Read Aloud the Text Content

This audio was created by Woord's Text to Speech service by content creators from all around the world.

Text Content or SSML code:

Abstract Advances in technology change the way consumers search and shop for products. Emerging is the trend of homeshopping devices such as Amazon’s Alexa and Google Home, which allow consumers to search or order products. We investigate how consumer brand and technology preferences may interact with the functionalities of technologyenabled shopping TES devices to determine the channel structure and market competition. In specific, we break the functionalities of the TES devices into two 1 the shopping support functionality SSF, and 2 the ordering convenience functionality OCF. Via a series of experiments, we document that stronger brand preferences are negatively correlated with the willingness to use a TES device that offers SSF. However, there is no association with brand preferences and desire to use a TES device when it offers OCF. We build an analytical model integrating the findings from these experiments, and then derive the equilibrium channel and pricing strategies for two competing retailers. Our findings show that the functionality of TES devices results in vastly different distribution and pricing strategies in retail markets. In particular, consumers’ heterogeneous valuation of the SSF results in a monopolistic adoption of TES devices by the retailers in equilibrium, and generates Pareto improvements for both. In contrast, when the TES devices offer OCF, in equilibrium, retailers adopt TES channels competitively, resulting in a prisoners dilemma outcome. In the extensions, studying a thirdparty technology developers decision to invest in OCF and SSF technologies, we show that the contrast between the channel strategies under the OCF and the SSF also impact the incentives to develop TES. We show that in some cases, in an effort to mitigate downstream retail competition, the provider may prefer not to offer the best possible OCF technology to consumers. These findings shed light on the future adoption and the functionalities of shopping technologies offered by retailers. Previous article in issue Next article in issue Keywords Shopping technologyChannel designTechnology investment 1. Introduction Past few years witnessed a quiet revolution in the number of technologyenabled shopping TES devices—devices that offer an alternative channel where consumers use interactive technologies to shop—introduced to the market. By the end of 2019, for instance, an estimated 111.8 million people in the U.S.—nearly one third of the population—owned a voiceassisted device, such as Amazon’s Alexa or Google Home Petrock, 2019. Up to one third of those who own a smart device use it to shop for goods in Europe Kinsella, 2018 where the sales through these devices reached 2.1 billion in 2018 Thakker, 2019. In the U.S. too, the use of TES devices is growing, with triple yearonyear growth between 2017 and 2018 for Amazon’s Alexa Toplin, 2018. Consumer spending on Alexa mostly ranges between 25 and 199 per item Edison Research, 2018 and common categories of shopping intention include groceries 45 of individuals, clothing 46, and specialty products such as books or pet supplies 49 Capgemini, 2018. These devices and new purchasing options are duly noticed by retail executives, too. As a VP of ecommerce, digital marketing, and innovation at Lands’ End commented, “Given the explosive growth in voice search and equally explosive growth in shipments for Echo devices, it is the time to start to test and lean into this emerging channel” Berthene, 2017. In this paper, we take a first rigorous look at a firm’s strategic entry decision on this emerging channel. Our research aims to explore how the technology preferences and brand preferences of consumers may influence the structure of the market competition and the channel of technologyenabled shopping TES devices. TES devices come with skills designed primarily to help consumers to review product information, confirm product fit, and navigate through the shopping journey. Using Amazon’s Alexa as an example, “MySomm” is a skill which suggests the most suitable wine for its users, and “Kit” is a skill which answers users’ product queries such as “what are good coffee makers” Trotter, 2017. We refer to this category of skills as shopping support functionality SSF. There are also skills that primarily facilitate a purchase by making ordering experience more convenient and pleasurable. For example, the skill “Ask Peapod” allows consumers to order groceries by voice Thakker, 2019, “Grubhub Alexa” makes it possible for users to enjoy the convenience of handsfree reordering1. We refer to these skills as a device for ordering convenience functionality OCF. This distinction between SSF and OCF for new retail technologies is not new and has been discussed in the literature Burke, 2002. Based on our classification, out of the top shopping skills which have been rated by at least 10 users, 38 are primarily for SSF, 50 for OCF, and the rest for both SSF and OCF. In this paper, we will investigate and show that consumers’ preferences for the device functionalities are actually related to their brand preferences. As the use of TES devices in retailing is in early stages, much of their vast potential is still unfolding, and competing retailers have just begun to reach their customers through TES devices. Branding in this new environment has thus become important to scholars and practitioners alike. Meyersohn 2018 suggests, for instance, in this new environment, brand names will become more important because consumers may interact with the devices ordering a brand by name to purchase items. What is even more important is to build trust between consumers and new technologies, as pointed out by Dawar 2018. As we shall discuss shortly, issues related to the interactions between consumers and devices, as well as to consumer buying behaviors in a technologyassisted environment, are critical managerial issues that will shape how the technology evolves. In this paper, we contribute to the emerging literature on TES devices by theoretically investigating the equilibrium channel structure and adoption of devices as a strategic choice by competing retailers and technology providers. Our research focuses on a set of questions that naturally arise as the use of TES devices grows. First, when should competing retailers embrace TES devices Should retailers always try to secure the first mover advantage in the new channel, as some executives suggest Berthene, 2017 Of course, a firm’s decision to integrate shopping technologies should depend on how the consumers may embrace the technology. This brings up a second research question. What kinds of consumers prefer which kinds of TES devices The answers to this question will shed light on the incentive environment in which retailers and technology providers would make their decisions. Given this incentive environment, a third and managerially more important question is whether competing retailers face different incentives in adopting TES devices with different functionalities. If retailers do vary their adoption strategies based on TES functionalities, what incentives do those strategies provide to the technology provider so as to shape the channel structure This last question begs a cluster of other questions. Would a technology provider favor SSF or OCF For a functionality it chooses to provide, does it have an incentive to make the functionality as good as feasible and make the technology broadly available to all the competing retailers In this paper, we will develop an analytical model to provide some preliminary answers to all these questions. Our research starts with the empirical documentation that consumers’ technology preferences are related to their brand preferences. We investigate experimentally whether a consumer with a strong or weak preference for a brand may prefer a TES device of a particular functionality, SSF or OCF. By incorporating this preference structure into our gametheoretical modeling, we show that competing retailers may or may not want to embrace the TES device of a certain functionality depending on three factors the customers who favor the new channel, the benefits to a retailer from the price discrimination enabled by the new channel, and the functionality of an available TES device. As the technology provider is the one who sells access to TES devices, the consumers’ preference over TES devices and the retailers’ adoption decisions will then underlie the incentives for the technology provider to develop TES devices of different functionalities. This research strategy enables us to generate a number of insights into a new, evolving phenomenon. We first show, through experiments, that consumers with weaker preferences for a brand derive higher positive utility from a TES device with SSF, whereas OCF provides a homogeneous and positive utility, independent of brand preferences. This preference structure implies that, in a channel with OCF benefits, competitive entry of retailers will take place even though it results in a prisoner’s dilemma outcome. In contrast, competitive entry will not occur in a channel with SSF benefits, but unilateral entry by a retailer can happen to the benefit of both retailers. Given these outcomes facing the retailers, the technology provider, for its own profitability, may not have the same incentive to develop a TES device with SSF as one with OCF. When both SSF and OCF can be incorporated into a TES device, the technology provider has an incentive to provide the most advanced level of SSF, but underprovides OCF, even when the cost of offering the technology is zero. Our paper contributes to the existing literature in three ways. First, there is a considerable amount of literature on consumer reactions to the use of technology in the marketplace. This literature has documented that consumer reactions tend to be heterogeneous in that some may embrace new technology for convenience e.g., Dekimpe, Geyskens, Gielens, 2020 and better decision making e.g., Soll and Larrick, 2009, Surowiecki, 2004 or for the reasons of trust e.g., Logg, Minson, Moore, 2019. Yet some others may display technology aversion as they feel uncomfortable about or distrustful of relying on machines to make a decision Dietvorst et al., 2016, Dietvorst et al., 2015, Yeomans et al., 2019. Apparently, this heterogeneity depends on the nature of the decisions to be made Chen, 2009, Longoni et al., 2019, Dzindolet et al., 2002, ingroup bias Brewer, 1979, Sinha et al., 2001, Önkal et al., 2009, cost of learning Mick and Fournier, 1998, Goodman, 1988, and the fear of technology George, 2014, Ferrari et al., 2016, Conniff, 2011. It also depends on whether one is an expert in a field Logg et al., 2019, Castelo et al., 2019, Highhouse, 2008. Our contribution is to show that this heterogeneity also depends on the strength of consumer preference for a brand. Second, there is a growing literature on consumer interactions with technology Kleinberg, Lakkaraju, Leskovec, Ludwig, Mullainathan, 2017 and the implications of such interactions on firm strategy Srinivasan et al., 2002, Ram and Sheth, 1989, Sriram et al., 2010. We contribute to this stream of literature by investigating how the functionalities of technology, as articulated by Burke 2002, can play a critical role in consumer interactions with technology, which in turn incentivizes the competing retailers to use different technology adoption strategies and the technology provider to develop TES devices with different functionalities. Third, our paper also adds to a long stream of literature on channel structure Lee Staelin, 1997, channel coordination Gerstner and Hess, 1995, Choi, 1991, Lal, 1990, Jeuland and Shugan, 2008, Moorthy, 1988, and channel management Coughlan Wernerfelt, 1989. In this regard, our contribution is to show that consumer technology preference can indeed help to shape a channel structure and call for some new managerial imperatives. In our study, the preference for a channel depends on the preference for the brands, and thus firms can use this information to price discriminate across channels Bergemann, Brooks, Morris, 2015. Price discrimination as a function of channel preference has been a cornerstone in economics Gerstner et al., 1994, Cavallo, 2017 and marketing research Zettelmeyer, 2000, Besanko et al., 2003, Liu and Zhang, 2006. However, there is, to our knowledge, little research demonstrating that consumers’ technology preference can be a tool for price discrimination, and we also fill this gap in the literature. The rest of the paper is structured as follows. We describe the experimental setting and the findings from them in Section 2.1. Next, in Sections 2.2 Study 1, 2.3 Study 2 we develop a theoretical model and Section 3 discusses extensions. Finally, in Section 4, we conclude. 2. Model Firm decisions depend on consumer choices, and consumer choices depend on consumer preferences for brands and technology. The interaction of these two kinds of consumer preferences may provide different incentives for a firm to embrace or not a particular shopping technology. Consumer brand and technology preferences can be related to each other positively or negatively, or they may not be related to each other at all. Past literature has not, to our knowledge, shed light on this relationship of growing importance. Thus, while in our theoretical exercise, we can explore all possible permutations of this relationship, our analysis will be much more relevant if we have a more concrete understanding of this relationship to motivate our modeling assumptions. To ground our analysis in realistic assumptions, we first investigate this relationship experimentally and pin it down with accuracy. 2.1. Brand and technology preferences Experimental evidence Past research has indicated that consumers vary in their response to algorithms and thus to technologyenabled devices that use them Dietvorst et al., 2016, Logg et al., 2019. While some consumers see a benefit from TES devices, others see little benefit and even purposefully refrain from using them. Burke 2002 discusses two relevant dimensions of benefits from shopping technologies that people seek when they buy utilitarian goods 1 detailed product information that supports consumers when they are shopping, and 2 a fast and convenient ordering experience. We shall refer to the former as shopping support functionality abbreviated as “SSF” and the latter as ordering convenience functionality abbreviated as “OCF”. A TES device in our model will be characterized by either SSF or OCF, or both functions. As Burke 2002 points out, “it is not the technology per se, but how it is used to create value for customers that will determine its success.” Of course, different customers may use different technologies differently, and hence derive different values from them. To the extent that consumers with different brand preferences may use a technology differently, it is rather expected that their brand and technology preferences may be related. Past research has shown some plausible evidence that value from SSF is negatively correlated with the strength of one’s brand preference. Consumers who hold a strong preference towards a brand should also be more familiar with its products Coupey et al., 1998, Wright, 1975, so they tend to search less when considering existing alternatives Johnson and Russo, 1984, Bettman and Park, 1980. They are therefore less likely to benefit from SSF and may even consider it a nuisance. Indeed, the theory on the value of information Raiffa Schlaifer, 1961 for consumers facing some uncertainty in a purchasing environment also suggests the same insight. A strong brand preference will reduce the importance of uncertainty in consumer purchase decision, and hence reduce the value of SSF. In contrast, there is no evidence, as far as we know, about any heterogeneity in brand preferences and OCF benefits. We test these negative and no correlation hypotheses in the two studies described below. The scripts for the experiments are in Appendix A.3. 2.2. Study 1 In Study 1, we recruited 251 subjects on Amazon’s MTurk. The study used a withinsubject design. The survey was advertised as a 5min survey about shopping preferences. Participation in the MTurk Survey was restricted to respondents located within the U.S. and over 18 years old. The median respondent took about 3.2 min to complete the survey. The payment scheme was on par with other MTurk surveys Bottan PerezTruglia, 2020. Participants received a fixed fee of 0.50 for participation. Each subject was first asked to indicate hisher preference between two headphone brands Sony Bose no preference as well as how strong the preference was if there was any. We coded the responses “strongly preferring Sony or Bose,” “weakly preferring Sony or Bose,” and “no preference,” as 2, 1, and 0, respectively, as a subject’s brand preference strength. Each subject was then introduced a shopping device with either SSF or OCF, which described the device either “aims to provide shopping support by, for example, providing product information, and telling you how well a certain product fits you” SSF or “aims to provide convenient ordering experience by, for example, enabling voice interactions, ‘oneclick’ ordering, and quick payment” OCF. They were then asked, compared to buying a headphone in stores, how much more or less benefit shopping for it via the described device would give them a 5point scale was used, and we code them as 2, 1, 0, , and where positive numbers indicate more benefit and negative numbers indicate less benefit. Finally, they were asked to describe in a few words why they thought the shopping device would provide less or more benefit than buying in store. The whole procedure repeated for the other functionality. In other words, this is a withinsubject design where subjects saw both SSF and OCF and the order was randomized. There was a multiplechoice question for attention check after each functionality of the device was introduced. Among all the 251 subjects, 234 of them passed the attention check, whose answer will be considered for further analysis. We regress subjects’ perceived benefit from the device on their brand preference strength for a given brand, for conditions SSF and OCF separately. Table 1 describes the regression results. The results show that a consumer with a weaker brand preference gets a higher benefit from using a device with SSF compared to one with a stronger brand preference, and the effect is significant at the level . As a contrast, only an insignificant positive relationship between consumers’ brand preference strength and benefit from using an OCF device has been found. Table 1. Regression results for Study 1. Empty Cell Dependent variable Empty Cell Benefit from SSF Benefit from OCF Empty Cell 1 2 Brand Preference Strength −0.146∗ 0.101 0.082 0.085 Constant 0.412 0.256 0.105 0.109 Observations 234 234 R2 0.013 0.006 Adjusted R2 0.009 0.002 Residual Std. Error df 232 1.039 1.079 F Statistic df 1 232 3.167∗ 1.399 Note p0.1 p0.05 p0.01. 2.3. Study 2 In Study 2, we abstract away from a specific product category to see if the results in Study 1 are generalizable. We recruited 1,028 subjects on Amazon’s MTurk. The study used a 2 2 design technologyenabled shopping device functionality ordering convenience OCFshopping support SSF product preference strongweak. We followed the same recruitment requirements and procedure as in Study 1. The median respondent took about 3 min to complete the survey. Towards the end of the survey, we introduced an attention check, similar to the one used in Bottan and PerezTruglia 2017. A total of 97.5 of the respondents passed the attention check. After removing those who failed the attention check, we ended up with 1,002 subjects. Participants were told that they were about to buy a product, and there were only two brands, A and B, in the market. They were then randomly assigned to a condition indicating their preference for product A over B, where a following statement indicated that they had a “strong preference for Brand A over Brand B” or that they had a “weak preference for Brand A over Brand B.” These two conditions will help us to conveniently put all consumers on the Hotelling line as we develop our analytical model. Each subject was then randomly assigned to a second condition of device functionality, SSF or OCF. Based on the pretest in Study 1 where we asked subjects why they think a shopping device with SSF or OCF provides less or more benefit than buying in store, we identified the major 3–4 weaknesses and strengths mentioned for both SSF and OCF from Study 1. A participant under SSF or OCF would be taken to a page where they would observe both the pros and cons of using the device. They were then asked, compared to buying Brand A in store, how much benefit they derive from shopping for Brand A via the described device. Participants used a sliding scale ranging from to with 0.1 precision to indicate how much the device benefited them. In Fig. 1, we present the consumer preference for shopping with the device, relative to shopping in store under OCF and SSF treatments. In particular, the experiment shows that, while there were no statistical differences between the panelists assigned to the strongweak brand preference conditions under the OCF treatment, those under the SSF treatment showed a significant difference. In particular, panelists assigned to the weak preference condition said they found the SSF device more valuable, and had a higher preference to shop with the technologyassisted device.