تعمیم در پذیرش نوآوری مصرف کننده : فرا تحلیل در مورد محرک های قصد و رفتار

Previous research has shown that consumer intentions to adopt innovations are often poor predictors of adoption behavior. An important reason for this may be that the evaluative criteria consumers use in both stages of the adoption process weigh differently. Using construal level theory, we develop expectations on the influence of innovation characteristics across the intention and behavior stages of the adoption process. Using meta-analysis, we derive generalizations on drivers of intentions and actual innovation adoption behavior. The results show important differences across both stages. Consumers show higher levels of adoption intention for innovations that are more complex, better match their needs, and involve lower uncertainty. However, consumers are found to actually adopt innovations with less complexity and higher relative advantages. Adopter demographics are found to explain little variance in adoption intention and behavior, whereas adopter psychographics are found to be influential in both stages. These findings have implications for innovation adoption theory, for managers involved in new product and service marketing, and for future research on innovation adoption.

Understanding whether and why consumers will adopt a new product or service is a critical insight for managers involved in marketing innovations. It is common practice to obtain such an understanding based on market research of consumers' attitudes toward the innovation and their purchase intention. However, many marketers have found out the hard way that consumers who “talk the talk” in surveys do not always “walk the walk” when it comes to innovation adoption. Consider for example the videophone. As early as 1964, AT&T tested its version of this innovation, the Picturephone, during the New York World Fair (Schnaars & Wymbs 2004). Market researchers interviewed almost 700 individuals making transcontinental videophone calls. Respondents rated the service favorably, and 45% indicated a need for the service at home. However, the launch of the innovation in the consumer market failed, as few consumers adopted the innovation. After a later unsuccessful launch in the business market, AT&T eventually decided to terminate the Picturephone by the mid-1970s. The case of the videophone can hardly be considered an isolated example. A recent report by Synovate that reviewed studies on product purchase intention and behavior from diverse categories, such as fast moving consumer goods, cars, PCs, appliances, clothing, and home furnishings, suggested that “91% of the variance [in purchase behavior] is not captured by purchase intent” (Synovate 2007, p. 4). Managerial practice shows that intentions are often used as proxy measures for adoption behavior (Van Ittersum and Feinberg, 2010 and Young et al., 1998). The case of the videophone painfully illustrates that market research showing favorable evaluation and high adoption intention of an innovation can be misleading. Indeed, academic research on the adoption of innovations has shown that intentions are far from perfect predictors of behavior. A meta-analysis by Sheppard, Hartwick, & Warshaw (1988) reported a correlation of .53 between intention and behavior. Moreover, Morwitz, Steckel, and Gupta (2007) found that the correlation between intention and behavior was significantly lower for new products than for existing ones. Several reasons have been suggested for this gap (e.g., Morwitz et al., 2007, Sun and Morwitz, 2010 and Van Ittersum and Feinberg, 2010), including consumers' change of intentions over time (Morrison, 1979), the use of biased estimates in research (Van Ittersum & Feinberg, 2010), and the inability of the consumer to anticipate unexpected events that may affect the adoption decision (Morwitz et al., 2007). Typically, the evaluation of a product or service, such as an innovation, is a goal-directed process in which consumers evaluate its attributes with certain use purposes and situations in mind (e.g., Gardial et al., 1994 and Vandecasteele and Geuens, 2010). Innovation adoption is best represented by a process of multiple stages through which an individual passes, from first awareness to continued use of the innovation (Rogers, 2003). During this complex decision process, the potential adopter forms perceptions of the characteristics of the innovation (e.g., Castaño et al., 2008 and Wood and Lynch, 2002) and weighs them in a choice decision (e.g., Bettman 1979). At different stages of the innovation adoption process, use purposes and situations may be perceived differently, thus affecting the weight of evaluative criteria in the decision process. Consumers may therefore weigh attributes differently in situations of purchase intention versus purchase behavior, resulting in an imperfect relationship between intention and behavior (Gollwitzer 1999). For example, in the case of the videophone, the high quality of personal visual communication may have led consumers to a favorable pre-adoption evaluation of the innovation, indicating high adoption intentions, whereas the importance of its perceived costs may have prevented consumers from actual purchase. Thus, it is best to distinguish intention and behavior as distinct dependent variables (Bemmaor, 1995 and Jamieson and Bass, 1989) that represent different, subsequent stages of the innovation adoption process (Rogers, 2003). A rich body of research has developed in the past decades that addresses factors affecting innovation adoption decisions by consumers in marketing science (Hauser et al., 2006 and Rogers, 2003). However, only more recently have insights into how antecedents of consumer innovation adoption differ between adoption process stages been developed in the literature (e.g., Alexander et al., 2008, Castaño et al., 2008 and Wood and Moreau, 2006). Practitioners could substantially benefit from a better understanding of the antecedents of consumers' intentions to adopt an innovation versus those of their actual behavior. Although Tornatzky and Klein (1982) have previously provided insight on innovation adoption drivers based on a meta-analysis of academic research, their study did not discriminate between intention and behavior. Moreover, their study was conducted almost three decades ago, thus excluding a large body of research conducted since then. The objective of this paper is to shed more light on whether and, if so, how drivers of innovation adoption that have been considered as indicators of innovation acceptance in the literature vary across the intention and behavior stages of the adoption process. To do so, this study uses meta-analysis (e.g., Assmus, Farley, & Lehmann 1984) on antecedents of both adoption intention and adoption behavior. As such, this study aims to obtain more insight in a field of research (i.e., consumer innovation adoption) rather than in a specific relation. This meta-analysis focuses on studies in marketing literature that address the adoption of new products by consumers. We further assess whether and how contextual and methodological factors have moderated the effects found on innovation adoption. We generalize the findings of 77 studies related to consumer innovation adoption published in marketing from 1970 to mid-2007. This method allows us to obtain generalized findings on both adoption outcomes and their antecedents. The main results of this analysis include the following: • Innovation characteristics have a strong but different effect on adoption process stages: ○ Benefits affect both intention and behavior, with compatibility being a stronger driver of intention and relative advantage of behavior; ○ Complexity has a positive effect on intention, but negatively affects adoption behavior; ○ Perceived uncertainty shows a stronger effect on intention than on adoption behavior. • Adopter demographics show minor influence on innovation adoption. • Adopter psychographics are found to be powerful drivers of innovation adoption, with respect to both intention and behavior. This study contributes to the literature by showing that drivers of innovation adoption to a large extent affect intention and behavior differently. Therefore, the findings show that it is important to take a dynamic perspective of innovation adoption. The study also suggests new directions for future research and provides implications for managers involved in new product and service marketing. This paper is organized as follows. First, we provide the theoretical background of the study. Second, we discuss the procedures that were used to conduct the literature review and the development of the database and elaborate on the methods employed to analyze the data. Third, we present the findings of the meta-analysis pertaining both to the substantive information on the effects and the existence of contextual and methodological moderators. Fourth, we discuss the findings and draw implications for practitioners dealing with the marketing of new products. Finally, we discuss the limitations of the present study and implications for future research on innovation adoption.

The objective of this meta-analysis was to investigate whether and, if so, how antecedents of innovation adoption vary across the intention and behavior stages of the adoption process. The findings of this research provide generalizations on the key drivers of consumer innovation adoption, revealing important differences between the influence and explanatory power of antecedents across the adoption process stages. These findings thus stress the relevance of distinguishing between evaluative criteria that help to explain consumers' purchase intention of an innovation and those that explain actual purchase behavior. As such, they have implications for research on innovation adoption and for managers involved in marketing innovations. Before turning to the implications, we first discuss the surprising and unexpected results of this study. Based on the Construal Level Theory (Trope & Liberman, 2003), we expected the benefits of an innovation to be most influential on a consumer's innovation adoption decision at the intention stage. We do find, as expected, that compatibility has a stronger positive effect on intention than on behavior. Relative advantage, however, has a somewhat stronger positive effect on behavior than on intention. An explanation for this finding could be that relative advantages of an innovation represent experience qualities that can best be assessed when consequences of using the innovation are evaluated, consistent with Means-End Chain Theory (e.g., Gutman, 1982). As these consequences are difficult to determine prior to purchase (Gardial et al., 1994), the effect of relative advantage on intention may be less strong. Similarly, studies in health psychology on the adoption of new healthy behaviors report that individuals attach more advantages to these new behaviors in the behavior stage than in the intention stage (Prochaska et al., 1994). The effects found for trialability and observability on behavior were also unexpected. Both trialability and observability show negative effects at this stage of the innovation adoption process, although their effect sizes are very small. An explanation may be found in research by Karahanna et al. (1999), which suggests that the relevance of trialability vanishes once the innovation is in use. Observability becomes less relevant as well because of personal experience with the innovation. Taken together, the findings lend partial support to the notion that benefits are most influential at the intention stage, especially in terms of the innovation's perceived compatibility. Contrary to expectations, perceptions of relative advantages are more pronounced at the behavior stage, which may be caused by context-specific experiences with the innovation. Perceived complexity was found to more strongly inhibit behavior than intention, as expected. Contrary to expectations, however, we found its effect on intention to be positive. An explanation for this finding may be that prior to adoption, at the intention stage, consumers underestimate the (potentially negative) role of complexity and are overconfident about the usability of the innovation (Wood & Moreau, 2006). Likewise, Thompson, Hamilton, and Rust (2005) showed that consumers value product features (which increase the innovation's complexity) more positively and attach less weight to the innovation's usability before use than after. Taylor and Todd (1995) suggest that complexity may signal higher quality. Rather than a cost, complexity could therefore signal newness and advancement. In this vein, complexity functions as a ‘trigger of interest’ for adoption intention (Berlyne, 1971 and Messinger, 1998), but becomes a barrier for behavior. Perceived uncertainty has a negative effect on both intention and behavior, as expected, although with a stronger effect on intention than on behavior. Consistent with the small effect found at the behavior stage, Demoulin and Zidda (2009) reported that perceived risk does not discriminate between adopters and non-adopters of a new loyalty card among consumers. An explanation for the relatively weak effect of uncertainty at the behavior stage may be that adopters were compared to a group of non-adopters that included consumers in any stage of the adoption process prior to behavior. These consumers were likely to be very heterogeneous in terms of their perceptions of innovation uncertainty, as some of them may not even have been aware of the innovation. Furthermore, the role of different types of uncertainty may also account for the effects found. Because we were not able to distinguish between alternative types of uncertainty (cf. Castaño et al., 2008), the results we found may have been affected by different types of uncertainty having received different attention in previous studies that were included in the meta-analysis. With regard to adopter characteristics, we found limited support for the impact of socio-demographics. Steenkamp and Gielens (2003) showed that the moderating role of market factors, such as the number of brands in a market, has a strong effect on the influence of age and income, resulting in non-significant main effects. Based on the meta-analysis, we conclude that socio-demographics (i.e., age, education, and income) do not have a generalizable systematic impact on adoption intention and adoption behavior. Adopter psychographics, however, including product involvement, innovativeness, and opinion leadership, account for a relatively high percentage of the explained variance of adoption intention and, to a lesser extent, of adoption behavior. Involvement has an especially strong effect on intention to adopt an innovation. Consumers who have greater familiarity with a product category need less cognitive effort to evaluate the innovation, which makes them more likely to form adoption intentions (Gatignon & Robertson, 1985). Although the empirical literature reports inconsistent effects of consumer innovativeness on adoption (Im, Bayus, & Mason, 2003), the meta-analysis suggests a positive influence on both intention and behavior (with a stronger effect on the latter). With respect to opinion leadership, we only found a positive effect on behavior. Research by Goldenberg, Sangman, Lehmann, and Hong (2009) suggests that “social hubs appear to adopt earlier because of their larger number of connections rather than innate innovativeness” (p. 10).