Computers in Industry 64 (2013) 436–447 Contents lists available at SciVerse ScienceDirect Computers in Industry journal homepage: www.elsevier.com/locate/compind Sales configurator capabilities to avoid the product variety paradox: Construct development and validation Alessio Trentin *, Elisa Perin, Cipriano Forza Università di Padova, Dipartimento di Tecnica e Gestione dei sistemi ind.li, Stradella S. Nicola 3, 36100 Vicenza, Italy A R T I C L E I N F O A B S T R A C T Article history: Received 27 April 2012 Accepted 4 February 2013 Available online 7 March 2013 Sales configurators are applications designed to support potential customers in choosing, within a company’s product offer, the product solution that best fits their needs. These applications can help firms avoid the risk that offering more product variety and customization in an attempt to increase sales, paradoxically results in a loss of sales. Relatively few studies, however, have focused on the characteristics sales configurators should have so as to avoid this paradox. Furthermore, empirical investigation on the effectiveness of the recommendations made by these studies has been hindered by the lack of psychometrically sound measurement items and scales. This paper conceptualizes, develops and validates five capabilities that sales configurators should deploy in order to avoid the product variety paradox: namely, focused navigation, flexible navigation, easy comparison, benefit-cost communication, and user-friendly product-space description capabilities. It is hoped that this study will provide a parsimonious measurement instrument to advance theory testing in the field. Moreover, this instrument may be a useful diagnostic and benchmarking tool for companies seeking to assess and/or improve sales configurators they use or develop. ß 2013 Elsevier B.V. All rights reserved. Keywords: Product configuration Software capabilities Measurement development Mass customization 1. Introduction A trend toward an increase in product variety and customization has been observed worldwide in many diverse industries [1– 4]. The promise of increased product variety and customization is that by offering customers exactly what they want, or at least something closer to their ideal product solutions, companies will be able to charge higher prices and/or to gain higher market shares [5–7], thereby increasing revenues. There is a risk, however, that a strategy of product proliferation and customization backfires, leading to lower rather than greater revenues, as increasingly suggested in literature [7–13]. Potential customers, for example, may feel so confused and overwhelmed by the number of product configurations offered by a company that they choose not to make a choice at all [8] and the company loses potential sales. Firms offering product variety and customization may therefore experience what has been termed the ‘‘product variety paradox’’ [14]: offering more product variety and customization in an attempt to increase sales paradoxically results in a loss of sales. * Corresponding author. Tel.: +39 0444 998742; fax: +39 0444 998884. E-mail addresses: alessio.trentin@unipd.it, alextren@tiscalinet.it (A. Trentin), perin@gest.unipd.it (E. Perin), cipriano.forza@unipd.it (C. Forza). 0166-3615/$ – see front matter ß 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.compind.2013.02.006 This risk is exacerbated by the fact that the Web has made immense product choice and a significant amount of productrelated information potentially available to customers [15,16]. Confronted with this information explosion, customers face no easy task searching for their ideal product configurations without salespeople’s support [16]. An important role in alleviating the risk of experiencing the product variety paradox can be played by sales configurators [14,17,18]. A sales configurator is a subtype of software-based expert systems (or knowledge-based systems) with a focus on the translation of each customer’s idiosyncratic needs into complete and valid sales specifications of the product solution that best fits those needs within a company’s product offer [19,20]. The fundamental functions of a sales configurator include presenting a company’s product space, meant as the set of product solutions that a firm offers [21], and guiding customers in the generation or selection of a product variant within that space, thus preventing inconsistent or unfeasible product characteristics from being defined [18,22]. Additional functionalities of a sales configurator may include providing real-time information on price and/or delivery terms of a product variant, making quotations [23,24] and recommending a product solution that can be further altered [17]. Sales configurators may be stand-alone applications or modules of other applications, known as product configurators, which support not only translation of customer needs into sales specifications, but also translation of sales specifications into the product data A. Trentin et al. / Computers in Industry 64 (2013) 436–447 necessary to build the product variant requested by the customer, such as bill of materials, production sequence, etc. [25]. Many studies on sales configurators and, more generally, on product configurators have investigated technical or application development issues, such as the modeling of configuration knowledge or the algorithms to make configurators faster and more accurate [e.g., 26–32]. Many other studies have provided detailed accounts of the introduction and use of a configurator in a single company, focusing mainly on implementation challenges and operational performance outcomes from the company perspective [e.g., 23,24,33–36]. In this vein, large-scale hypothesis-testing studies on the effects of product configurator use on a firm’s operational performance have recently appeared as well [37,38]. Instead, less attention has been given in literature to which characteristics of sales configurators reduce the effort involved in the specification process and drive users’ satisfaction with this process [18], thereby alleviating the risk that companies experience the product variety paradox [14]. Huffman and Kahn [8], Kamis et al. [39], and Valenzuela et al. [11] find that enabling choice by product attributes, rather than among complete product alternatives, eases the customer decision process, especially as the number of attributes and their values increase. Randall et al. [40] suggest that inexperienced customers should be allowed to specify the relative importance of their needs, rather than the values of design parameters of the product, whereas expert customers should be allowed to directly specify design parameters. Dellaert and Stremersch [41] find that customer perceived effort is reduced if only the price of the configured product, and not also the price of each choice option, is presented. A broader set of recommendations is made by Randall et al. [17] and Salvador and Forza [14], such as providing an initial configuration that the customer can subsequently alter, supporting incremental refinement, and structuring customer–company interaction. However, the empirical study of how sales configurators should be designed to ease the customer decision process and to increase configuration processrelated value for the customer is still in its infancy [18,42]. To help narrow this research gap, the present paper conceptualizes, develops and validates five sales configurator capabilities that are expected to motivate and facilitate further empirical investigation in the field. Moreover, by distilling the capabilities implied by prior research recommendations, the present paper enhances understanding of when and why sales configurators generate value for customers and thus support mass customization strategies. 2. Literature review and construct definitions 2.1. The product variety paradox Prior research suggests several mechanisms that explain why a company’s strategy of product proliferation and customization might prove detrimental, rather than beneficial to the company’s revenues [13]. In particular, four inter-related mechanisms link product variety and customization to the difficulty experienced by potential customers in configuring the product solutions that best fit their needs within a company’s product space. The experience of difficulty accompanying a potential customer’s decision process may become an input to his/her evaluation of the decision outcome itself [11,13,43,44]. Consequently, greater decision difficulty for potential customers may translate into lower satisfaction with the configured products and, eventually, into reduced willingness to make a purchase [11,13]. A first explanation for the product variety paradox relies on choice complexity, defined as the amount of information processing necessary to make a decision [11]. As product variety and 437 customization increase, so too does choice complexity, since more alternatives have to be processed in order for a potential customer to make a decision based on rational optimization. The amount of information processing is a widely acknowledged source of decision difficulty [45]. If potential customers are provided with ‘‘too much’’ information at a given time, such that it exceeds their processing limits, information overload occurs [46]. Information overload, in turn, may lead potential customers to choose from competing brands that do not require such cognitive effort [7] thus reducing the company’s revenues. A related explanation for the product variety paradox relies on anticipation of post-decisional regret, which is a cognitively determined negative emotion that individuals experience when realizing or imagining that their present situation would have been better, had they acted differently [47]. When choice complexity becomes excessive, potential customers may become unable to invest the requisite time and effort in seeking the best option for them based on rational optimization and may turn from compensatory decision strategies, which process all of the available information, to non-compensatory heuristics, which reduce information processing demands by ignoring potentially relevant information [45,48,49]. Furthermore, potential customers may have uncertain preferences because of poorly developed preferences or poor insight into their preferences [49–51], so that their wants at the time of choice can have low correlations with their likes at the time of consumption [10]. When potential customers are unable to engage in rational optimization and/or have uncertain preferences, they may anticipate the possibility of post-decisional regret due to poor fit between the selected product configuration and the customer’s preferences [9,10,52]. If this is the case, potential customers take into account this possibility into their decision processes, seeking to avoid or minimize postdecisional regret [10,52]. The goal of minimizing post-decisional regret makes potential customers’ decision processes more difficult [9] and may lead them to delay their purchase decisions [9,52] or to prefer a standard product to a customized one [10]. A third related explanation for the product variety paradox relies on responsibility felt by potential customers for making a good decision. As product variety and customization increase, potential customers feel more responsible for their choices, given the greater opportunity of finding the very best option for them [9,13]. These enhanced feelings of responsibility promote anticipated regret, as subjectively important decisions, for which individuals feel more responsible, will result in more intense post-decisional regret when things go awry [47,52]. By amplifying anticipated regret and the resulting decision difficulty, responsibility for making a good decision magnifies the negative impact of choice complexity on customers’ willingness to make a purchase. Finally, a fourth mechanism relating product variety and customization to decision difficulty relies on conflict between product attributes that are linked to highly valued goals for potential customers [7,11,45,53]. To increase product variety and customization, companies need to broaden the range of product attributes on which they allow their potential customers to make a choice [54]. As the number of product-differentiation attributes increases, so too does the likelihood that potential customers have to make trade-offs among attractive attributes. This happens because offering all the possible combinations of all the different levels of all the product-differentiation attributes may be economically unfeasible, owing to insufficient manufacturing process flexibility and limited product modularity [55]. Explicit trade-offs among attractive attributes not only increase the cognitive effort required of potential customers to process all of the available information [7], but also cause potential customers to experience negative emotions such as anticipated regret [7]. This happens because trade-off resolution involves consideration of 438 A. Trentin et al. / Computers in Industry 64 (2013) 436–447 potential unwanted consequences and threatens one’s reputation of self-esteem as a decision maker [56]. The negative emotions associated with between-attribute trade-offs are another mechanism that links increased product variety and customization to greater subjective experience of choice task difficulty [11] and decreased satisfaction with the chosen product [13], thus explaining the product variety paradox. 2.2. Sales configurator capabilities to avoid the product variety paradox In the following subsections, we propose five sales configurator capabilities that help companies avoid the product variety paradox by hindering operation of at least one of the mechanisms outlined in the previous section. For each proposed capability: (i) we provide its conceptual definition as well as empirical illustration by means of one existent Web-based sales configurator, (ii) we link development of that capability with a number of recommendations made by relevant previous research and, finally (iii) we explain why that capability is expected to alleviate the risk of experiencing the product variety paradox. 2.2.1. Focused navigation capability We define focused navigation capability as the ability to quickly focus a potential customer’s search on a product space subset that contains the product configuration that best matches his/her idiosyncratic needs. An example of Web-based sales configurator deploying this capability is Toshiba’s laptop customization site, which offers potential customers a laptop finder to narrow their search down to a specific color, processor, and/or any other desired product-differentiation attribute value (Fig. B1). Indeed, a fundamental way of improving focused navigation capability is to allow potential customers to sequence their choices concerning the value of each product-differentiation attribute from the least uncertain choice to the most uncertain one [14]. This is because, according to the attribute being considered, a customer’s preferences may be more or less uncertain [50] and preference uncertainty is an antecedent of anticipated regret [10,57]. If the early choices a potential customer is required to make are those for which his/her preferences are best developed, then he/she is enabled to narrow down search more quickly, as anticipated regret associated with those choices is lower. Noteworthy, a prerequisite for this way of structuring the customer–company interaction is the by-attribute presentation of the company’s product space, meaning that the customer is asked which value he/she prefers for each product-differentiation attribute instead of being required to choose from among a set of fully specified product configurations, as happens with the byalternative presentation [8]. Another option to enhance focused navigation capability is to provide one or more starting points, where a starting point is defined as an initial product configuration that is close to the customer’s ideal solution and that may be further altered [17]. Starting points can be recommended with little or no effort on the customer’s part, based on his/her past purchases and/or customer input concerning simple demographics, intended product usage and his/her best developed preferences [30,58]. Noteworthy, this solution requires complementing the by-attribute presentation of the product space with the by-alternative presentation. The same applies to another way of improving focused navigation capability, that is to allow a potential customer to completely exclude certain product solutions from consideration if he/she does not wish for them [59]. Focused navigation capability helps to avoid the product variety paradox by reducing choice complexity and by mitigating anticipated regret. A sales configurator with this capability does not force potential customers to go through and evaluate a number of product options that they regard as certainly inappropriate for themselves. Therefore, this capability reduces the amount of information processing necessary to make a decision without potential customers experiencing anticipated regret [10,47,52,57]. Furthermore, by quickly reducing the size of the search problem, this capability enables potential customers to invest more time and effort in exploring the product options for which their preferences are less certain. Potential customers can learn more about both these options and the value they would derive from them, especially when focused navigation capability is complemented with the capabilities discussed in the subsequent sections. In addition, potential customers can rely on more time-consuming, compensatory decision strategies that enable rational resolution of between-attribute conflicts [49], if any. As a consequence, once a potential customer has selected his/her most preferred product configuration, he/she is more confident that the chosen solution is the one that best fits his/her needs within the company’s product space. Reduced uncertainty on the superior fit of the selected product configuration with the customer’s preferences, in turn, translates into less anticipated regret [52]. 2.2.2. Benefit-cost communication capability We define benefit-cost communication capability as the ability to effectively communicate the consequences of the available choice options both in terms of what the customer gets (benefits) and in terms of what the customer gives (monetary and nonmonetary costs). An example of Web-based sales configurator deploying this capability is Dell’s laptop customization site. At each step of the configuration process, this site gives potential customers the possibility to click on the ‘‘Help Me Choose’’ button, which opens up a page with a list of recommendations suggesting the advantages of every choice option (Fig. B2a). Moreover, the site communicates the price variation that selecting each of the available options would cause with respect to the price of the current configuration (Fig. B2b). Indeed, a fundamental way of improving benefit-cost communication capability is to explain what potential needs a given choice option contributes to fulfill and to what extent it does so [14]. This explanation is especially important when choice options involve design parameters of the product, such as specifications of product components, because potential customers are often unable to relate design parameters to satisfaction of user needs [17]. According to the product attribute being considered, this explanation may be more effectively provided by means of different media, including texts, photos, animations or other simulations of the real product on a computer [60]. Besides the benefits, it is also important to communicate monetary and nonmonetary costs of each option, for example by displaying the prices of the individual product components from among which potential customers can choose or by warning potential customers that certain options imply longer delivery lead-times [14]. Benefit-cost communication capability helps avoid the product variety paradox by mitigating anticipated regret. During the sales configuration process, potential customers seek to anticipate the value they will perceive from consumption of the product being configured [61]. Perceived product value is defined as the customer’s ‘‘overall assessment of the utility of a product based on perceptions of what is received and what is given’’ [14,62]. By delivering clear prepurchase feedback on the effects of the available choice options, a sales configurator with high benefitcost communication capability fosters potential customers’ learning about the value they would derive from these options [63,64]. This learning process makes a potential customer more confident that the product configuration he/she has selected is the one that best fits his/her needs within the company’s product space. Reduced uncertainty on the superior fit of the chosen A. Trentin et al. / Computers in Industry 64 (2013) 436–447 product configuration with the customer’s preferences, in turn, translates into less anticipated regret [52], thus lowering choice task difficulty [9]. At the same time, however, higher benefit-cost communication capability may lead to greater choice complexity, with negative effects on decision difficulty. For instance, individual pricing of the available choice options may make cost-benefit trade-offs more salient and, hence, may increase information processing demands [41]. To fully realize the potential advantages of benefit-cost communication capability, therefore, this capability needs to be complemented with the focused navigation one, which lowers choice complexity by quickly reducing the size of the search problem for potential customers. As a result, the learning process enabled by benefit-cost communication capability focuses only on those choice options for which potential customers’ preferences are less certain and, thus, the possible negative effects of this capability on choice complexity are mitigated. 2.2.3. Flexible navigation capability We define flexible navigation capability as the ability to minimize the effort required of a potential customer to modify a product configuration that he/she has previously created or is currently creating. An example of Web-based sales configurator deploying such capability is Converse’s site, which allows potential customers to customize sport shoes through a multiple step configuration process, where each step corresponds to one customizable feature of the product. This configuration process is depicted by a progress bar made up of as many boxes as are configuration steps. Potential customers can modify any previously selected feature by simply clicking on the related box, without losing any other choice that they have previously made (Fig. B3). Indeed, a fundamental way of improving flexible navigation capability is to allow sales configurator users to change the choice made at any previous step of the configuration process without having to start it over again [17]. Furthermore, after changing the choice made at a given step, potential customers should not be required to go through all the subsequent steps up to the current one. Instead, they should be asked to revise only those choices, if any, that are no longer valid because of the change they have just made [65]. Another option to enhance flexible navigation capability is to allow potential customers engaged in configuring their products to bookmark their work [17]. Bookmarks enable potential customers who are exploring alternative product configurations to immediately recover a previous configuration in the case that they decide to reject the newly created one. Flexible navigation capability helps to avoid the product variety paradox by mitigating anticipated regret. A sales configurator with this capability enables potential customers to quickly make and undo changes to previously created product configurations. Consequently, the number of product solutions a potential customer can explore in the time span he/she is willing to devote to the sales configuration task is larger. Stated otherwise, potential customers can conduct more trial-and-error tests to evaluate the effects of initial choices made and to improve upon them. Trialand-error experimentation promotes potential customers’ learning about the value they would derive from consumption of the product being configured [63,64], especially when flexible navigation capability is complemented with the benefit-cost communication one as well as those discussed in the subsequent sections. This learning process makes potential customers more confident that the product configuration they have selected is the one that best fits their needs within the company’s product space. Reduced uncertainty on the superior fit of the chosen product configuration with the customer’s preferences, in turn, translates into less anticipated regret [52]. 439 2.2.4. Easy comparison capability We define easy comparison capability as the ability to minimize the effort required of a potential customer to compare previously created product configurations. An example of Web-based sales configurator deploying this capability is K SWISS’s site, which enables potential customers to save configured shoes in their ‘‘My Account’’ areas and, subsequently, access them at any time (Fig. B4). Once logged in, the site user finds all his/her previously saved shoes portrayed on the same Webpage, so that they can be easily compared. Indeed, a fundamental way of improving easy comparison capability is to allow potential customers to save a product configuration they have just created and, then, to compare previously saved configurations side-by-side in the same screen [17]. The advantages of providing an overview of previous configurations can be enhanced by highlighting commonalities and differences among them, especially if the sales configuration process involves many choices. In this manner, a potential customer can immediately understand, for example, which configuration choices have caused the price or weight difference between two configurations he/she is comparing. Another solution to enhance easy comparison capability is to rank-order previously created configurations in terms of fit to the customer’s preferences or profile [50]. This can be accomplished with little or no effort on the customer’s part, based on his/her past purchases and/or customer input concerning simple demographics, intended product usage and his/her best developed preferences [30,58]. Easy comparison capability helps avoid the product variety paradox by reducing choice complexity and by mitigating anticipated regret. A sales configurator with this capability fosters potential customers’ learning about the value they would derive from consumption of the product being configured. This happens because, in assessing the value of a particular product solution, customers tend to rely on comparisons with other alternatives that are currently available or that have been encountered in the past [50,66]. In particular, the possibility of easily comparing complete product configurations is of greatest assistance when global performance characteristics, which arise from the physical properties of most if not all of the product components [55], are important to potential customers. In brief, easy comparison capability gives potential customers practice at evaluating alternative configurations and provides anchors for the evaluative process [8]. Consequently, potential customers improve their confidence that the configuration they have eventually selected is the one that best fits their needs within the company’s product space. In turn, reduced uncertainty on the superior fit of the chosen product configuration with the customer’s preferences translates into less anticipated regret [52]. A sales configurator with high easy comparison capability also alleviates choice complexity, by reducing information processing necessary to make comparisons. Potential customers do not need to rely on their limited working memory to recover configurations they have previously created. Moreover, potential customers do not need to rely on their limited computational abilities to decompose the configurations they want to compare to find out similarities and differences among them. 2.2.5. User-friendly product-space description capability We define user-friendly product-space description capability as the ability to adapt the product space description to the needs and abilities of different potential customers, as well as to different contexts of use. An example of Web-based sales configurator deploying such capability is Volkswagen’s site, which allows potential customers to customize Polo utility cars. For each available choice option, users are allowed to opt for a brief description or a detailed one with more technical information by 440 A. Trentin et al. / Computers in Industry 64 (2013) 436–447 selecting the ‘‘More Info’’ button (Fig. B5a). In addition, choice options affecting the esthetics of the car are described using both text and product images, with the latter changing automatically as the potential customer selects different options (Fig. B5b). Indeed, one way of improving user-friendly product-space description capability is to employ content adaptation techniques (cf. [67]) to provide optional detailed information pertaining to the available choice options. In this manner, potential customers with higher involvement for the product, who are more interested in acquiring product information [68], are allowed to learn more about the choice options for which their preferences are less developed. Conversely, customers with lower involvement, who feel less responsible for making a good decision [52], are not forced to process product information they are not interested in. In this respect, a promising approach is to design multimedia-based interfaces that enable potential customers to retrieve rich information and explanations about specific product parts/ features while looking at an illustration of the product and without breaking the continuity of their product evaluation processes [69]. Another option to enhance user-friendly product-space description capability is to adapt information content presented to potential customers according to their prior knowledge about the product [17,59]. Particularly, novice customers should be allowed to use a needs-based interface, where the available choice options involve desired product performance and functions, while expert customers should be enabled to employ a parameter-based interface, where the available choice options include design parameters such as specifications of product components [14,17]. User-friendly product-space description capability can also be improved by presenting the same information content by means of different media, so as to streamline human–computer interaction based on potential customers’ characteristics, such as cognitive abilities, age, motivation, cultural background, etc. [70,71]. User-friendly product-space description capability helps avoid the product variety paradox by reducing choice complexity and by mitigating anticipated regret. A sales configurator deploying this capability provides potential customers with the information content they value most according to their individual characteristics or usage contexts and does not bother users with communications they do not need [59]. In addition, a sales configurator with this capability augments or switches modalities of presentation of the same information content in such a way that each individual user’s information processing is enhanced [72]. By tailoring both information content and information format, this capability reduces information overload and eases the customer decision process [73–75]. In particular, this capability allows for aligning the way in which the product space is presented to a potential customer with the way in which he/she is able or willing to express his/her requirements [63,64]. As potential customers interact with a sales configurator in their customary language, they become able to assess the fit of the configured product with their needs more easily and in less time [40]. This means that, once a potential customer has selected his/her most preferred product configuration, he/she is more confident that the chosen solution is the one that best fits his/her needs within the company’s product space. Reduced uncertainty on the superior fit of the selected product configuration with the customer’s preferences, in turn, translates into less anticipated regret [52]. 3. Instrument development and validation Consistent with prior studies [e.g., 76–78], our instrument development and validation is based upon a two-stage approach. In the first phase, the measurement items for each construct are generated and tentative indications of reliability and validity are provided. In the second phase, we further refine and validate our multi-item scales using large-scale sample data and confirmatory analyses to derive stronger assessments of the psychometric properties of our measures. 3.1. Instrument development and preliminary assessment The items for the five sales configurator capabilities were generated based upon the relevant literature and extensive interviews with practitioners involved with the development and use of sales configurators. All the items were measured by means of a 7-point Likert scale (7 = completely agree, . . ., 1 = completely disagree). Additionally, we used only positive statements, as negatively worded questions with an agree– disagree response format are often cognitively complex [79] and may be a source of method bias [80]. The original set of items was vetted through three steps in order to remove potential for measurement error from the new scales. First, the items were reviewed by a group of six people with different experiences and perceptions relative to sales configuration, who were questioned about the appropriateness and completeness of the instrument. Second, to replicate as closely as possible data collection procedures to be used in our large-scale study, we pretested the instrument with 20 engineering students from our university, who were asked to comment on any problems encountered while responding, such as interpretation difficulties, faulty instructions, typos, item redundancies, etc. Based on the feedback from the focus group and field pretesting, redundant and ambiguous items were either modified or eliminated. Finally, the resulting instrument was evaluated through a Q-sort procedure for establishing tentative indications of construct validity and reliability [77,81]. Each of 10 practitioners who are experienced in developing or using sales configurators was given a questionnaire containing short descriptions of the proposed capabilities, together with a randomized list of the items. Subsequently, these expert judges were asked to assign each item to one or none of the defined capabilities. All the items were placed in the target construct by at least 75% of the judges and, therefore, were retained for our large-scale study [61]. The final instrument consisted of 17 items and is reported in Appendix A. 3.2. Large-scale refinement and validation Each of the proposed sales configurator capabilities indicates a fundamental benefit potential customers should experience and perceive during the sales configuration process if the product variety paradox is to be avoided, regardless of how such benefits are delivered. This is consistent with the capability perspective of routines, which tends to treat routines as a ‘‘black box’’ and is mainly interested in the purpose or motivation for routines [82]. Notice that sales configurators embody rules and procedures to generate or select the product variant that best fits each customer’s idiosyncratic needs within a company’s product space and, as such, are repositories of organizational procedural knowledge: namely, they are software-embedded routines [83]. Accordingly, to measure the proposed sales configurator capabilities, we needed to collect data on sales configurations experiences made by potential customers using sales configurators. Specifically, data for our large-scale study were gathered on a sample of 630 sales configuration experiences made by 63 engineering students at the authors’ university (age range: 24– 27; 29% females) using Web-based sales configurators for consumer goods relevant to the participants. As a result, our data are biased in favor of young, male, and fairly adept persons who are familiar with the Internet. At the same time, however, young people adept at using Internet also represent the majority of 441 A. Trentin et al. / Computers in Industry 64 (2013) 436–447 Table 1 Discriminant validity. Square root of AVE (%) Benefit-cost communication capability Easy comparison capability User-friendly product-space description capability Flexible navigation capability Focused navigation capability *** Correlations Benefit-cost communication capability Easy comparison capability User-friendly product-space description capability Flexible navigation capability Focused navigation capability 0.854 0.853 0.861 – 0.237*** 0.727*** 0.237*** – 0.358*** 0.727*** 0.358*** – 0.338*** 0.356*** 0.486*** 0.760*** 0.334*** 0.772*** 0.822 0.834 0.338*** 0.760*** 0.356*** 0.334*** 0.486*** 0.772*** – 0.393*** 0.393*** – Significant at p < 0.001. business-to-consumer sales configurator users [42,84]. Each participant was asked to configure a product according to his/ her individual preferences on 10 different, preassigned, Web-based sales configurators and to fill out a questionnaire for each of these configuration experiences. We decided to control for possible effects of participants’ characteristics before assessing the psychometric properties of our measurement scales. Consequently, consistent with prior studies [38,85], we regressed our 17 indicators on 63 dummies representing the participants in our study and used the standardized residuals from this linear, ordinary least square regression model as our data in all the subsequent analyses. Prior to conducting confirmatory analyses, we also evaluated common method bias, which is often mentioned as a major source of concern in studies involving self-report measures [86]. Our analyses suggest that any common method bias that exists in our study is unlikely to be problematic. First, we conducted Harman’s single-factor test, using confirmatory factor analysis (CFA) to test the hypothesis that a single factor can account for all of the variance in our data [87], and we found a very poor fit of the model testing such hypothesis with our data (RMSEA (90% CI) = 0.194 (0.188; 0.200), x2/df (df) = 24.57 (119)). Furthermore, we controlled for the effects of a single, unmeasured latent method factor [87], and found that, on average, only 7.55% of the variance of our indicators is due to the method factor [cf. 88,89]. Subsequently, CFA was employed to assess unidimensionality, convergent validity, discriminant validity, and reliability of our measurement scales. We used a variance-covariance matrix of the 17 indicators to input data, maximum likelihood method to estimate the model, and LISREL 8.80 to conduct the analysis. Unidimensionality and convergent validity were assessed by estimating an a priori measurement model, where the empirical indicators were restricted to load on the latent factor they were intended to measure. The model showed a good fit to the data: RMSEA (90% CI) = 0.047 (0.040; 0.054), x2/df (df) = 2.39 (109), CFI = 0.991, NFI = 0.984. Furthermore, inspection of the standardized factor loadings (see Appendix A) indicated that each was in its anticipated direction (i.e., positive correspondences between latent constructs and their posited indicators), was greater than 0.50, and was statistically significant at p < 0.001. Altogether, these results suggested that every item was significantly associated with one and only one latent factor and that, for each scale, all items in the scale were convergent [78,90–92]. Discriminant validity was tested using [93] procedure. For each latent construct, the square root of the average variance extracted (AVE) exceeded the correlation with each of the other latent variables (see Table 1), thereby suggesting that our measurement scales represent distinct latent variables [93]. Reliability was assessed using both AVE and the Werts, Linn, and Joreskog (WLJ) composite reliability method [94]. All the WLJ composite reliability values were greater than 0.70 and all the AVE scores exceeded 0.50, indicating that a large amount of the variance is captured by each latent construct rather than due to measurement error [93,95]. Finally, we examined the predictive validity of our constructs by investigating whether they exhibit relationships with other constructs in accordance with theory [76]. Our proposed sales configurator capabilities are posited to help firms avoid the risk that offering more product variety and customization to increase sales, paradoxically results in a loss of sales. Accordingly, these capabilities are hypothesized to positively influence both choice satisfaction and purchase intention. In the same way as Valenzuela et al. [11], we measured choice satisfaction as follows: ‘‘How satisfied or dissatisfied are you with the product that you have customized? (seven-point scale: very dissatisfied/very satisfied)’’. Following Schlosser et al. [96], we measured purchase intention by means of three items, each rating the same statement (‘‘If I needed this type of product, I do think I would buy the product that I have just configured’’) on a different seven-point scale (‘‘Unlikely/likely’’ ( 3 to +3), ‘‘Impossible/possible’’ ( 3 to +3), ‘‘Improbable/ Table 2 Predictive validity. Choice satisfaction Path coefficient Benefit-cost communication capability Easy comparison capability User-friendly product-space description capability Flexible navigation capability Focused navigation capability *** ** * y Significant at p < 0.001. Significant at p < 0.01. Significant at p < 0.05. Significant at p < 0.10. 0.277 0.071 0.133 0.101 0.302 Purchase intention t-Value *** 4.322 2.198* 1.972* 3.048** 4.631*** Path coefficient t-Value 0.137 0.103 0.131 0.095 0.375 2.189* 2.941** 1.904y 2.590** 5.267*** 442 A. Trentin et al. / Computers in Industry 64 (2013) 436–447 probable’’ ( 3 to +3), respectively). This measure of purchase intention proved to be both reliable and valid (WLJ composite reliability: 0.972; AVE = 0.920; square root of AVE exceeded the correlation with each of the other latent variables). The structural model testing the hypotheses that the proposed sales configurator capabilities positively influence both choice satisfaction and purchase intention, showed a good fit to the data: RMSEA (90% CI) = 0.0432 (0.0372; 0.0493), x2/df (df) = 2.18 (169), CFI = 0.993, NFI = 0.987. All the path coefficients are positive and statistically significant (see Table 2), indicating that each of the five sales configurator capabilities has a significant positive effect on both choice satisfaction and purchase intention and thus establishing the predictive validity of our constructs. 4. Discussion, managerial implications and future research directions In the last few years, literature has repeatedly warned against the risk that companies offering more product variety and customization in an attempt to increase sales, paradoxically suffer from a loss of sales [7,10–13]. Sales configurators can play an important role in avoiding this product variety paradox, yet relatively few studies have focused on the characteristics these applications should have so as to yield this benefit [18]. Drawing upon prior research concerning sales configurators and the customer decision process, the present paper conceptualizes five capabilities that sales configurators should deploy in order to help avoid the product variety paradox: namely, focused navigation, flexible navigation, easy comparison, benefit-cost communication, and user-friendly product-space description capabilities. Overall, these capabilities support personalization of the sales configuration experience according to each individual user’s characteristics and context of usage. Benefit-cost communication capability combined with user-friendly product-space description capability supports personalization on the content and presentation levels [cf. 97], while focused navigation, flexible navigation, and easy comparison capabilities support personalization on the interaction level [cf. 97]. Personalization of the sales configuration experience is essential to build successful sales configurators, which improve fit between selected product configuration and customer needs while limiting search effort [cf. 97,98]. The ultimate goal would be to simulate the adaptive and heuristic behavior that makes salespeople effective and aids in improving both the shopping experience and the final product choice [99,100]. By conceptualizing the abovementioned sales configurator capabilities, our paper also enhances understanding of how information technology may help achieve mass customization [e.g., 19,101–103]. As recently acknowledged in literature [42,61,84,104,105], mass customization involves not only improving compatibility between product customization and the firm’s operational performance, but also augmenting the customer perceived value of both the customized product and customization experience. Both product utility and customer satisfaction with the customization experience are negatively affected by the complexity and effort involved in the sales configuration process [8,41,104]. The present paper provides additional insight into the mechanisms through which sales configurators may prevent these negative effects: namely, reduction of choice complexity and mitigation of anticipated regret through user-friendly description of the company’s product space, focused and flexible navigation of this space, easy comparison among alternative points of this space, and effective communication of the benefits and costs of the available choice options. By illuminating these mechanisms, the present paper enhances understanding of when and why sales configurators generate value for customers and, therefore, support mass customization strategies. Consistent with the capability perspective of routines [82], the proposed capabilities indicate five fundamental benefits that sales configurator users should experience and perceive during the customization experience, regardless of how such benefits are delivered. While the solutions to develop the proposed capabilities are likely to be context-specific and subject to variation due to technological changes, the proposed set of capabilities is expected to provide more general indications to companies selecting or building a sales configurator. Developing and/or implementing such a system involves costs and, for these costs to make economic sense, the system must yield benefits, otherwise spectacular failures are possible [42]. The capabilities proposed in this paper point to five fundamental requirements that any sales configurator should meet, regardless of the specific design solutions adopted, in order to ease customers’ decision processes and, consequently, generate value for them. Another contribution of this study is the development and validation of an instrument to measure the proposed set of capabilities. The instrument was rigorously tested for content validity, unidimensionality, convergent validity, discriminant validity, predictive validity, and reliability. In particular, we found that each of the proposed capabilities significantly predicts both choice satisfaction and purchase intention, in accord with the theoretical argument that these capabilities help avoid the product variety paradox. Admittedly, our large-scale validation study involved hypothetical rather than real purchase experiences, only focused on sales configurators for consumer goods, and used students as subjects for research. Therefore, future studies should strengthen the proposed instrument through a series of further refinements and tests across different populations and settings, including truly representative samples of potential customers, sales configurators for industrial goods, etc. Though conscious that development of a measurement instrument is an ongoing process [106], we believe our instrument will be a useful diagnostic and benchmarking tool for companies seeking to assess and/or improve their sales configurators. Further, we believe the instrument developed in this paper will be of use to researchers not only as a basis for refinement and extension, but also directly. Future studies could develop and test hypotheses linking the proposed capabilities to the various dimensions of the value of customization that have been discussed in literature [42,61,84,104,105]. In particular, further research is needed to empirically investigate complementarities among the proposed capabilities, meaning that the effects of one capability on the customer perceived value of customization is reinforced by another capability, as our paper suggests. Besides the direct and joint effects of the proposed capabilities on the value of customization, future empirical studies could also examine the antecedents of these capabilities. Experimental studies, for instance, could be designed to understand how the proposed capabilities are influenced by specific technical solutions such as the use of virtual reality, which is gaining increasing utility for a variety of applications in product development [107] and could therefore aid customers in designing their own products through Web-based sales configurators. 443 A. Trentin et al. / Computers in Industry 64 (2013) 436–447 Appendix A. Measurement instrument Benefit-cost communication capability (WLJ composite reliability: 0.890; AVE: 0.729) Thanks to this system, I understood how the various choice options influence the value that this BCC1 product has for me Thanks to this system, I realized the advantages and drawbacks of each of the options I had to BCC2 choose from This system made me exactly understand what value the product I was configuring had for me BCC3 Easy comparison capability (WLJ composite reliability: 0.913; AVE: 0.727) The system enables easy comparison of product configurations previously created by the user EC1 EC2 The system lets you easily understand what previously created configurations have in common EC3 The system enables side-by-side comparison of the details of previously saved configurations The systems lets you easily understand the differences between previously created configurations EC4 User-friendly product-space description capability (WLJ composite reliability: 0.896; AVE: 0.742) UFDC1 The system gives an adequate presentation of the choice options for when you are in a hurry, as well as when you have enough time to go into the details UFDC2 The product features are adequately presented for the user who just wants to find out about them, as well as for the user who wants to go into specific details The choice options are adequately presented for both the expert and inexpert user of the product UFDC3 Flexible navigation capability (WLJ composite reliability: 0.861; AVE: 0.675) FlexN1 The system enables you to change some of the choices you have previously made during the configuration process without having to start it over again With this system, it takes very little effort to modify the choices you have previously made during FlexN2 the configuration process Once you have completed the configuration process, this system enables you to quickly change any FlexN3 choice made during that process Focused navigation capability (WLJ composite reliability: 0.901; AVE: 0.696) FocN1 The system made me immediately understand which way to go to find what I needed The system enabled me to quickly eliminate from further consideration everything that was not FocN2 interesting to me at all The system immediately led me to what was more interesting to me FocN3 FocN4 This system quickly leads the user to those solutions that best meet his/her requirements a Standardized factor loadinga Measurement error variancea 0.888 0.211 0.828 0.314 0.844 0.288 0.879 0.938 0.679 0.891 0.227 0.120 0.538 0.207 0.842 0.291 0.918 0.157 0.821 0.326 0.730 0.468 0.872 0.239 0.856 0.267 0.785 0.752 0.384 0.435 0.879 0.911 0.228 0.169 All standardized factor loadings and measurement error variances are significant at p < 0.001. Appendix B. Illustrative sales configurators Figs. B1–B5. Fig. B1. Example of sales configurator deploying focused navigation capability. 444 A. Trentin et al. / Computers in Industry 64 (2013) 436–447 Fig. B2. (a) Example of sales configurator deploying benefit-cost communication capability. (b) Example of sales configurator deploying benefit-cost communication capability. A. Trentin et al. / Computers in Industry 64 (2013) 436–447 445 Fig. B4. Example of sales configurator deploying easy comparison capability. Fig. B3. Example of sales configurator deploying flexible navigation capability. Fig. B5. (a) Example of sales configurator with user-friendly product-space description capability. (b) Example of sales configurator with user-friendly product-space description capability. 446 A. Trentin et al. / Computers in Industry 64 (2013) 436–447 References [1] B.J.I. Pine, Mass Customization: The New Frontier in Business Competition, Harvard Business School Press, Boston, MA, 1993. [2] L.F. Scavarda, A. Reichhart, S. Hamacher, M. Holweg, Managing product variety in emerging markets, International Journal of Operations & Production Management 30 (2) (2010) 205–224. [3] M. Bils, P.J. Klenow, The acceleration in variety growth, The American Economic Review 91 (2) (2001) 274–280. [4] W.M. Cox, R. Alm, The right stuff: America’s move to mass customization, Annual Report-Federal Reserve Bank of Dallas, 1998, pp. 3–26. [5] S. Kekre, K. Srinivasan, Broader product line: a necessity to achieve success? Management Science 36 (10) (1990) 1216–1231. [6] W.L. Berry, M.C. Cooper, Manufacturing flexibility: methods for measuring the impact of product variety on performance in process industries, Journal of Operations Management 17 (2) (1999) 163–178. [7] J.T. Gourville, D. Soman, Overchoice and assortment type: when and why variety backfires, Marketing Science 24 (3) (2005) 382–395. [8] C. Huffman, B.E. Kahn, Variety for sale: mass customization or mass confusion? Journal of Retailing 74 (4) (1998) 491–513. [9] S.S. Iyengar, M.R. Lepper, When choice is demotivating: can one desire too much of a good thing? Journal of Personality and Social Psychology 79 (6) (2000) 995–1006. [10] N. Syam, P. Krishnamurthy, J.D. Hess, That’s what I thought I wanted? Miswanting and regret for a standard good in a mass-customized world, Marketing Science 27 (3) (2008) 379–397. [11] A. Valenzuela, R. Dhar, F. Zettelmeyer, Contingent response to self-customization procedures: implications for decision satisfaction and choice, Journal of Marketing Research 46 (6) (2009) 754–763. [12] X. Wan, P.T. Evers, M.E. Dresner, Too much of a good thing: the impact of product variety on operations and sales performance, Journal of Operations Management 30 (4) (2012) 316–324. [13] K. Diehl, C. Poynor, Great expectations?! Assortment size, expectations, and satisfaction, Journal of Marketing Research 47 (2) (2010) 312–322. [14] F. Salvador, C. Forza, Principles for efficient and effective sales configuration design, International Journal of Mass Customisation 2 (1–2) (2007) 114–127. [15] J. Alba, J. Lynch, B. Weitz, C. Janiszewski, R. Lutz, A. Sawyer, S. Wood, Interactive home shopping: consumer, retailer, and manufacturer incentives to participate in electronic marketplaces, Journal of Marketing 61 (3) (1997) 38–53. [16] B. Xiao, I. Benbasat, E-commerce product recommendation agents: use, characteristics, and impact, MIS Quarterly 31 (1) (2007) 137–209. [17] T. Randall, C. Terwiesch, K.T. Ulrich, Principles for user design of customized products, California Management Review 47 (4) (2005) 68–85. [18] M. Heiskala, J. Tiihonen, K.-S. Paloheimo, T. Soininen, Mass customization with configurable products and configurators: a review of benefits and challenges, in: T. Blecker, G. Friedrich (Eds.), Mass Customization Information Systems in Business, IGI Global, London, UK, 2007, pp. 1–32. [19] C. Forza, F. Salvador, Application support to product variety management, International Journal of Production Research 46 (3) (2008) 817–836. [20] A. Haug, L. Hvam, N.H. Mortensen, Definition and evaluation of product configurator development strategies, Computers in Industry 63 (5) (2012) 471–481. [21] M.M. Tseng, T.F. Piller, The Customer Centric Enterprise: Advances in Mass Customization and Personalization, Springer Verlag, Berlin, Germany, 2003. [22] N. Franke, F.T. Piller, Key research issues in user interaction with user toolkits in a mass customization system, International Journal of Technology Management 26 (5/6) (2003) 578–599. [23] J. Vanwelkenhuysen, The tender support system, Knowledge-based Systems 11 (5–6) (1998) 363–372. [24] L. Hvam, S. Pape, M.K. Nielsen, Improving the quotation process with product configuration, Computers in Industry 57 (7) (2006) 607–621. [25] C. Forza, F. Salvador, Product Information Management for Mass Customization, Palgrave Macmillan, London, UK, 2007. [26] S.M. Fohn, J.S. Liau, A.R. Greef, R.E. Young, P.J. O’Grady, Configuring computer systems through constraint-based modeling and interactive constraint satisfaction, Computers in Industry 27 (1) (1995) 3–21. [27] T. Soininen, J. Tiihonen, T. Männistö, R. Sulonen, Towards a general ontology of configuration, Artificial Intelligence for Engineering, Design, Analysis & Manufacturing 12 (4) (1998) 357–372. [28] A. Felfernig, G. Friedrich, D. Jannach, Conceptual modeling for configuration of mass-customizable products, Artificial Intelligence in Engineering 15 (2) (2001) 165–176. [29] S.K. Ong, Q. Lin, A.Y.C. Nee, Web-based configuration design system for product customization, International Journal of Production Research 44 (2) (2006) 351–382. [30] X. Luo, Y. Tu, J. Tang, C.K. Kwong, Optimizing customer’s selection for configurable product in B2C e-commerce application, Computers in Industry 59 (8) (2008) 767–776. [31] P.T. Helo, Q.L. Xu, S.J. Kyllönen, R.J. Jiao, Integrated vehicle configuration system – connecting the domains of mass customization, Computers in Industry 61 (1) (2010) 44–52. [32] G. Hong, D. Xue, Y. Tu, Rapid identification of the optimal product configuration and its parameters based on customer-centric modeling for one-of-a-kind production, Computers in Industry 61 (3) (2010) 270–279. [33] J.R. Wright, E.S. Weixelbaum, G.T. Vesonder, K.E. Brown, S.R. Palmer, J.I. Berman, H.H. Moore, A knowledge-based configurator that supports sales, engineering, and manufacturing at AT&T network systems, AI Magazine 14 (3) (1993) 69–80. [34] L. Hvam, Mass customisation in the electronics industry: based on modular products and product configuration, International Journal of Mass Customisation 1 (4) (2006) 410–426. [35] C. Forza, F. Salvador, Managing for variety in the order acquisition and fulfilment process: the contribution of product configuration systems, International Journal of Production Economics 76 (1) (2002) 87–98. [36] C. Forza, F. Salvador, Product configuration and inter-firm co-ordination: an innovative solution from a small manufacturing enterprise, Computers in Industry 49 (1) (2002) 37–46. [37] A. Trentin, E. Perin, C. Forza, Overcoming the customization–responsiveness squeeze by using product configurators: beyond anecdotal evidence, Computers in Industry 62 (3) (2011) 260–268. [38] A. Trentin, C. Forza, E. Perin, Organisation design strategies for mass customisation: an information-processing-view perspective, International Journal of Production Research 50 (14) (2012) 3860–3877. [39] A. Kamis, M. Koufaris, T. Stern, Using an attribute-based decision support system for user-customized products online: an experimental investigation, MIS Quarterly 32 (1) (2008) 159–177. [40] T. Randall, C. Terwiesch, K.T. Ulrich, User design of customized products, Marketing Science 26 (2) (2007) 268–280. [41] B.G.C. Dellaert, S. Stremersch, Marketing mass-customized products: striking a balance between utility and complexity, Journal of Marketing Research 42 (2) (2005) 219–227. [42] N. Franke, M. Schreier, Why customers value self-designed products: the importance of process effort and enjoyment, Journal of Product Innovation Management 27 (7) (2010) 1020–1031. [43] G.J. Fitzsimons, Consumer response to stockouts, Journal of Consumer Research 27 (2) (2000) 249–266. [44] N. Novemsky, R. Dhar, N. Schwarz, I. Simonson, Preference fluency in choice, Journal of Marketing Research 44 (3) (2007) 347–356. [45] S. Chatterjee, T.B. Haeth, Conflict and loss aversion in multiattribute choice: the effects of trade-off size and reference dependence on decision difficulty, Organizational Behavior & Human Decision Processes 67 (2) (1996) 144–155. [46] N.K. Malhotra, Information load and consumer decision making, Journal of Consumer Research 8 (4) (1982) 419–430. [47] M. Zeelemberg, W.W. van Dijk, A.S.R. Manstead, Reconsidering the relation between regret and responsibility, Organizational Behavior & Human Decision Processes 74 (3) (1998) 254–272. [48] J.W. Payne, J.R. Bettman, E.J. Johnson, Adaptive strategy selection in decision making, Journal of Experimental Psychology: Learning, Memory, & Cognition 14 (3) (1988) 534–552. [49] J.R. Bettman, M.F. Luce, J.W. Payne, Constructive consumer choice processes, Journal of Consumer Research 25 (3) (1998) 187–217. [50] I. Simonson, Determinants of customers’ responses to customized offers: conceptual framework and research propositions, Journal of Marketing 69 (1) (2005) 32–45. [51] I. Simonson, Regarding inherent preferences, Journal of Consumer Psychology 18 (3) (2008) 191–196. [52] M. Zeelemberg, Anticipated regret, expected feedback and behavioral decision making, Journal of Behavioral Decision Making 12 (2) (1999) 93–106. [53] R. Dhar, Consumer preference for a no-choice option, Journal of Consumer Research 24 (1997) 215–231. [54] F. Salvador, C. Forza, M. Rungtusanatham, Modularity, product variety, production volume, and component sourcing: theorizing beyond generic prescriptions, Journal of Operations Management 20 (5) (2002) 549–575. [55] K. Ulrich, The role of product architecture in the manufacturing firm, Research Policy 24 (3) (1995) 419–440. [56] M.F. Luce, Choosing to avoid: coping with negatively emotion-laden consumer decisions, Journal of Consumer Research 24 (4) (1998) 409–433. [57] J. Nasiry, I. Popescu, Advance selling when consumers regret, Management Science 58 (6) (2012) 1160–1177. [58] A. De Bruyn, J.C. Liechty, E.K.R.E. Huizingh, G.L. Lilien, Offering online recommendations with minimum customer input through conjoint-based decision aids, Marketing Science 27 (3) (2008) 443–460. [59] S. Spiekermann, C. Parashiv, Motivating human–agent interaction: transferring insights from behavioral marketing to interface design, Electronic Commerce Research 2 (3) (2002) 255–285. [60] A.G. Sutcliffe, S. Kurniawan, J.-E. Shin, A method and advisor tool for multimedia user interface design, International Journal of Human–Computer Studies 64 (4) (2006) 375–392. [61] A. Merle, J.-L. Chandon, E. Roux, F. Alizon, Perceived value of the mass-customized product and mass customization experience for individual consumers, Production & Operations Management 19 (5) (2010) 503–514. [62] V. Zeithaml, Consumer perceptions of price, quality, and value: a means-end model and synthesis of evidence, Journal of Marketing 52 (3) (1988) 2–22. [63] E. von Hippel, PERSPECTIVE: user toolkits for innovation, Journal of Product Innovation Management 18 (4) (2001) 247–257. [64] E. von Hippel, R. Katz, Shifting innovation to users via toolkits, Management Science 48 (7) (2002) 821–833. [65] B. Yu, J. Skovgaard, A configuration tool to increase product competitiveness, IEEE Intelligent Systems 13 (4) (1998) 34–41. [66] I. Simonson, A. Tversky, Choice in contexts: tradeoff contrasts and extremeness aversion, Journal of Marketing Research 29 (3) (1992) 281–295. [67] A. Kobsa, J. Koenemann, W. Pohl, Personalised hypermedia presentation techniques for improving online customer relationships, The Knowledge Engineering Review 16 (2) (2001) 111–155. A. Trentin et al. / Computers in Industry 64 (2013) 436–447 [68] J.L. Zaichkowsky, Measuring the involvement construct, Journal of Consumer Research 12 (3) (1985) 341–352. [69] Z. Jiang, W. Wang, I. Benbasat, Multimedia-based interactive advising technology for online consumer decision support, Communications of the ACM 48 (9) (2005) 93–98. [70] J.H. Gerlach, F.-Y. Kuo, Understanding human–computer interaction for information systems design, MIS Quarterly 15 (4) (1991) 527–549. [71] L.M. Reeves, J. Lai, J.A. Larson, S. Oviatt, T.S. Balaji, S. Buisine, P. Collings, P. Cohen, B. Kraal, J.-C. Martin, M. McTear, T. Raman, K.M. Stanney, H. Su, Q.-Y. Wang, Guidelines for multimodal user interface design, Communications of the ACM 47 (1) (2004) 57–59. [72] K. Stanney, S. Samman, L. Reeves, K. Hale, W. Buff, C. Bowers, B. Goldiez, D. Nicholson, S. Lackey, A paradigm shift in interactive computing: deriving multimodal design principles from behavioral and neurological foundations, International Journal of Human–Computer Interaction 17 (2) (2004) 229–257. [73] A. Ansari, C.F. Mela, E-customization, Journal of Marketing Research 40 (2) (2003) 131–145. [74] T.-P. Liang, H.-J. Lai, Y.-C. Ku, Personalized content recommendation and user satisfaction: theoretical synthesis and empirical findings, Journal of Management Information Systems 23 (3) (2007) 45–70. [75] H. Berghel, Cyberspace 2000: dealing with information overload, Communications of the ACM 40 (2) (1997) 19–24. [76] S. Li, S.S. Rao, T.S. Ragu-Nathan, B. Ragu-Nathan, Development and validation of a measurement instrument for studying supply chain management practices, Journal of Operations Management 23 (6) (2005) 618–641. [77] J.K. Stratman, A.V. Roth, Enterprise resource planning (ERP) competence constructs: two-stage multi-item scale development and validation, Decision Sciences 33 (4) (2002) 601–628. [78] L. Menor, A.V. Roth, New service development competence in retail banking: construct development and measurement validation, Journal of Operations Management 25 (4) (2007) 825–846. [79] F.J. Fowler, Survey Research Methods, Sage Publications, Newbury Park, CA, 1993. [80] H.W. Marsh, Positive and negative global self-esteem: a substantively meaningful distinction or artifactors? Journal of Personality & Social Psychology 70 (4) (1996) 810–819. [81] G.C. Moore, I. Benbasat, Development of an instrument to measure the perceptions of adopting an information technology innovation, Information Systems Research 2 (3) (1991) 192–222. [82] A. Parmigiani, J. Howard-Grenville, Routines revisited: exploring the capabilities and practice perspectives, The Academy of Management Annals 5 (1) (2011) 413–453. [83] L. D’Adderio, Configuring software, reconfiguring memories: the influence of integrated systems on the reproduction of knowledge and routines, Industrial & Corporate Change 12 (2) (2003) 321–350. [84] N. Franke, M. Schreier, Product uniqueness as a driver of customer utility in mass customization, Marketing Letters 19 (2) (2008) 93–107. [85] G.J. Liu, R. Shah, R.G. Schroeder, Linking work design to mass customization: a sociotechnical systems perspective, Decision Sciences 37 (4) (2006) 519–545. [86] P.E. Spector, Method variance in organizational research, Organizational Research Methods 9 (2) (2006) 221–232. [87] P. Podsakoff, S. MacKenzie, J. Lee, N. Podsakoff, Common method biases in behavioral research: a critical review of the literature and recommended remedies, Journal of Applied Psychology 88 (5) (2003) 879–903. [88] L.J. Williams, J.A. Cote, M.R. Buckley, Lack of method variance in self-reported affect and perception at work: reality or artifact, Journal of Applied Psychology 74 (1989) 462–468. [89] J.A. Cote, M.R. Buckley, Estimating trait, method, and error variance: generalizing across 70 construct validation studies, Journal of Marketing Research 24 (3) (1987) 315–318. [90] D.W. Gerbing, J.C. Anderson, An updated paradigm for scale development incorporating unidimensionality and its assessment, Journal of Marketing Research 25 (2) (1988) 186–192. [91] J.C. Anderson, D.W. Gerbing, Structural equation modeling in practice: a review and recommended two-step approach, Psychological Bulletin 103 (3) (1988) 411–423. [92] J.F.J. Hair, R.E. Anderson, R.L. Tatham, Multivariate Data Analysis, third ed., Macmillan Publishing Company, New York, 1992. [93] C. Fornell, D.F. Larcker, Evaluating structural equation models with unobservable variables and measurement error, Journal of Marketing Research 18 (1) (1981) 39–50. [94] C.E. Werts, R.L. Linn, K.G. Jöreskog, Intraclass reliability estimates: testing structural assumptions, Educational & Psychological Measurement 34 (1) (1974) 25–33. [95] S.W. O’Leary-Kelly, R.J. Vokurka, The empirical assessment of construct validity, Journal of Operations Management 16 (4) (1998) 387–405. [96] A.E. Schlosser, T.B. White, S.M. Lloyd, Converting Web site visitors into buyers: how Web site investment increases consumer trusting beliefs and online purchase intentions, Journal of Marketing 70 (2) (2006) 133–148. [97] G. Kreutler, D. Jannach, Personalized needs acquisition in Web-based configuration systems, in: T. Blecker, G. Friedrich (Eds.), Mass Customization, [98] [99] [100] [101] [102] [103] [104] [105] [106] [107] 447 Concepts – Tools – Realization, Proceedings of the International Mass Customization Meeting 2005 (IMCM’05), GITO-Verlag, Berlin, Germany, 2005 , pp. 293–302. D. Jannach, A. Felfernig, G. Kreutler, M. Zanker, G. Friedrich, Research issues in knowledge-based configuration, in: T. Blecker, G. Friedrich (Eds.), Mass Customization Information Systems in Business, IGI Global, London, UK, 2007 , pp. 221–236. D. Jannach, G. Kreutler, Rapid development of knowledge-based conversational recommender applications with advisor suite, Journal of Web Engineering 6 (2) (2007) 165–192. A.V. Lukas, G. Lukas, D.L. Klencke, C. Nass, System and method for optimizing a product configuration. Patent Number US 7,505,921 B1, Finali Corporation, Westminster, CO (US), 2009. T. Blecker, G. Friedrich, Mass Customization Information Systems in Business, IGI Global, London, UK, 2007. K. Steger-Jensen, C. Svensson, Issues of mass customisation and supporting ITsolutions, Computers in Industry 54 (1) (2004) 83–103. J. Warschat, M. Kürümlüoglu, R. Nostdal, Enabling IT for mass customisation: the IT architecture to support an extended enterprise offering mass-customised products, International Journal of Mass Customisation 1 (2/3) (2006) 394–401. N. Franke, M. Schreier, U. Kaiser, The ‘‘I designed it myself’’ effect in mass customization, Management Science 56 (1) (2010) 125–140. F.S. Fogliatto, G.J.C. da Silveira, D. Borenstein, The mass customization decade: an updated review of the literature, International Journal of Production Economics 138 (1) (2012) 14–25. R.L. Hensley, A review of operations management studies using scale development techniques, Journal of Operations Management 17 (3) (1999) 343–358. C. Noon, R. Zhang, E. Winer, J. Oliver, B. Gilmore, J. Duncan, A system for rapid creation and assessment of conceptual large vehicle designs using immersive virtual reality, Computers in Industry 63 (5) (2012) 500–512. Alessio Trentin is an assistant professor at the Università di Padova (Italy), where he got a PhD in operations management in 2006. In 2007–2008, he was visiting assistant research professor at the Zaragoza Logistics Center (Zaragoza, Spain), a joint research center of MIT (USA) and Aragona government (Spain). His research interests mainly concern form postponement, mass customization, build-to-order supply chains, product configuration, and sustainable operations management. His work has been published in Computers in Industry, the International Journal of Operations & Production Management, the International Journal of Production Economics, the International Journal of Production Research, and the International Journal of Mass Customisation. Elisa Perin is a Ph.D. student in operations management at the Università di Padova (Italy). She holds an MS in management engineering from the Università di Padova (Italy). Her research interests are related to mass customization, product configuration and sustainable operations management. Her work has been published in Computers in Industry, the International Journal of Production Economics, and the International Journal of Production Research. Cipriano Forza is a full professor of operations management at the Università di Padova (Italy). He is also on the faculty at the European Institute of Advanced Studies in Management, where he teaches research methods in operations management. He has been visiting scholar at Minnesota University (USA), London Business School (UK), and Arizona State University (USA). His research focuses on product variety management. Currently he serves as an associate editor for the Journal of Operations Management and the Decision Sciences Journal. He is researching such topics as mass customization, concurrent productprocess-supply chain design, and product configuration. He has been successfully assisting numerous companies in these topics. He has published in the Journal of Operations Management, the International Journal of Operations & Production Management, the International Journal of Production Research, Computers in Industry, the International Journal of Production Economics, Industrial Management & Data Systems, and other journals. In 2003 and 2007, he published two books with McGraw-Hill and Palgrave Macmillan, respectively, on product information management for mass customization.