Assessing multifunctional innovation adoption via an integrative model

Not all consumers are equally equipped to categorize new products (Petty et al. 1983). Consumers’ knowledge, or familiarity and expertise, determines their ability to categorize new products and consequently guides their comprehension of the MFIs (Alba and Hutchinson 1987). In particular, high-tech products are complex and are often described in technical terms, and therefore knowledge of the relevant product category is imperative for assessing product benefits and functionality.

In learning about DFIs, consumers can possess knowledge about two product categories—the “primary” and “supplemental” categories (Gill and Dubé 2007; Gill 2008; Moreau et al. 2001). Consistent with the literature, we define the primary product category as the one that most resembles the innovation in terms of functionality. In the English language, this product category is listed as the right-most category in a category label (Gagné and Shoben 1997; Wisniewski 1997; Wisniewski and Love 1998). The supplemental category signifies how the DFI differs from other primary category products (Moreau et al. 2001; Wisniewski 1997). This product category is listed as the left-most category in a category label (Gagné and Shoben 1997; Wisniewski 1997; Wisniewski and Love 1998). For example, in the case of a DFI such as a “camera cell phone,” “cell phone” is the primary category and “camera” is the supplemental category that signals to consumers this cell phone can also be used as a camera and is, therefore, distinct from ordinary cell phones.

Experts in both primary and supplemental categories possess more knowledge than do novices (Alba and Hutchinson 1987). That is, experts have more elaborate knowledge structures and richer product attribute representations and connections. Thus, they are more likely than novices are to realize the DFI’s dual functionality and correctly categorize the product as such. They see the discrepancy between the DFI and other category products and can correctly see what separates the DFI from others. Novices, on the other hand, are likely to have trouble visualizing a DFI’s dual functionality since their knowledge structures are impoverished. They likely consider the DFI as another category product with added features. The two different types of knowledge are complementary; consumers integrate their knowledge from primary and supplemental categories to reach a DF categorization (Gill and Dubé 2007). Primary category knowledge helps consumers see the improvement in product functionality in the product class and how the product relates to other products in that category. Supplemental category knowledge helps consumers see how a DF product is different from other primary category products (Wisniewski 1996). Thus, the two types of knowledge help consumers in different ways to reach DF categorization.

So far, we have assumed that adequate resources are available to categorize a DF product. When substantial resources are not allocated to the categorization process, even experts may have difficulty in seeing the DF product as having two functionalities. This assertion follows from the earlier observation that correctly categorizing a DF product can be relatively effortful.

Two types of mental resources, i.e., need for cognition and product involvement, have been shown to enhance correct categorization of a new product. Following Cacioppo and Petty (1982), need for cognition (NFC) is “the difference among individuals in their tendency to engage in and enjoy thinking.” The effect of knowledge on MFI categorization is likely to be influenced by NFC. Numerous studies have demonstrated that individuals with a high NFC enjoy elaborating on new product information and consequently engage in detailed processing since it satisfies their goals (e.g., Priluck and Till 2004; Tidwell et al. 2000). Instead of being influenced by peripheral cues such as endorser attractiveness (Haugtvedt et al. 1992), spokesperson credibility (Petty and Cacioppo 1986), or the number of arguments presented (Cacioppo et al. 1983), they tend to be influenced by message-relevant thoughts (Mantel and Kardes 1999). This implies that individuals with a high NFC make careful and well thought-out judgments (Mantel and Kardes 1999).

Within the context of DF products, this would mean that high NFC experts in the primary category pay attention to the product and its functionalities and are thereby more likely to realize its DF purpose. As NFC increases attention to details, high NFC experts in the primary category should be more likely to realize the DF product’s dual functionality. Experts in the supplemental category are also aided by NFC. As the supplemental category helps consumers see how the DF product is different from other primary category products (Moreau et al. 2001; Wisniewski 1996), they will elaborate on these differences to a greater extent. In contrast, consumers low in NFC tend to focus on readily available details of the DF product and are thereby unlikely to grasp its dual functionality. Consumers low in NFC tend to categorize products holistically rather than analytically, and they thereby focus on the product’s appearance (rather than functionality) in categorization (Creusen and Schoormans 2005). MF products resemble products from the primary category in their appearance, and they encourage consumers low in NFC to consider the MF product to be an SF product (Gregan-Paxton et al. 2005). Since low NFC consumers process neither the primary nor supplemental category in depth, this effect is likely to hold across both categories. Thus, we posit:

Product involvement is the degree to which consumers perceive the personal relevance (Zaichkowsky 1985) or importance of a product category (Mantel and Kardes 1999). Higher levels of involvement have been shown to lead to greater processing of information (e.g., Celsi and Olson 1988; Petty et al. 1983; Richins and Bloch 1986). Highly involved individuals spend more time considering and evaluating accessible attributes during product evaluation (Celsi and Olson 1988).

For high-involvement consumers, knowledge is likely to be positively related to MF categorization. In contrast, if knowledge of the product categories is lacking, the correct categorization of MF products is less likely. High involvement only increases consideration and deliberation of a product if product-related attributes are already accessible from memory (Mantel and Kardes 1999). If high-involvement consumers do not have sufficient knowledge to categorize the new product, the advantages gained from involvement are likely to be lost (Mantel and Kardes 1999). As there are important product-related attributes in the primary and supplemental categories (Wisniewski 1997), we expect this effect to hold for both categories.

Consumers who are involved with a product category have been found to elaborate on products to a broader extent and to a greater degree and to reach more accurate conclusions regarding products than low-involvement consumers (Zaichkowsky 1985). Thus, they should be more likely to be able to correctly categorize a DFI as a dual-functionality product than their low-involvement counterparts. In contrast, for low-involvement consumers, processing a DF product is a taxing process in which they have no incentive to engage. Instead, they tend to rely on categorization cues that are easily visible, such as the product’s appearance (Creusen and Schoormans 2005). As DF products tend to look like primary category products, such a categorization strategy is likely to lead consumers to categorize the product as an SF product (Gregan-Paxton et al. 2005). Equal levels of effort are required to process both the primary and supplemental category (Gagné and Shoben 1997); thus, these effects should be visible in both categories.

When faced with an MF product, consumers have the opportunity to categorize it as either an SF or MF product. As compared to SF products, MFIs can contribute to the attainment of multiple goals and are thus likely to be preferred to those that only contribute to a single goal if consumers realize the product’s dual functionality (Gill 2008). Multiple goals are preferred to a single goal if consumers view a DF product as having a dual functionality since it evokes goal complementarity (Fishbach and Ferguson 2007; Fishbach and Dhar 2007; Fishbach and Zhang 2008). Complementary goals contribute to a positive product evaluation by suggesting synergistic effects across the goals and highlighting the new (emergent) features of the DF product (Fishbach and Zhang 2008; Kruglanski et al. 2002).³ DF products have new features that emerge from the combination of the products (Estes and Ward 2002; Wilkenfeld and Ward 2001), such as multimedia messaging in camera cell phones (which emerges from having a camera and a cell phone in the same package), and highlighting these features increases positive evaluations of the DF product. If consumers see the DF product as having only a single functionality, they are likely to form a more negative view of the product. Missing out on the product’s dual functionality is likely to lead to its evaluation as an SF product. As an SF product, MF features are not likely to be that appealing, since the SF counterparts of MFIs are often more powerful and perform better (Han et al. 2009). In this case, consumers are likely to be confused and frustrated about the poor performance of the MF product and find it less attractive. However, recognition of a DFI’s dual functionality is likely to lead consumers to accept the lower functionality for the sake of a specific benefit, such as added convenience (Han et al. 2009). For the specific benefit denoted for the DFI, the new attributes that emerge from the combination can come in handy, such as using multi-media messaging with a camera cell phone on-the-go. However, these advantages of a DFI are apparent only if consumers categorize it as such. Given that an SF categorization of a DFI tends to receive lower evaluations (as consumers compare the DFI to SF products) whereas a DFI product categorized as a dual-functionality product tends to receive higher evaluations (as it supports the attainment of multiple goals and has unique advantages as compared to other products), a positive relationship between DFI categorization (as compared to an SF categorization of the same product) and product evaluation is expected.

Consumer evaluations of MFIs are influenced by the three attributes of relative advantage, complexity, and compatibility (e.g., Bauer 1960; Rogers 1962, 2003; Tornatzky and Klein 1982). Innovation characteristics initially included two other characteristics: observability and trialability. However, a meta-analysis by Tornatzky and Klein (1982) demonstrated that the effects of these characteristics on adoption were weak. Thus, observability and trialability have not been used in subsequent studies (e.g., Moreau et al. 2001). Two other factors have also emerged as impacting innovation adoption propensity and are often used in conjunction with innovation characteristics: perceived risk (Bauer 1960) and product newness (Robertson 1971). Numerous studies demonstrate the importance of these factors in predicting innovation adoption (Hoeffler 2003; Moreau et al. 2001). The literature has also established that the compatibility and relative advantage dimensions overlap; that is, the more compatible a product, the greater its advantages that are linked to the compatibility of the product (Meuter et al. 2005; Taylor and Todd 1995), as “there can be no advantage to an innovation that does not reflect an adopter’s needs” (Moore and Benbasat 1991). Compatibility and complexity have also been suggested to overlap as they both measure learning requirements (Gatignon and Robertson 1985; Moreau et al. 2001). Given this overlap, we capture MFI compatibility through its relative advantage and complexity (cf. Moreau et al. 2001).

The literature on innovation adoption concludes that greater likelihood of product acceptance is positively associated with higher perceived advantage and product newness,⁴ and inversely associated with perceived risk and product complexity (Dickerson and Gentry 1983; Gatignon and Robertson 1985; Holak 1988; Moreau et al. 2001; Rogers 2003; Taylor and Todd 1995). We incorporate advantages of a DFI and its associated risks in our model via net benefits (advantages—risks). The marketing literature is relatively consistent with respect to the effectiveness of net benefits in predicting adoption of innovations (e.g., Moreau et al. 2001; Wood and Moreau 2006). Taken together, we anticipate that net benefits and newness will have a positive effect and complexity will have a negative effect on product evaluation. In addition to these measures, Rogers (1962, 2003) points out that the decision to adopt or reject an innovation is influenced by the consumer’s evaluation of the product. Thus, we propose the following hypotheses:

Motivation pertains to the readiness and intention of consumers to interpret new products (MacInnis et al. 1991). The interpretation of technological products is dependent on two important factors: the degree of technological anxiety consumers experience when faced with such products and how playful they feel when interacting with them.

Technological anxiety is the tendency of individuals to be uneasy, apprehensive, or fearful about using technological products (Igbaria and Parasuraman 1989). Technologically anxious consumers demonstrate a lower propensity to interact with complex products, and therefore complexity is likely to be negatively related to the evaluation of DFIs (see Meuter et al. 2003). Technologically sophisticated consumers, on the other hand, are more willing to embrace the more complex DF products. These products tend to be more advanced and perform a wider variety of functions than basic models, may offer customization possibilities and, since not all consumers know how to effectively use technologically sophisticated devices, they can even boost a consumer’s ego (Holak 1988; Holak and Lehmann 1990). As such, we expect:

Consumer reaction to the newness of products is also influenced by technological anxiety (Meuter et al. 2005). MFIs constitute new products, which have been found to heighten uncertainty (Hoeffler 2003). Technologically anxious consumers experience difficulty managing such uncertainties. Anxious consumers experience a negative emotional reaction toward new products that evokes fear and confusion (Venkatesh 2000). The situation escalates given the ambiguous nature of MFIs. In contrast, new products represent a learning opportunity for technologically sophisticated consumers. Technologically confident consumers are likely to enjoy interacting with new products and learning about the benefits they provide. As such, we expect:

Playfulness is an individual’s tendency to interact spontaneously, inventively, and imaginatively with technological products (Webster and Martocchio 1992). It involves a desire for fun, exploration, discovery, challenge, and curiosity (Venkatesh 2000). Playful consumers interact with products for the fun of it and without concern for the utilitarian benefit derived from such interaction (Holbrook et al. 1984). These consumers can imagine using and interacting with MFIs and enjoy products that provide opportunities to use the product in varied ways.

Playfulness is likely to alter consumer perceptions toward product complexity. Although product complexity is traditionally perceived as harmful (Rogers 1962), we propose that playful consumers are likely to enjoy it. Complex products have many features, and playful consumers like to explore new products and discover new ways of using the products. In interacting with products, playful consumers perceive that they exert little effort (Moon and Kim 2001; Venkatesh 2000) and find increased enjoyment in such interactions, even in the case of complex devices (Webster and Martocchio 1993). As such, playful consumers are likely to find product complexity alluring, which enhances their product evaluations. However, for consumers scoring low in playfulness, the evaluation of a product is likely to be weakened by product complexity. These consumers do not enjoy interacting with products and, in particular, complex devices demand heightened effort to do so (Fang et al. 2005–2006). In this case, product complexity is likely to have a negative effect on product evaluation, because less playful consumers do not enjoy discovering new ways of learning about a new product and, as such, product complexity poses a hurdle for adoption. Thus, we posit:

Newness is arousing and prompts product elaboration (Mandler 1982). Furthermore, MFIs are more stimulating when they are moderately new rather than slightly new (Meyers-Levy and Tybout 1989). Playful consumers enjoy elaborating about new products and find the arousal resulting from newness to be stimulating and fun. MFIs are new and require effort to be successfully assessed. Accordingly, in our conceptual framing playfulness interacts with product newness. For playful consumers, resolving the product is likely to be perceived as a fun challenge that they happily engage in as it satisfies their curiosity and provides them with mental stimulation. In contrast, for less playful consumers, newness is likely to be negatively perceived. Assessing a product’s newness demands effort and less playful consumers who do not enjoy interacting with new products are likely to feel overwhelmed by them.

The participants for this investigation were members of a consumer panel consisting of 1,600 business professionals. A random sample of 400 panel members who owned a cell phone was selected, and 250 participated in the study (a response rate of 62.5%). A total of 330 product impressions were obtained from respondents, each of whom evaluated one to three multifunctional products and completed a questionnaire. The participants were told that the researcher was interested in their impressions of new products. The questionnaire was sent to the respondents by postal mail. We found no evidence for non-response bias based on age, gender, education, or income (p < .01). Characteristics of the sample are shown in Table 2.

We use a high-technology product context in our study as recommended in the literature (Gill 2008; Gregan-Paxton et al. 2005). The following DF products were selected as stimuli for this study: camera cell phone, cell phone PDA, PDA watch, and cell phone Internet (a service). We initially conceived ten product announcement formats using publicly available information from business and trade publications and the Internet, coupled with a series of personal interviews with experts familiar with product development. Following Hirschman and Holbrook (1982), a combination of pictorial and verbal stimuli was used to accurately capture both cognitive and affective consumer responses to new products. To refine and select the stimuli for the main study, we extensively pre-tested our stimuli in two stages.

First, following recommendations from Roberts (2000) and von Hippel (1986), experts from each of the product categories were contacted and asked to help in refining and choosing the most appropriate product announcements (out of the ten initially developed). Next, our experts assisted with the development of the announcements. Our guidelines for the experts specified that the announcements were meant to be realistic and believable in every respect, including being up-to-date with market developments. Five experts in total (one from each product category) assisted in the development and selection of the product announcements that were used in the study. Product category experts also lent assistance in choosing appropriate attributes for presenting each product. Second, product announcements were further pre-tested with 20 consumers for novelty, product benefits, relevancy, and compatibility for the relevant target groups. All of the products were considered novel (M = 4.9), relevant (M = 6.2) and compatible (M = 5.5) and provided multiple benefits for the relevant target group (M = 3 benefits). The categorizations of the products were in line with those of other studies, with most of the pre-test sample considering our DF products to possess a single functionality (60%) but with a significant minority (40%) considering the DF products to have dual functionality. Based on these pre-tests, we considered the product announcements to be suitable for our study. All pre-test subjects were eventually asked about the dual nature of their respective stimuli. Based on their responses, it is evident that once the nature of the stimuli were disclosed, subjects fully comprehended the dual-functionality of the products, further affirming the appropriateness and complementary nature of the product concepts used to develop each stimulus.

Print announcements for all four products were professionally prepared and shown to the participants in a random fashion to avoid order effects. Each announcement featured the name of the product (in bold print) and a color picture. Below this, nine product properties described the product, three of which were in the primary category, three of which were in the supplemental category, and three of which were shared product properties. Moreover, the body of the copy briefly described the product (see Appendix 1).

Ability is represented by expertise, familiarity, need for cognition, and product involvement. Expertise was assessed on a ten-item scale measuring factual knowledge about all five product groups involved in the study (cameras, cell phones, PDAs, the Internet, and watches) using a multiple choice/fill in the blanks test (Moreau et al. 2001, Study 2).⁵ Objective measures of familiarity better correspond to the content of consumer knowledge structures (Moreau et al. 2001) and were thus adopted for this research. Need for cognition was measured using Cacioppo et al. (1984) revised 18-item scale. In line with other studies (Cotte and Wood 2004; Kopalle and Lindsey-Mullikin 2003; Shrum et al. 2005), we used a shorter version of the scale, comprised of six items. These items loaded on a single factor (α = .81). Product category involvement was measured along the ten-item revised personal inventory scale developed by Zaichkowsky (1994) using the five items denoted to measure cognitive involvement. These items also loaded on a single factor (α = .85).

Consistent with our conceptual framework, opportunity is assessed via categorization. Categorization for each DF product was assessed by asking consumers to state whether the product has a single or dual functionality, followed by a question on the product’s categorization. A consumer was given the value 1 if s/he considered the product to have a single functionality, and 2 if the person considered the product to have a dual functionality. In the categorization question, we gained insight into what product category the respondent thought the product would belong. For example, consumers categorizing a DF product as an SF product consider the camera cell phone as “a cell phone” or “a camera,” whereas consumers categorizing a DF product as possessing dual functionality consider it “a camera cell phone” or “a combination of a cell phone and a camera.” In the categorization, the complementary roles of the product categories were also often noted, such as a camera cell phone for taking pictures on-the-go, and a PDA watch for keeping up with one’s calendar when in between meetings. A member of the research team and a naïve judge (uninformed about our hypotheses) coded the responses based on single vs. dual functionality. The inter-rater reliability (calculated via Cohen’s kappa) was 0.96. Disagreements were resolved by discussion and subsequent consensus.

Innovation evaluation measures consist of net benefits, complexity, and newness, and the adoption decision. Following Moreau et al. (2001) and Wood and Moreau (2006), net benefits were measured by asking the participants to list each product’s benefits and weaknesses. The difference between the two constitutes net benefits.⁶ Complexity was represented by two items: “complex/not complex” and “difficult/easy to use” (α = .76), and product newness was assessed in terms of “new/not at all new” and “advanced/not advanced” (α = .86). Product evaluation was measured via an eight-item, seven-point scale adapted from Roehm and Sternthal (2001). Of these, five items loaded on a single scale and were included in subsequent analysis (α = .92).

Consumer adoption decisions were measured by an open-ended question that offered respondents an opportunity to accept or reject the DFI, by asking them whether they would buy the product. The open-ended question was subsequently converted into a three-point scale by assessing consumer willingness to buy the DFI (where 1 = would not buy, 2 = purchase would be contingent upon certain conditions, e.g., if s/he could afford it, and 3 = would buy).⁷

Technological anxiety was measured along a ten-item construct, using seven-point scales adapted from Parasuraman (2000) and Meuter et al. (2003). Five of these items loaded on a single scale and were used for subsequent analysis (α = .87). We combined two existing playfulness scales by Venkatesh (2000) and Moon and Kim (2001) for the technology playfulness scale, which consists of a ten-item, seven-point scale. Of these, nine items loaded on a single factor and were included in subsequent analysis (α = .92).

We used AMOS 18.0 to test the proposed hypotheses. To avoid dependency in the data structure, the order in which the products were shown to subjects was randomized and included in our model as dummy variables. The order effects were non-significant (p > .10). We also autocorrelated the errors across the four product evaluations to determine whether there is dependency in the data. The correlation coefficients were non-significant (p > .10). Therefore, dependency across product evaluations is not influenced by stimuli order effect, and hence the data for all four DF products were pooled. DF categorization (i.e., single vs. dual functionality) was coded as a dummy variable (0 or 1). AMOS handles dummy variables using Bayesian estimation, in which the means and the intercepts of each variable are calculated using Bayes’ theorem and used as analogs to parameter estimates and standard errors (Byrne 2010).