delivery to manned or unmanned pickup points (Nguyen, de Leeuw, and Dullaert 2018; Pan et al. 2017). Home delivery appears to be the most preferred option for consumers (Colla and Lapoule 2012; Morganti et al. 2014), and when it is chosen the consumer’s home becomes a part of the supply chain. Researchers and retailers have acknowledged this in several ways; for example, it has been observed that home delivery can easily generate knotty logistical problems, particularly in urban areas (Colla and Lapoule 2012; Morganti et al. 2014; Pan et al. 2017). Moreover, several studies have examined the influence of specific last mile aspects on consumer preferences and choices, such as delivery fees, delivery speed, timeliness of delivery, and the possibility to track orders (Nguyen et al. 2019). Previous studies have also covered motivations for online shopping, such as convenience, saving time and saving money (Maat and Konings 2018).

So far, however, marketing and retailing-related studies have not taken into account that the consumer’s home is more than an end destination for goods ordered online. That is to say, for most of us, home is a particularly significant space within which we experience strong social, psychological, and emotive attachments (Easthope 2004), it has rich and multidimensional meanings (Smith 1994a; Somerville 1992), and it is a context in which we humans perform our routines of existence and in which we realize our desire to persevere in our being (Noble 2004). It has also been argued that home is a unique place where a person’s past, present, and future selves are reflected and come to life (Graham, Gosling, and Travis 2015). The significance and uniqueness of home can be seen as a function of what is actually happening in people’s homes; diary-based studies (e.g., Smith 1994a) reveal that the high number of different activities carried out at home is unlikely to be matched by any other place. The covid-19 pandemic, which forced people to spend more time at home, has indeed underlined the home as a multi-activity place, and many consumers claim that they intend to permanently adopt much of their home-oriented pandemic behaviors (Heisler 2020).

Moreover, for many individuals, home is positively charged, in the sense that it offers a comfortable, friendly, and safe space where one can retreat and relax (Cristoforetti, Gennai, and Rodeschini 2011; Mallett 2004; Smith 1994b; Somerville 1997). Indeed, when people are exposed to the English word ‘home’, it produces a relatively high mean level of positive affect (Bradley and Lang 1999). For other individuals, however, home is not a safe haven; home can be a prison and a place full of obligations, terror, and abuse (Mallett 2004; Moore 2000). It may be noted that home is a place in which a substantial number of violent crimes are committed. In the US, for example, and according to the Bureau of Justice Statistics, no other place generates a higher number of rapes and sexual assaults. When home is not a safe haven, it seems as if home delivery would not be a very attractive offer – particularly given the present state of low timing precision in many home deliveries (Pan et al. 2017), so that the consumer is forced into an unwelcome ‘house arrest’ situation in which even more time must be spent at home. In any event, a main assumption in the present study is that (a) the centrality of home for us humans and (b) the valenced charge of home create a potential for home aspects to influence consumers’ evaluations of commercial offers involving their homes. In more general terms, it is assumed that repeated exposure to everyday environments can influence evaluations of specific offers (Berger and Fitzsimons 2008).

Given the lack of attention to home-related issues from the consumer’s point of view in previous research, and given also contemporary retailers’ intense struggle to find viable last mile solutions, the purpose of the present study is to explore if the e-shopper’s experience of his or her home can affect satisfaction with home delivery solutions. This exploration contributes to the literature on consumer experiences in an online shopping context in which the consumer’s experience of home is not explicitly acknowledged (e.g., Bilgihan, Kandampully, and Zhang 2016; Rose et al., 2016). In other words, the present study is an attempt to ‘emplace’ online retailing with respect to consumers’ homes. Moreover, even though home in viewed as a construct comprising more than a place in the present study, home is indeed situated in a physical environment. In that sense, the present study is an attempt to contribute to the literature stressing the general importance of the physical environment for human behavior (e.g., Fried 1982) and to the retail-related literature in which it is recognized that the place where products are bought and consumed can be influential for consumers’ decisions and evaluations (Kotler 1973–1974; Spies, Hesse, and Loesch 1997). The present study also contributes with empirically based results to the literature concerned with the general importance of home – a literature that is often based more on reasoning than on empirical data (e.g., Easthope 2004; Noble 2004; Porteous 1976; Somerville 1992). To this end, two empirical studies were carried out with customers who regularly receive home delivery of groceries purchased online.

2. Theoretical framework and hypotheses

The assumptions in the present study involve three satisfaction constructs in an online retailing setting in which goods are delivered to customers’ homes. It is assumed that (a) home satisfaction has a positive influence on satisfaction with home delivery from a particular retailer, and that (b) home delivery satisfaction influences overall satisfaction with the retailer from which goods are purchased online. This thus implies that home satisfaction can have an indirect impact on overall satisfaction with one particular retailer.

2.1 Home as a special place and home satisfaction

We humans are strongly place bound (Thin 2016) and therefore it is not surprising that we can develop an affection for specific places (sometimes referred to as topophilia). Of all possible places than may be targets of humans’ affections, however, home seems to be special. It has been described as a highly significant socio-spatial entity invested with meanings, emotions, experiences, and relationships that lie at the heart of human life (Blunt and Varley 2004; Easthope 2004), a setting in which basic psychological processes are regularly played out (Graham, Gosling, and Travis 2015), and a mode of being not usually experienced in other places (Molony, McDonald, and Palmisano‐Mills 2007). Moreover, home is the space in which most individuals spend most time and a major reference point for the structuring of reality (Porteous 1976). Consequently, to be without a home – to be homeless – is to be almost non-existent (Gieryn 2000). Given the centrality of home along those lines, the authors of the present study assume that home issues are more or less constantly on people’s minds. That is to say, content in the category ‘home’ is likely to be easily accessible. It is also assumed that home issues are affectively charged (Blunt and Varley 2004; Easthope 2004; Graham, Gosling, and Travis 2015); indeed, it has been argued that home is an ‘emotional warehouse’ in which many emotions are experienced (Easthope 2004).

In the present study, and given that life can be represented by many discrete domains from which satisfactions are derived, such as family, job, housing, health, and friendships (Day 1987; Michalos 1980; Neal, Sirgy, and Uysal 1999), it is assumed that home is one particular domain, which generates satisfaction. Therefore, in the present study, home satisfaction is used as a variable to capture the overall evaluation of home. Previous attempts to use the notion of home satisfaction do exist; it has been argued, for example, that home provides humans with all the satisfactions that territory provides for animals (Easthope 2004; Porteous 1976). In addition, there have been several previous attempts to capture home-related satisfaction(s) in terms of housing satisfaction (Ginsberg and Churchman 1984) and residential satisfaction (Adriaanse 2007; Fried 1982). In any event, in the present study, home satisfaction is viewed as a domain-specific and experience-based evaluation response of the same type as other satisfaction variables related to one particular domain (e.g., job satisfaction and customer satisfaction).

2.2 Home satisfaction can influence evaluations of commercial offers

A central tenet in the present study is that home satisfaction can influence evaluations of commercial offers that are home-related. One reason for an expected influence of this type has to do with spreading activation. That is to say, exposure to an object in one particular cognitive category can activate content in other categories (Berger and Fitzsimons 2008; Collins and Loftus 1975; Janiszewski and Wyer 2014), particularly when the categories are semantically related (Fazio 2001; Neely 1976). For example, exposure to material about a Mars expedition can activate content in the category ‘Expeditions to Mars’, which in turn can activate content in the category ‘Candy bars’ in such a way that people come to think of (and buy more) Mars candy bars (Berger and Fitzsimons 2008). Given spreading activation of this type, it is expected that a commercial offer with home-related attributes (e.g., it is referred to as homemade, home-brewed, home cinema, or home delivery) can prime the receiver so that it activates content in the receiver’s general mental category for home.

Moreover, it is assumed that category content in general is affectively tagged, in the sense that it has a valenced affective charge (Bower 1981; Spielman, Pratto, and Bargh 1988). Given that the affective tone of the material in one category can carry over (in a valence-congruent way) to the activated content in other categories (Fazio 2001), it is expected that exposure to one object can result in that the affective tone of an activated category can bounce back and influence affect related to the first object. More specifically, then, and with this view, it is expected that a product presented as home-related (e.g., ‘this ice cream is homemade’) can activate content in a general home category and that the satisfaction associated with home can have a valence-congruent impact on affect and evaluations in relation to the product.

A related reason why home satisfaction can have an impact on evaluations of home-related offers has to do with affect infusion. That is to say, the affect evoked by one particular object can inform (again in a valence-congruent way) evaluations of another object (Pham 2004). One reason why we humans would allow such affect-based inferences, despite the fact that they may lead us astray, is that they are speedy (ibid.). In other words, they allow for high processing fluency – something that is valuable for the effort-avoiding human brain. With respect to home-related affect and its ability to affect satisfaction (of other objects and activities than home itself), Judge and Ilies (2004) has shown that mood experienced at home influences subsequent mood at work – which in turn influences job satisfaction. Affect transfer seems to be particularly likely when the relation between objects is place-based (e.g., two objects are in the same place or one object is ‘in’ another object). Presumably, affect transfer of this type is a main reason why several studies have shown that one particular product is evaluated more positively when it appears in a positively charged environment than in a less positively charged environment (e.g., Edwards et al. 2003). Of particular interest here is Cardello, Bell, and Kramer (1996) who found that each of 12 food items consistently generated higher liking scores when they were presented as served at home compared to served at other places (restaurants, airlines, hospitals, etc.).

Given spreading activation and affect transfer mechanisms, and given that home delivery in an online setting can evoke content in a general home category (both semantically and in terms of home being the actual place to which good are delivered), it is expected that home satisfaction can positively influence the overall evaluation of the home delivery option chosen by online customers. The latter type of evaluation is referred to here as home delivery satisfaction, and the following is hypothesized:

H1: Home satisfaction is positively associated with home delivery satisfaction

2.3 Home delivery satisfaction and overall retailer satisfaction

Home delivery is one of several existing solutions to the last mile issue in online retailing, and such solutions represent the end point in a process with several activities. In general, people’s overall evaluations of events that occur in terms of a process are well predicted by what happens in the end of the process (Fredrickson 2000). One reason is that the end is particularly easy to remember (Chase and Dasu 2001). Another reason, when the end represents the outcome of goal-oriented behavior (such as purchasing goods online), is that outcomes are highly relevant for evaluations (Fredrickson 2000). Empirical results underlining the importance of the ending for overall evaluations in consumer settings are presented by, for example, Baumgartner, Sujan, and Padgett (1997).

Therefore, it is expected that home delivery satisfaction has a positive impact on the customer’s overall evaluation of the retailer from which the goods are purchased. Here, in the present study, such evaluations are referred to as overall customer satisfaction with the retailer. If such a positive impact does exist, it would be consonant with statements in the literature suggesting that the last mile delivery is a crucial aspect in consumers’ purchase process (Nguyen et al. 2019) as well as with empirical studies showing that customers’ perceptions of the performance of delivery aspects in an e-retailing setting are positively associated with overall retailer satisfaction (e.g., Dholakia and Zhao 2010; Koufteros et al. 2014). The following, then, is hypothesized:

H2: Home delivery satisfaction is positively associated with overall customer satisfaction with the retailer

2.4 Mediation and moderation

The reasoning involved in H1and H2 implies that home satisfaction can have an indirect influence on overall customer satisfaction with the retailer from which goods are purchased. The possibility of carry-over effects from satisfaction derived from one facet of life to satisfaction with other facets has indeed been recognized by several authors (e.g., Judge and Ilies 2004; Wright, Bennett, and Dun 1999). To assess this possibility explicitly with respect to home satisfaction, the following is hypothesized in an online retailing context:

H3: Home satisfaction has a positive and indirect influence on overall customer satisfaction with the retailer, and this influence is mediated by home delivery satisfaction

Moreover, in general, social aspects represent a key dimension of the context that can moderate associations between variables (Johns 2006). Social aspects are particularly important in making home a central construct for most people (Easthope 2004; Blunt and Varley 2004; Sixsmith 1986), but it must be acknowledged that the social nature of home can be subject to variation between individuals. For example, a main trend in many countries is an increasing number of single-person households (DePaulo 2014), indicating that home can be a relatively unsocial place for many individuals.

In the present study, it is assumed that the more socially important home is, the more likely it is that content in a general home category will be easily accessible. And accessibility makes it more likely that home-related aspects can influence evaluations of commercial offers that are home-related. Therefore, and in an online retailing context, it is expected that home satisfaction will have a relatively stronger impact on home delivery satisfaction for those customers for which the social importance of home is high as opposed to low. That is to say, it is assumed that the home satisfaction–home delivery satisfaction association (i.e., H1) is moderated by the social importance of home:

H4: The positive influence of home satisfaction on home delivery satisfaction is stronger when the social importance of home is high compared to low

Two empirical studies were used to assess the hypotheses; both studies comprised customers who regularly purchase groceries online on a home delivery basis.

3. Study 1

Study 1 was conducted to test H1-H4 with questionnaire data collected from participants who purchase groceries online and whose dominant delivery mode is home delivery.

3.1 Data collection and participants

The variables in the hypotheses were measured with items in a questionnaire that was distributed online to members of the Prolific panel. Two hundred and fifty panel members who were UK residents, and who were experienced in purchasing groceries online, were invited to the study. From this group, the authors of the present study selected the participants for which home delivery is the dominant mode of receiving groceries purchased online (n = 230), and after deleting 4 participants who failed one or two of the included attention check items, this resulted in 226 participants to be used for the hypothesis testing (M_age = 39.68; 172 women and 54 men). Hence, most of the invited online customers preferred home delivery as a last mile solution, which reflects that home delivery is by far the most common arrangement for groceries ordered online in the UK (Allen et al., 2017). It may also be noted that the online share of grocery shopping in the UK is higher than in most other countries (Munson, Tiropanis, and Lowe 2017). Moreover, the participants were asked from which place they typically order groceries online, and 219 of them stated that they do so from their homes (and 7 participants stated that they place their orders from their workplace). Thus, for the majority of the participants, ‘home’ was not only the end of the customer journey; it was also the point where it begins.

3.2 Measures

Home satisfaction was measured with four items designed for the present study, namely ‘Being at home satisfies me’, ‘My home is a source of satisfaction’, ‘To be at home is a pleasant experience’, and ‘I would like to spend more time at home’. They were scored on a scale ranging from 1 (strongly disagree) to 7 (strongly agree). Somewhat similar items appear in the home attachment scale used by Billig (2006). Cronbach’s alpha for the scale was .90, and the unweighted average of the responses to the four items was used as a measure of home satisfaction.

To assess the validity of the home satisfaction variable, the participants were asked to ‘write what home is for you and how you feel about it’ (an open-ended space was provided to capture the responses). The participants were instructed that at least 100 words were expected about this topic. In the next step, the authors of the present study used the Linguistic Inquiry and Word Count (LIWC) software to compute the emotional tone of each text. The emotional tone score produced by LIWC takes on values in the range from 0 (very negative) to 100 (very positive), and it has been used as an emotion indicator in previous research, for example, by Ludwig et al. (2013) and Cavanaugh, Bettman, and Luce (2015). In the present study, the response range was 1.23–99.00 (M = 86.12, SD = 24.59). The relatively high mean level in the sample is in tune with previous results stemming from exposing people to the English word ‘home’ (Bradley and Lang 1999) and with authors stressing that home issues are affectively charged (Blunt and Varley 2004; Easthope 2004; Graham, Gosling, and Travis 2015). However, in relation to what was said in the introduction (i.e., home can be both positively and negatively charged), the high sample mean for emotional tone reflects a high positive charge. Then, the association between the home satisfaction variable and the emotional tone of the participants’ texts about home was examined. These two variables were positively and significantly associated (r = .28, p < .01), which suggests that the home satisfaction variable was able to capture the valence of the participants’ home-related thoughts.

To measure home delivery satisfaction, the participants were first asked to (1) name the specific firm that is their main online grocery retailer and (2) indicate the most frequent way in which they receive the groceries after purchasing from this firm (home delivery or store pick-up). Tesco, Asda and Morrisons were the most frequently mentioned online retailers. Then, the questionnaire was programmed so that the participants’ answers to (1) and (2) appeared in the wording of the measurement items for home delivery satisfaction. Only a few participants indicated that their most frequent way to receive groceries was store-pick up, and they were excluded from the present study. The authors of the present study used the three Fornell (1992) items, which have been employed frequently in national customer satisfaction barometers and in academic research, including research on online purchases (Hult et al. 2019). After having adapted these items to the participants’ answers to (1) and (2), this was the specific wording of the items: ‘How satisfied are you with getting the groceries via home delivery from [selected retailer]?’ (1 = very dissatisfied, 7 = very satisfied), ‘To what extent does getting the groceries by home delivery from [selected retailer] meet your expectations?’ (1 = not at all, 7 = totally), and “Imagine a mode of grocery delivery that is perfect in every respect. How near or far from this ideal do you find getting the groceries by home delivery from [selected retailer]? (1 = very far from ideal, 7 = cannot get any closer). Cronbach’s alpha for this scale was .84, and the unweighted average of the responses to the three items was used as a measure of home delivery satisfaction.

Overall satisfaction with the retailer was measured with the same three satisfaction items as above, and a 7-point scale, but the items were adjusted for a focus on the retailer rather than the mode of delivery (i.e., ‘How satisfied are you, overall, with your main online grocery retailer [selected retailer]?’, ‘To what extent does your main online grocery retailer [selected retailer] meet your expectations?’, and ‘Imagine an online grocery retailer that is perfect in every respect. How near or far from this ideal is your online grocery retailer [selected retailer]?’). For this scale alpha was .92, and again the unweighted average of the responses was employed as a satisfaction measure. As a validity check for this satisfaction measure, and given that empirical research typically shows that overall customer satisfaction with an e-retailer is positively associated with repurchase intentions and word-of-mouth intentions (e.g., Koufteros et al. 2014), the authors of the present study included single-item measures of such intentions (‘How likely is it that you would repurchase from [selected retailer] in the near future?’ and ‘If someone you know would ask you for advice regarding an online grocery retailer, how likely is it that you would recommend [selected retailer]?’. Both were scored on a scale ranging from 1 to 7. The overall retailer satisfaction variable was positively associated with both intention variables (r = .56, p < .01; r = .73, p < .01), which indicates predictive validity in the overall satisfaction variable.

Finally, the social importance of home, the moderator variable involved in Hypothesis 4, was assessed with the items ‘In my home, there are people I can turn to in times of need for understanding’, ‘When I am in my home I feel connected to others’, ‘My home gives me an opportunity to talk about my emotions with others’, and ‘In my home, I can share my personal thoughts with others’. The responses were scored on a scale ranging from 1 (strongly disagree) to 7 (strongly agree). Cronbach’s alpha was .94, and the unweighted average of the responses to the four items was used as the variable in the analysis. As an additional indicator of the social importance of home, the authors of the present study used the number of persons in the participant’s household (as reported by the participants; M = 3.22 in the sample). These two indicators were positively associated among the participants (r = .25, p < .01).

3.3 Analysis and results

There was a positive and significant association between home satisfaction and home delivery satisfaction (r = .22, p < .01), so H1 was supported. Moreover, there was a positive and significant association between home delivery satisfaction and overall satisfaction with the retailer (r = .86, p < .01). Thus, H2 was supported. The strength of the latter association signals a potential problem, namely that the participants may have been unable to distinguish between the two types of satisfactions. A discriminant analysis assessment, however, showed that both satisfaction measures conformed to the Fornell-Larcker criteria for discriminant validity.

To test H3, Hayes’ Model 4 for analysis of mediation (Hayes 2012) was employed. Home satisfaction was the independent variable, home delivery satisfaction was the mediating variable, and overall satisfaction with the retailer was the dependent variable. This approach indicted a significant indirect effect from the bootstrap analysis of .18 (95% CI [.06, .30]). Thus H3 was supported. It may be noted that the direct effect was significant, too (b = 0.09, p < .01), which indicates that complimentary mediation was at hand (Zhao, Lynch, and Chen 2010).

H4 was assessed with Hayes Model 1 for moderation analysis (Hayes 2012). That is to say, the authors of the present study examined if the social importance of home would affect the strength of the home satisfaction–home delivery satisfaction association. With the four-item measure of home’s social importance, the analysis indicated that there was indeed such an influence, in the sense that the interaction between home satisfaction and the social importance of home had a positive impact on home delivery satisfaction (b = 0.07, p < .05). That is to say, the association between home satisfaction and home delivery satisfaction was stronger for participants with high as opposed to low values of home’s social importance. The same pattern was reproduced when the number of persons in the participant’s household was used as an alternative indicator of home’s social importance (b = 0.08, p < .01). Thus, H4 was supported for both indicators of the social importance of home.

The supported H4 indicates that a high level of social importance of one’s home boosts the impact of home satisfaction on another home-related variable (i.e., satisfaction with home delivery). The analysis, however, did not assess the extent to which the social importance of home can boost also other effects of home satisfaction. To make the moderation assessment more complete, and to examine if H4 would survive also another type of assessment, an additional analysis was performed: a moderated mediation assessment with Hayes (2012) Model 8. This, then, comprises Model 4 described above, but extended with the social importance of home as a potential moderator of both the home satisfaction–home delivery satisfaction association and the home satisfaction–overall satisfaction association. For social importance of one’s home measured with the four-item measure, the results show that there was a significant moderation effect of the social importance of home on the home satisfaction–home delivery satisfaction association (b = .05, p < .05). However, there was no significant moderation effect for the home satisfaction–overall satisfaction association (b = – .003; p = .84). The same pattern was obtained when the number of persons in the participant’s household was used as an indicator of home’ social importance; the there was a significant moderation effect of the social importance of home on the home satisfaction–home delivery satisfaction association (b = .08, p < .01), but the moderation was not significant when it comes to the home satisfaction–overall satisfaction association (b = .005, p = .75). This, then, provides additional support for H4, and it indicates that there are limits to what the social importance of one’s home can do to associations when the independent variable is home satisfaction.

4. Study 2

The purpose of Study 2 was to re(test) H1-H4. More specifically, Study 2 was an attempt to examine if the Study 1 results would be replicated with data collected from other participants in another national context – a context in which the online share of total grocery purchases is different from what is at hand in the UK. The main target group was the same as in Study 1, however, namely customers who purchase groceries online and whose dominant delivery mode is home delivery.

4.1 Data collection and participants

The variables were measured with items in a questionnaire that was distributed online to members of the MTurk panel. Panel members who were US residents, and who were experienced in purchasing groceries online, were invited to participate in the study (n = 382). The share of total grocery sales stemming from online purchases is lower in the US than in the UK (Dawes and Nenycz-Thiel 2014), and the availability of home delivery is lower for those who do purchase groceries online in the US. In any event, as in Study 1, the authors of the present study selected the participants for which home delivery is the dominant mode of receiving groceries purchased online (n = 162). Five of these participants failed the attention check used in this study and they were removed. This resulted in 157 participants to be included in the hypothesis testing (M_age = 37.66; 88 women and 69 men). The most frequently mentioned retailers used by the participants for online purchases of groceries were Amazon, Walmart, and Whole Foods.

4.2 Measures

Home satisfaction (alpha = .89) and social importance of home (alpha = .92 for the four-item measure) were measured in the same way as in Study 1. Moreover, as in Study 1, the authors of the present study asked the Study 2 participants to write a text about their homes. Also, in Study 2, the emotional tone of the text they provided had a high positive charge (M = 85.34, SD = 27.02). Moreover, again as in Study 1, the emotional tone of home was positively associated with the home satisfaction variable (r = .30, p < .01). When the Fornell (1992) three-item measures from Study 1 for home delivery satisfaction and overall retailer satisfaction were used, however, the ideal item reduced Cronbach’s alpha (i.e., the ideal item was relatively weakly correlated with the other two items). Therefore, for both these satisfaction variables, the unweighted mean of the responses to the two remaining items was used in the analysis. The internal consistency for the home delivery satisfaction measure was r = .70 (p < .01) and for the overall retailer satisfaction measure it was r = .73 (p < .01).

4.3 Analysis and results

There was a positive and significant association between home satisfaction and home delivery satisfaction (r = .37, p < .01), so H1 was supported. Moreover, there was a positive and significant association between home delivery satisfaction and overall satisfaction with the retailer (r = .82, p < .01). This means that H2 was supported. As in Study 1, the strength of this association could indicate that the participants have been unable to distinguish between the two types of satisfactions. Again, however, a discriminant analysis assessment showed that both satisfaction measures conformed to the Fornell-Larcker criteria for discriminant validity.

For H3, Hayes’ Model 4 for analysis of mediation (Hayes 2012) was used. Home satisfaction was the independent variable, home delivery satisfaction was the mediating variable, and overall satisfaction with the retailer was the dependent variable. This analysis showed that there was a significant indirect effect from the bootstrap analysis of 0.28 (95% CI [0.13, 0.50]). This means that H3 was supported. The direct effect (b = 0.16) was also significant (p < .01), which indicates complimentary mediation (Zhao, Lynch, and Chen 2010).

H4 was first assessed with Hayes Model 1 for moderation analysis (Hayes 2012). Home satisfaction was the independent variable, home delivery satisfaction was the dependent variable, and social importance of home was the moderator. When the four-item indicator of social importance was used, this analysis resulted in a significant influence of the interaction between home satisfaction and social importance of home on home delivery satisfaction (b = 0.11, p < .01). More specifically, the association between home satisfaction and home delivery satisfaction was stronger for participants with high as opposed to low levels of home’s social importance. When the number of persons in the participant’s household was employed as an alternative indicator of home’s social importance, the same pattern was produced (b = 0.14, p < .01). H4 was thus supported for both indicators of the social importance of home.

The same extended analysis as in Study 1 was conducted with Hayes’ Model 8 to assess if (a) H4 would be supported also with an alternative analysis and (b) there are limits to what the social importance of one’s home is capable of moderating when it comes to the influence of home satisfaction on other (and less-home related) outcomes. For the social importance of one’s home measured with the four-item measure, the there was a significant moderation effect of the social importance of home on the home satisfaction–home delivery satisfaction association (b = .11, p < .01). However, there was no significant moderation effect for the home satisfaction–overall satisfaction association (b = −0.03; p = .19). The same pattern was obtained when the number of persons in the participant’s household was used as an indicator of home’ social importance; in this case, there was a significant moderation effect of the social importance of home on the home satisfaction–home delivery satisfaction association (b = .14, p < .01), but the moderation was not significant when it comes to the home satisfaction–overall satisfaction association (b = .02, p = .43). This provides additional support for H4. As in Study 1, it also indicates that there are limits to what the social importance of one’s home can do to associations when the independent variable is home satisfaction.

5. Conclusions

5.1 Summary of main findings

Research over many years has shown that people’s experiences and perceptions of their everyday environments can influence their well-being (Saxbe and Repetti 2010), and the present study examined if one particular everyday environment, home, would influence evaluations of commercial offers. The main outcome, from both Study 1 and Study 2, suggests that this is indeed likely: for customers who purchase groceries online, and who use home delivery, the main finding was a positive association between home satisfaction and home delivery satisfaction as well as a mediated influence of home satisfaction on overall retailer satisfaction.

5.2 Discussion

Increased mobility, globalization, progress in transportation and communication technologies, and the fast growth of online retailing may perhaps suggest that places do not matter much anymore. This, however, seems to be a wrong conclusion. Places have not lost their meaning for us humans; copious studies show that attachment to places continues to be strong (Gieryn 2000; Lewicka 2011). Indeed, places saturate social life; it is a medium through which social life happens (Gieryn 2000). In light of this, the present study has been an attempt to ‘emplace’ online retailing from the customer’s point of view by explicitly recognizing home as the place in which what is ordered ends up. More specifically, the present study contributes with an additional variable, home satisfaction, to the growing literature on online shopping and last mile deliveries. Home satisfaction has been neglected in this literature, even in ethnographic attempts to provide rich contextualized case studies of online shopping practices (Elms, De Kervenoael, and Hallsworth 2016). The importance of home, however, has been stressed in fields outside marketing and retailing (e.g., Easthope 2004; Graham, Gosling, and Travis 2015; Noble 2004; Sixsmith 1986; Somerville 1992), so the present study should also be seen as an attempt to build bridges between literatures that would benefit from closer contact.

The result that home satisfaction can influence other satisfaction variables, particularly satisfaction with a home-related offer, is not surprising given that several authors over the years has argued – and shown empirically – that the atmosphere of brick and mortar stores can influence customer responses in such stores (e.g., Donovan et al. 1994; Kotler 1973–1974; Spies, Hesse, and Loesch 1997). That is to say, in light of store atmosphere research, it seems reasonable to expect that the home atmosphere can influence customer responses, too – particularly in a situation in which the home is an integrated part of the distribution process. Similarly, the results are not surprising from a priming point of view, because priming effects are assumed to be particularly strong when the content of cognitive categories is easily accessible (Vohs 2015). And given the centrality of home for most people (which makes home highly accessible), a home-related offer is expected to activate the content of a general home category (which in turn can inform evaluations of a home-related offer in a valence-congruent way). To date, however, there has not been much priming research in a consumer context comprising home-related stimuli.

The authors of the present study believe that the home satisfaction variable (as an upstream variable, capable of influence in the beginning of a chain of variables), and its association with overall retailer satisfaction, can provide additional explanatory material to results obtained in previous studies (in which home satisfaction was not examined). In other words, explicitly including home satisfaction in the nomological net of variables in online shopping studies can add further precision to the understanding of relationships between online shopping variables. For example, there have been several attempts to compare what drives customer satisfaction in online and offline settings (e.g., Hult et al. 2019). Given that the content of a general home category becomes more accessible (and thus becomes more causally potent) in an online setting with home delivery compared to an offline setting, an explicit account of home aspects may add more precision to offline-online comparisons. Moreover, it has been observed that loyalty tends to be higher online than offline, and one reason is that saveable lists make customers less eager to switch (Dawes and Nenycz-Thiel 2014). However, given that also overall satisfaction with a retailer can boost retailer loyalty, and given that home satisfaction influences overall satisfaction, it seems possible that higher levels of loyalty for online customers may at least partly stem from home satisfaction. Similarly, it has been argued that perceived price and perceived risk affect perceived value for online customers, which in turn has a positive impact on purchase intentions (Kim and Gupta 2009). But given that also overall retailer satisfaction is an antecedent to purchase intentions, and given again that home satisfaction can boost overall retailer satisfaction, it seems as if home satisfaction would have a role as an (upstream) variable in this case, too.

The results in the present study, as well as the assumption that home aspects influence consumer behavior variables, also call for some caution in researchers’ use of wording in studies of online shopping behavior. It is not uncommon, for example, that researchers make statements such as ‘15 % of all retail spending now occurs online’ (Munson, Tiropanis, and Lowe 2017). It is also common that online grocery shopping is viewed as taking place in a new ‘environment’ that allows customers to find products in a number of different ways (Benn et al. 2015). Such statements and views, however, conceal that the online shopper is always emplaced somewhere and that this somewhere is an environment, too. By ignoring this, one may easily also ignore that the physical environment is still fully capable of influencing us humans.

5.3 Managerial implications

Given that the positive home satisfaction–home delivery satisfaction association is a function of spreading activation and affect transfer, firms offering home delivery may benefit if they make deliberate attempts to facilitate customers’ accessibility to content in the home category. This implies, for example, that deliveries to the customer’s home should indeed be labelled home delivery. Marketing material describing what such deliveries comprise may also be used to stress home aspects by a careful use of words and pictures with home content. Such material, then, may be used to attract new customers to choose home delivery and it may boost the satisfaction of existing customers. Given the centrality and valenced charge of home, it may even be possible for firms that do not offer home delivery to capitalize on home-related stimuli to evoke content in the home category. For example, online retailers who are now rushing into a click-and-collect format, in which ordered goods are not delivered to consumers’ homes (Gielens, Gijsbrechts, and Geyskens 2021), may want to stress that this type of delivery brings the goods closer to customers’ homes or that it allows customers to spend more time at home.

Deliberately priming consumers with home-related stimuli, however, is likely to backfire if it is used for customers with low home satisfaction. Several authors have stressed that home can be very negatively charged for some individuals (Mallett 2004; Moore 2000), and the identification of such individuals should be taken seriously by marketers who wish to use home-related elements in a home delivery context. One possible indicator variable is length of residence in one particular home; several studies indicate that it is associated with home attachment (Lewicka 2011; Somerville 1997). Clearly, variables of this type are less sensitive and therefore easier to use in data collections than those that tap directly into the potentially dark side of what may go on in people’s homes.

5.4 Limitations and suggestions for further research

The present study comprised online purchases of groceries – a product category with items that are typically consumed at home. This may make the content of a general home category more accessible compared to other products that are not consumed at home, which in the next step may boost the influence of home satisfaction on other satisfaction variables. Further research is therefore needed to examine if the responses in the present study would be obtained also for evaluations of retailers’ delivery solutions and overall retailer satisfaction with respect to other products than groceries. Such research is needed also because of another reason: online sales of groceries represents only a modest part of the total consumer spending on groceries. That is to say, grocery retailing has not followed suit in relation to the booming of online retail sales in other sectors (Gielens, Gijsbrechts, and Geyskens 2021). Similarly, many products that are purchased offline are consumed at home, and researchers should explore the causal potency of home satisfaction for evaluations of such products. A more detailed picture of the role of home satisfaction would also be obtained if consumer satisfaction is compared between customer groups who consume the (same) product at home and in other places.

Moreover, the present study comprised consumers from the UK (Study 1) and the US (study 2). Even though the development of various last mile delivery solutions is a relatively recent phenomenon, it seems as if different countries are already moving in different directions with respect to the most prevalent last mile delivery solutions (Hübner, Kuhn, and Wollenburg 2016; Morganti et al. 2014). This calls for further research on consumers in other geographical areas.

The measurement of home satisfaction and other home-related variables (such as the social importance of home) is another issue that calls for caution. One aspect has to do with the distribution of such variables; in the present study, home satisfaction was negatively skewed (i.e., the majority of the participants reported high rather than low levels of home satisfaction). Home satisfaction, then, was subject to the same skewed distribution as customer satisfaction typically is (Fornell 1992; Peterson and Wilson 1992). Skewness is relatively unproblematic from a conceptual point of view in the case of customer satisfaction; on a market with several competing suppliers, and with information about various offers available before consumers make their choices, high levels of customer satisfaction is what one would expect (Fornell 1992). Home, however, is unlikely to be subject to the same freedom of choice. So why was home satisfaction a negatively skewed variable? For participants in research projects, responding with low levels of home satisfaction may signal personal failure to a much larger extent than reporting low customer satisfaction, so a skewed home satisfaction distribution may reflect social desirability in the responses. Further research on home satisfaction needs to find ways to assess this possibility.

Further research is also needed to examine more closely the mechanisms by which home satisfaction influences other variables. In the present study, it was assumed that home-related stimuli are able to prime a receiver in such a way that evaluations of home-related offers are influenced. It should be noted, however, that several attempts to replicate findings of priming effects have resulted in modest or no effects on dependent variables (e.g., Caruso, Shapira, and Landy 2017; Gill and El Gamal 2014). Although the assumed priming mechanism in the present study (an offer with home-related content is likely to activate material in our cognitive category for home and the valenced charge of this material is likely to carry over to evaluations of the offer) may have been less complex than in some cases for which replications were not successful, the correlational nature of the present study does not test the assumed mechanism. Further research, then, is needed to assess the proposed priming-based mechanism in the present study.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Notes on contributors

Reema Singh

Reema Singh is a post-doctoral researcher at Department of Economics and Management, Lund University, Sweden, where she is extending her doctoral thesis work to understand consumers’ responses to grocery retailer attributes. Her research is focused on the customer experience, customer satisfaction and customers’ behavioral responses.

Magnus Söderlund

Magnus Söderlund is Professor of Marketing and Head of the Center for Consumer Marketing at Stockholm School of Economics, Sweden. He is also a Senior Fellow at Hanken School of Economics, Finland. His main research interest is how consumers react to various marketing-related activities.