Abstract
This research integrates marketing literature, design theory, interviews with world-renowned designers, and established scale development procedures to develop a reliable and valid instrument that measures the effectiveness of design communication (i.e., the information about product designs conveyed through the product, packaging, or advertisements) via consumer evaluations. The theoretical underpinnings and face validity engaged in the development of the Design Communication Assessment Scale (DCAS) progresses the field's understanding as to what constitutes the seven evaluative dimensions of design (form, function, solidity, usefulness, style, eco-consciousness, and uniqueness). Practically, DCAS's versatility provides managers with the ability to gauge consumer evaluations of design communications while enabling better communications with designers. In addition, the authors validate a shortened form of the DCAS for resource-constrained firms. Support for DCAS’s generalizability is provided across laboratory and field studies in which ecological validity is established. Together, these studies demonstrate that using DCAS leads to improved performance across a wide array of metrics, including click-through rates, email signups, and retail sales for a diverse set of products.
Traditional design methods provide substantial aid in the development of product designs (e.g., Kano et al. 1984), yet new products still fail at high rates (Emmer 2018). The reasons for failure are numerous, but one primary cause is an inability to effectively communicate design information to consumers (Hoegg and Alba 2011; Noseworthy and Trudel 2011).
Every aspect of design communications (i.e., the various product, packaging, and advertising elements that convey information about product designs) must be considered. Yet, marketers are often left wondering whether consumers truly understand what a product design does and the benefits it provides. Although useful product design scales have been developed (e.g., Blijlevens et al. 2017; Gilal, Zhang, and Gilal 2018; Homburg, Schwemmle, and Kuehnl 2015), the information provided by these instruments cannot fully answer these concerns, nor do they provide actionable insights for marketers. To address this gap, we adopt a designer lens to develop a scale that measures consumer evaluations of design communications, establishing an enhanced ability to predict important outcomes (e.g., ad engagement, product sales) and respond to design deficiencies.
We introduce the Design Communication Assessment Scale (DCAS), a tool anchored in centuries-old design theories that provides direction for marketers to meaningfully improve product design communications. As its name indicates, this scale is focused not only on improving design but also on enhancing the elements around it (e.g., advertising, packaging) that help communicate the main attributes and benefits of a product to consumers. Further, the use of influential theories of design expands the nomological network (beyond prior work) to more accurately capture how designs are understood by designers—those who create and adjust product designs and their associated communications. In an increasingly visually crowded market (Kane and Pear 2016), it is imperative that marketers be able to assess and adjust the effectiveness of their design communications, and DCAS facilitates this. Consequently, the contributions of our work are both theoretical and practical.
Theoretically, we develop an enriched framework utilizing marketing insights to reconcile and categorize two long-standing design theories that have laid the foundations for current design thinkers (e.g., Hayward 1998; Norman 2013; Rams 2014). Further, we demonstrate how two new dimensions have emerged as important aspects for consumer design assessment. Developed through the integration of designer input and marketing theory, our framework provides novel insights into consumers’ evaluations of design communications.
Practically, our work offers an actionable tool for marketers that can improve their design communications. By using DCAS, a firm can pinpoint the dimensions of a design that are (or are not) being effectively communicated to consumers, a crucial contribution, as designs are more likely to succeed when marketers can promptly respond to consumer concerns, insights, and experiences (Griffin and Hauser 1993). Additionally, due to our use of design's theoretical foundations and extensive face and nomological validity assessments in the development of DCAS (e.g., input and perspectives from world-renowned and practicing designers), our tool uses a common design language, thereby helping designers better adjust their work. Thus, DCAS represents an evaluative bridge that can help marketers engage with designers.
Our article is laid out as follows. First, we give an overview of existing design scales. We then provide the theoretical background for DCAS; this involves investigating design foundations and integrating them with marketing literature. Following this, we provide an overview of the generation and evaluation of DCAS using established scale-development best practices (Netemeyer, Bearden, and Sharma 2003). Next, we demonstrate the effectiveness of our seven-dimension instrument in two field studies that exhibit significantly improved outcomes (e.g., actual sales, email signups). We then validate a shortened version of DCAS. Finally, we discuss the theoretical and practical implications of our work and potential avenues for future research.
Prior Product Design Scales
Prior work has sought to inform the discipline via product design scales, yet, as shown in Table 1, there are theoretical, empirical, and practical limitations associated with these scales. 1
Comparison of Design Scales’ Composition and Construction.
Regarding theory, as designers have spent centuries determining what constitutes good design, an improvement of DCAS over the scales in Homburg, Schwemmle, and Kuehnl (2015) and Gilal, Zhang, and Gilal (2018) is that it is based on well-established design theories and interviews with designers. This shortcoming in previous scales results in a limited ability to help designers create offerings that effectively communicate product design. For instance, their symbolism/reflective dimension, which we do not consider, pertains to how consumers feel about a design in relation to other consumers. This is tangential to DCAS's purpose, is not mentioned by designers, and provides little direction for designers when adjusting a design. DCAS’s dimensions encompass other scales’ dimensions (Table 2), but our dimensions use more familiar terminology or are worded in a manner consistent with design theory. Although Blijlevens et al. (2017) utilize design theory and interview designers, their scale is focused on aesthetic pleasure, a subset of our scale, and is not intended to bridge the gap between design and marketing. To address these issues, we draw on established design theories, integrate these theories with marketing research, solicit designer expertise via interviews, and have a licensed designer on the author team.
Definitions, Sources, and Predicted Relationships of DCAS’ Seven Dimensions.
Notes: A highly rated (i.e., good) design does not require high evaluations for all dimensions, as trade-offs occur.
Empirically, Homburg, Schwemmle, and Kuehnl (2015) and Gilal, Zhang, and Gilal (2018) do not engage in face validity assessments. Unfortunately, this omission can be misleading in implementation. For instance, design theory and practicing designers use the term “functionality” in a substantially different manner than how it has been employed in marketing. For example, the dimension Homburg, Schwemmle, and Kuehnl refer to as functionality is described by designers as “solidity” (a dimension of our scale, informed by design theory and defined in a subsequent section). In the development of our instrument, we have taken care to use terms that have meaning and relevance to designers. Further, prior design scale papers in marketing have not connected their instruments to other important constructs (i.e., they did not engage in nomological validity assessments, thereby isolating their empirical measures from a theoretically appropriate network of relationships). The current research not only engages in a formal face validity assessment with practicing designers and undertakes nomological validity assessments but also includes experimental validity assessments and a test–retest, steps that are important to ensure consistent and meaningful assessments but have been absent in prior scales.
Practically, marketers and designers are limited by what can be done in response to the findings from prior design scales. DCAS's theoretical foundations, more-thorough scale development steps, and generalizable applicability provide greater insight than existing design instruments regarding what marketers should do when adjusting design communications.
Theoretical Background
To bolster marketers’ communications and collaborations with designers, we investigated the roots of design theorizing (aligning with recommendations from marketing researchers; e.g., Dahl 2011; Eppinger and Ulrich 2015; Luchs and Swan 2011). This revealed consistent usage of two centuries-old design theories—the Vitruvian Triad (Gwilt 1826) and “form follows function” (Sullivan 1896)—that have helped inform current design theorists (e.g., Norman 2013; Rams 2014). Yet, design literature does not fully address how these theories relate to each other. To reconcile and integrate these design theories, we focus on marketing research showing that consumers evaluate designs on two fundamental sets of dimensions, the intrinsic and extrinsic. The intrinsic dimensions are those salient, immutable attributes of a product design that, if changed, result in an alteration to the very nature of the design itself (e.g., a phone that cannot make calls is not a phone; Zeithaml 1988). In contrast to the intrinsic, extrinsic dimensions are those attributes ascribed to a design by individual consumers to assess potential personal benefits and future usage considerations (Zeithaml 1988). In the following sections, we explicate how form follows function assesses the intrinsic dimensions and the Vitruvian Triad assesses the extrinsic dimensions.
Intrinsic Dimensions
Intrinsic dimensions dictate how a design manifests and aid in the understanding of new designs (Moreau, Markman, and Lehmann 2001); thus, consumers use intrinsic dimensions to determine what a design is. This correlates with the design theory of “form follows function,” a phrase that has been understood by designers to mean that any design is composed of two primary dimensions—form and function (Sullivan 1896). These dimensions have been noted to play an important role in consumer evaluations of designs (Jindal et al. 2016; Noseworthy and Trudel 2011), and marketing researchers have suggested that to understand designs better, form and function should be studied together (Dahl 2011; Luchs and Swan 2011). However, as we discuss subsequently, these two dimensions alone do not fully capture consumers’ design responses.
Form
Form is the resultant physical manifestation of a product idea. Sullivan (1896) initially described form as a shape or an outward semblance, and it has been defined as a recognizable external appearance, a particular state, or the integration of elements such as shape, size, color, and texture to make a coherent image (Ching 2014). Examples of the form include the shape and materiality that constitute a coffee mug, a toothbrush, or a warehouse (for an overview of these perceptual factors, see Sample, Hagtvedt, and Brasel [2020]). Even though form is related to and has been correlated with aesthetics in some marketing literature (e.g., Bloch, Brunel, and Arnold 2003), in alignment with design theorists (Fiell and Fiell 2016; Rams 2014), we assert in the “style” subsection that aesthetics is distinct from form.
Function
Function is the intended purpose and outcome provided by the form. Sullivan (1896) described function as the purpose or ability provided by a design. Examples include the ability of a mug to hold coffee, the cleaning ability of a toothbrush, or the storage capacity of a warehouse. Therefore, form is said to follow function because one typically has an intended purpose before being able to create a form, even if the purpose is aesthetically oriented. Thus, any product in a category must have a recognizable function relevant to that category. For example, in making the initial iPhone, the form was the result of the design process, but it was the purpose of providing a sleek, portable phone without buttons (i.e., the function) that initially drove the iPhone's creation. Prior research has noted that consumers make assessments about function based on the form of designs (Hoegg and Alba 2011; Noseworthy and Trudel 2011) and that these assessments help consumers ascertain what a product does before they decide if they need or want it. This aligns with both design theorists (Fiell and Fiell 2016; Norman 2013; Rams 2014) and marketing researchers (Noseworthy and Trudel 2011), who argue that designs should be readily understood (i.e., discernible) by consumers.
Extrinsic Dimensions: The Vitruvian Triad
Beyond understanding what a design is, consumers must also determine the potential personal benefits from designs—how well a design may fit into their lifestyle, with the environment, and in relation to others’ products (Belk 1988; Haws, Winterich, and Naylor 2014; Tian, Bearden, and Hunter 2001). That is, consumers evaluate the goodness of designs, and this relates to the work of the 1st century BCE Roman author and architect Vitruvius. Vitruvius contended that good designs are solid, useful, and attractive, referred to as the Vitruvian Triad (Gwilt 1826). Since consumers draw on extrinsic attributes to assess potential benefits (Zeithaml 1988), it follows that the three dimensions of the Vitruvian Triad, which designers have used for over two millennia to evaluate designs, correlate with the extrinsic.
Solidity
Solidity is defined as the dependability of a design over the expected life of a product. Solid designs are made of appropriate materials that can be relied on to serve the intended function over time (Fiell and Fiell 2016; Gwilt 1826). This constitutes using materials effectively and integrating them in a way to provide consistent, reliable results.
Usefulness
Usefulness is defined as the ability to meet a consumer's need through a design. Useful designs have previously been noted as meeting user desires and/or needs (Gwilt 1826). Yet, usefulness should not be ascribed only to utilitarian products, for even if the usefulness of a design is the generation of positive affect (e.g., a piece of sculpture), this still meets a consumer need (Fiell and Fiell 2016). Further, usefulness is distinct from function, as function evaluates how discernible the purpose of a design is, not the potential benefits consumers may receive from a design.
Style
Style is defined as the sensorial appeal of a design. (Some designers refer to this dimension as “beauty.”) Style is a provision of aesthetics (Gwilt 1826) and can be an appeal to all senses (Hekkert 2006). Thus, various sensorial aspects can be considered aesthetically pleasing depending on the consumer (e.g., the taste of a fine wine, the sound of an operatic voice, the feel of cashmere, the smell of a perfume, the motion of a John Deere tractor). Style should not be equated only with hedonic objects, however, as utilitarian products can also be considered to have an appealing style.
It is important to note that style should not be confused with the intrinsic dimension of form. Whereas form and style have a close relationship, there is nuance to these distinct constructs, as style relates to the benefits provided, whereas form relates to the understanding of a design. Thus, even though these two constructs will highly correlate at times, this is not a given. For instance, a design, such as a particular car model, may not be attractive to a consumer (i.e., low style rating), but it can still adequately meet product category expectations (i.e., high form rating). Consequently, this is a beneficial and important distinction.
Additional Extrinsic Dimensions
Solidity, usefulness, and style have received extensive attention since the 1st century BCE for evaluating the goodness of designs (Hayward 1998; Norman 2013; Rams 2014). However, based on our literature review and interviews (as discussed in the empirical section), important additional extrinsic dimensions have emerged over the past century. The Industrial Revolution generated technological advances while simultaneously accelerating resource depletion (Allen 2009). The resultant mass production and standardization have led to substantial cost efficiencies (and prices affordable to more consumers) but at the expense of environmental damage and eroded product uniqueness. This has given rise to two novel and important measures of design: eco-consciousness and uniqueness. Consumers now place greater emphasis on these dimensions, and this is captured through an increasing prominence of these constructs within design and marketing literature (Liu et al. 2017; Luchs et al. 2010; Norman 2013; Rams 2014; Tezer and Bodur 2020). Although we considered other extrinsic dimensions in the creation of our scale, they were not mentioned by design experts, are already encompassed within one of the dimensions presented here (e.g., ergonomics is an aspect of usefulness), or are too distal from the intended purpose of DCAS (e.g., a designer would be uncertain how to adjust “symbolism”).
Eco-consciousness
Eco-consciousness is defined as the preservation, protection, and/or promotion of environmentally friendly behavior through a design or in the creation of a design. As natural resource exploitation and energy consumption have increased, society has become more focused on preserving and protecting the environment (Brophy and Lewis 2011). Consequently, we see an increasing emphasis on eco-consciousness in designs, as consumers value green products, are more likely to pay premiums for them, and can receive a boost in the consumption experience when purchasing them (Griskevicius, Tybur, and Van den Bergh 2010; Haws, Winterich, and Naylor 2014; Luchs et al. 2010; Tezer and Bodur 2020). Further, designers are increasingly focused on this construct (Norman 2013; Rams 2014).
Uniqueness
Uniqueness is defined as the manifestation of a design such that it is perceived as distinct from others. Mass production of virtually identical products via increased industrialization (Allen 2009) has led to a greater desire for uniqueness. This trend has been supported by an increase in societal wealth, intense market competition, a proliferation in consumer segments, and technological advances that allow for more efficient manufacturing of various products. As a result, consumers do not want product designs to be too similar to product category prototypes (Liu et al. 2017), and their design preferences are influenced by a need for uniqueness (Irmak, Vallen, and Sen 2010; Simonson and Nowlis 2000; Tian, Bearden, and Hunter 2001). In group settings, consumers make choices to appear more unique (Ariely and Levav 2000), and nonconformity leads to higher evaluations of consumers by others (Bellezza, Gino, and Keinan 2014), so it can be beneficial to stand out via uniqueness. This dimension can also manifest as scarcity (Hamilton et al. 2019). On the design side, leading designers have also noted the importance of uniqueness (Rams 2014; Norman 2013).
Dimensionality Summary
Our synthesis of these two design theories, technological advancements since the establishment of said theories, and the inclusion of designer perspectives (see the section titled “Stage 1: Item Generation and Content Validity Assessment of DCAS) indicate the need for seven design dimensions, as captured in Table 2. Theoretically, a product must have a form and function to belong to a particular product category, as these are necessary elements of designs. Once developed, a product design should have effectively fulfilled this requirement. The intrinsic dimensions give consumers the ability to understand a product and assess whether that product design meets their expectations. In contrast, consumers use the extrinsic dimensions to ascertain whether benefits will be realized from a design compared with design alternatives. Thus, design success is contingent on both the intrinsic and extrinsic, and consumer evaluations of products can be simultaneously based on all seven dimensions.
The success of a design is not reliant on all dimensions being perceived positively, however. Trade-offs are often made between different aspects of a design (e.g., it may be difficult to emphasize eco-consciousness when highlighting solidity [Luchs et al. 2010] or to be unique when striving for a prototypical product category design [Liu et al. 2017]). Further, as seen in judgment and decision making (Plous 1993), even if a consumer rates one dimension highly (e.g., the style of a wheelbarrow), this rating may not be useful in predicting consumer response if that consumer does not place any importance on that design dimension.
DCAS Development and Application: Overview
The development and application of DCAS is presented in four stages, as summarized in Table 3. For every data set and study, participants received course credit (undergraduate students) or cash (workers from Amazon Mechanical Turk [MTurk] and Prolific). Whole numbers of participants were recruited (e.g., 400), but this inevitably varied (e.g., 406). We chose stimuli from diverse product categories within the suggested new products section on Amazon. These designs, referenced by design set names A, B, and C, are presented in Figures A1, A2, and A3 in the Appendix.
Overview of the Development Process of DCAS.
Notes: Only the most pertinent Web Appendix tables are listed in this table.
In Stage 1, we develop a reliable scale that reflects the seven dimensions embedded in the conceptual model. We accomplished this via expert interviews, deductive and inductive item generation, and convergent and face validity assessments. Stage 2 establishes the discriminant, nomological, predictive, and experimental validity of DCAS, as well as its temporal stability. In Stage 3, we demonstrate the generalizability and applicability of DCAS through lab and field studies. Finally, in Stage 4, we validate a shortened form of DCAS, providing a condensed version of the tool for practitioner use. We present a summary of steps, data, and results in Table 3, with ancillary materials for each stage presented in the Web Appendix.
Stage 1: Item Generation and Content Validity Assessment of DCAS
Assessment of the DCAS Theoretical Framework
Before generating our initial items for DCAS, the first author interviewed ten Japanese and American designers, including world-renowned architects 2 Toyo Ito and Ryue Nishizawa, 2013 and 2010 winners of the Pritzker Prize (architecture's Nobel Prize equivalent). These interviews supplied insights into how expert designers consider and evaluate designs. Their responses frequently mentioned the seven dimensions of DCAS, whereas no other criteria were consistently identified. This provides qualitative support for our conceptual framework, pointing to the importance of generating items for all seven dimensions.
Deductive and Inductive Item Generation
We conducted the deductive step in item generation drawing on research and literature within the marketing and design fields (a representative list is provided in Web Appendix A). These references include Vitruvius’s writings (Gwilt 1826), Dieter Rams’s principles of good design (Rams 2014), Don Norman's understanding of design (Norman 2013), published marketing research on design (e.g., Dahl 2011; Eppinger and Ulrich 2015; Luchs and Swan 2011), and writings on design from other sources (e.g., Ching 2014; Fiell and Fiell 2016).
We conducted the inductive step in item generation with a new, mixed set of eight practicing designers (i.e., architects, artisans, graphic designers, and industrial designers). These designers were asked open-ended questions regarding how they evaluate designs. We used their responses to generate additional items not included from our literature review. Collectively, these deductive and inductive actions generated 200 initial items representing the seven dimensions of our proposed scale.
Dimensionality Assessment (Exploratory and Confirmatory Factor Analysis)
Using an initial data set, we pared a preliminary set of 200 items down to 140. We did this by assessing the convergent validity of items proposed for each of the seven dimensions through exploratory factor analysis (EFA). We dropped items that did not adequately load on the expected dimension (i.e., loadings less than .40) or that reduced the dimension's internal consistency (i.e., removal of the item increased Cronbach's alpha). We next examined the effectiveness of these 140 items via Data Set 1. To generate this data set, we had 406 undergraduate students (51.7% male, 47.9% female; Mage = 20.7 years) rate one of five randomly shown designs (Design Set A, Figure A1: an inflatable paddleboard, an anthropomorphic salt-and-pepper shaker, an innovative scooter, seaweed snacks, and a bath towel set) for all 140 items on seven-point Likert scales, ranging from 1 (“strongly disagree”) to 7 (“strongly agree”).
We repeated the convergent validity assessment steps of EFA and alpha examination for each dimension with Data Set 1. Further, we assessed cross-covariances between the separate dimensions to ensure that distinct constructs were being measured. For some of the dimensions, five or more appropriately loading items emerged, but these were subsequently reduced to three items per dimension for the sake of parsimony in use (Böckenholt and Lehmann 2015; Netemeyer, Bearden, and Sharma 2003). This resulted in a final 21-item scale. Table 4 reports these items, their loadings, and Cronbach's alphas for each data set from Stages 1 and 2 (Data Sets 2–4 are detailed in Stage 2).
CFA Factor Loadings and Alphas.
Indicates reverse-coded.
Notes: Alphas are listed in data set order (1, 2, 3, and 4). Avg. = average loadings across data sets.
Confirmatory factor analysis (CFA) indicates that the seven-factor correlated model meets recommended levels (Hu and Bentler 1999; Steiger 2007) in terms of goodness-of-fit (root mean square error of approximation [RMSEA] = .07; comparative fit index [CFI] = .93; Tucker–Lewis index [TLI] = .91; standardized root mean squared residual [SRMR] = .05). Additionally, all the average variances extracted (AVEs) meet the standard of being above .5 (Fornell and Larcker 1981), and the composite reliability for each dimension is above the recommended level of .70 (Hair et al. 1998). See Web Appendix A for this summary (Table W1), the descriptive statistics for this data set (Table W2), and the cross-loadings (Table W3). Furthermore, the values of α reported in Table 4 for all seven dimensions across all data sets are consistently above the threshold value of .70 (demonstrating adequate convergent validity between the three items proposed for each of the seven dimensions; Nunnally 1978), and the factor loadings across data sets and their averages (last column) are consistently above the recommended .70 (Bagozzi and Yi 1988).
Face Validity
After completing the CFA for Data Set 1, we invited the designers who initially provided input for our item generation to formally evaluate our final item sets. They were provided the dimension definitions and the three scale items for each dimension. They were asked, on a seven-point Likert scale ranging from “extremely inappropriate” (1) to “extremely appropriate” (7), to indicate how appropriate the three items were for measuring their intended constructs. The overall average (6.1) fell within the “moderately appropriate” to “extremely appropriate” range, providing strong face validity support (see Web Appendix A, Table W4).
Stage 1 Conclusion
Stage 1 provides support for the content viability of our proposed seven-dimensioned, 21-item scale (Table 4). We make further validity assessments in Stage 2.
Stage 2: Validity Assessment of DCAS
In Stage 2, we collected three primary data sets (2, 3, and 4), one subset (3a), and two additional Web Appendix data sets (A and B; Web Appendix B). We generated Data Set 2 to engage in discriminant validity assessments. We generated Data Set 3 to assess nomological validity, with a subset of these participants providing additional data to assess temporal stability through a test–retest (3a). We generated Data Set 4 to evaluate the predictive validity of DCAS while comparing DCAS with existing design scales. We also conducted experimental validity assessment during this stage, but as the studies in Stage 3 are similar in execution and results, we provide this in Web Appendix B. Refer to Table 3 for a summary of the steps, data used, and results in this stage.
Discriminant Validity (Data Set 2)
The discriminant validity of a response scale is highly dependent on the entity being evaluated (Haws, Sample, and Hulland 2023). Two or more dimensions may be highly correlated for one entity being evaluated, but these dimensions may be uncorrelated for another. For example, correlations (Web Appendix B, Table W2) between the dimensions of form and style are high for evaluations of the seaweed snack from Data Set 1 (r = .60), but low for the scooter (r = .05). Thus, to ensure an accurate assessment of discriminant validity for DCAS that is not overly influenced by the designs being evaluated, participants for Data Set 2 (n = 365 undergraduate students; 46.3% male, 53.1% female, .3% nonbinary, .3% prefer not to say; Mage = 20.2 years) evaluated two randomly presented product designs (one from Design Set A and one from Design Set B: an all-in-one breakfast center, an electric lawn mower, a kitchen tool, and a package of water balloons; see Figure A2), resulting in 730 unique evaluations.
We utilized Fornell–Larcker overlapping confidence intervals, constrained phi, and the heterotrait–monotrait ratio of correlations (HTMT) for discriminant validity assessment (Henseler, Ringle, and Sarstedt 2015). Results (Table 5) exceed the discriminant validity criteria for every dimension and analysis (Fornell and Larcker 1981; Hair et al. 1998), with form and style meeting the cutoff for the most conservative HTMT analysis (i.e., .85; Henseler, Ringle, and Sarstedt 2015; Voorhees et al. 2016). High correlations may emerge between form/style and solidity/usefulness (Web Appendix A, Table W2; Web Appendix B, Tables W1, W2, and W8), but this is predictable, considering the theoretical basis of these dimensions and given that this is a response scale (Haws, Sample, and Hulland 2023). Still, to ensure a proper discriminant validity assessment, we conducted the same assessments for Data Set 3, finding similar results (Web Appendix B, Table W3).
Discriminant Validity (Data Set 2: Design Sets A and B).
Chi-square difference tests indicate that the seven-factor model is significantly better than all alternatives.
The seven-factor model provides the best AIC, and this number was utilized to calculate the differences.
To provide further support for discriminant validity, we compared the proposed scale with several alternative models (Table 5 and Web Appendix B, Table W3): a six-factor model combining form and style, a six-factor model combining solidity and usefulness, a five-factor model combining form and style as well as solidity and usefulness, a two-factor model summing the intrinsic and extrinsic dimensions, a higher-order model with the two intrinsic and the five extrinsic dimensions, and a single-factor model summing all dimensions. Chi-square difference tests, which compare the seven-factor model with the more constrained alternative models (Bagozzi, Yi, and Phillips 1991; Jöreskog 1969), reveal that the suggested model significantly outperforms all alternatives (p < .001).
Additionally, all AVE values for both data sets meet the standard of being above .5 (Fornell and Larcker 1981), and the composite reliability for each dimension is above the recommended level of .70 (Hair et al. 1998). Change in Akaike information criterion (AIC) and other metrics also support the seven-factor model as a better fit, meeting recommended levels (Hu and Bentler 1999; Steiger 2007) in terms of goodness-of-fit (CFI = .96; TLI = .95; SRMR = .04; RMSEA = .06; Bagozzi and Yi 1988, 2012; Hu and Bentler 1995). Consequently, despite the tendency for a few dimensions to highly correlate at times (depending on the product design), these collective results provide confidence in the discriminant validity between the seven dimensions of DCAS.
Nomological Validity: (Data Set 3)
To be considered nomologically valid, the dimensions of DCAS must be shown to be empirically correlated with theoretically related constructs (Netemeyer, Bearden, and Sharma 2003). When looking at good design holistically, other researchers have posited that good designs should generate a general, positive affect that leads to consumer response (e.g., Srinivasan, Lovejoy, and Beach 1997). Thus, when assessing the nomological validity for the overall measure of DCAS, we utilized the Positive and Negative Affect Schedule (PANAS; Watson, Clark, and Tellegen 1988). We anticipated that evaluations of good designs should be highly correlated with the positive dimensions of PANAS, whereas the negative dimensions of PANAS should be negatively correlated. In addition to general positive feelings, good design has also been linked to feelings of achievement, joy, and inspiration (Givechi and Velázquez 2003); consequently, we predict that our dimensions should be positively correlated to these.
Furthermore, research has suggested that dimensions of DCAS are associated with other related constructs. Satisfaction (Han and Hong 2003) and hedonic benefits (Bloch 2011; Chitturi, Raghunathan, and Mahajan 2008; Hekkert 2006) should be positively correlated to the form and style dimensions. Utilitarian benefits should be positively correlated with function, solidity, and usefulness (Bloch 2011; Fiell and Fiell 2016; Norman 2013). Because style (Townsend and Sood 2012), eco-consciousness (Griskevicius, Tybur, and Van den Bergh 2010; Luchs et al. 2010), and uniqueness (Liu et al. 2017; Simonson and Nowlis 2000) are typically used by consumers to send signals to others, these dimensions should positively correlate with symbolic benefits (Belk 1988; Bloch 2011).
To assess these nomological relations, 301 U.S. MTurk workers (Data Set 3; 43.9% male, 56.1% female; Mage = 37.1 years) responded to these measures in addition to DCAS. Participants evaluated a randomly presented product from Design Set A; responded to PANAS (Watson, Clark, and Tellegen 1988); answered questions about achievement, joy, and inspiration (Givechi and Velázquez 2003); and responded to single-item measures about satisfaction (Han and Hong 2003), hedonic benefits (Bloch 2011), utilitarian benefits (Norman 2013), and symbolic benefits (Bloch 2011). When looking at the correlations between DCAS constructs and these theoretically related constructs, significant (but not overly high) correlations occur in the predicted directions, indicating sufficient nomological validity (for results, see Web Appendix B, Table W4).
Temporal Stability: Test/Retest (Data Sets 3 and 3a)
To ascertain the stability with which DCAS assesses each dimension over time, 200 MTurk workers from Data Set 3 were invited to reevaluate their previously assigned product three to five weeks later, yielding Data Set 3a (184 completed responses; 46.7% male, 53.3% female; Mage = 38.3 years). This moderate three- to five-week time frame prevented participants from remembering their previous answers while also avoiding concerns about substantial shifts in attitudes over longer periods of time. High correlations emerged between the dimensions over time (α = .69–.88; r = .53–.78; Web Appendix B, Table W5), indicating strong test–retest reliability.
Predictive Validity (Data Set 4)
We next assessed DCAS’s predictive ability compared with previously published design scales. 3 For this data set, 505 undergraduate students (44.2% male, 55.4% female, .2% nonbinary, .2% prefer not to say; Mage = 20.1 years), who reported having never seen their randomly assigned design before, evaluated one of five product designs (Design Set C, Figure A3: a light-up baseball, a bottle topper, a dog bike leash, a soap sponge, and an interactive whiteboard). Participants were asked to evaluate the presented design with DCAS and the scales in Blijlevens et al. (2017), Gilal, Zhang, and Gilal (2018), and Homburg, Schwemmle, and Kuehnl (2015). The ordering of scale presentation was randomized. This study design resulted in a 5 (product design seen) × 4 (design scale) mixed design. Afterward, participants were asked to report their agreement on seven-point Likert scales (with 7 being the most positive response) for the following predictive measures of behavioral intentions: purchase intent (“How likely would you be to purchase this product?,” “How do you feel about buying this product in the near future?”), willingness to pay (WTP; “How much would you be willing to pay?”), word of mouth (WOM; “How likely would you be to tell your friends and family about this product?,” “I would tell other people about this product”), and attitude (“My attitude toward this product is very positive”). Gender and age were collected at the end of the study.
To compare the predictive performance of the scales, we estimated hierarchical linear regression using an indexed measure of the six positive behavioral intention items as the dependent variable. This was done for each alternative scale, resulting in three separate comparisons (Table 6). The introduction of DCAS to each of these scales explains significantly more variance (ΔR2s > .16). Further, when adding in DCAS, several of the alternative dimensions no longer significantly predict performance. When considering this analysis in the reverse (i.e., DCAS initially in the model and then adding an alternative scale), the alternative scales also explain significantly more variance, although the incremental gains are much smaller (ΔR2s < .02). That is, the incremental variance explained by DCAS when added to one of the existing scales is an order of magnitude larger than that of one of the existing scales when added to DCAS.
Predictive Validity Individual Scale Comparisons Hierarchical Linear Regression: Data Set 4; Design Set C (n = 505).
Significant at .1. *Significant at .05. **Significant at .01. ***Significant at .001.
Notes: AP = aesthetic pleasure.
To address the potential issue of DCAS explaining more variance simply by having more dimensions, we estimated additional hierarchical regression estimations chronologically, with the first model being the oldest design scale in this set (i.e., Homburg, Schwemmle, and Kuehnl 2015), and then adding each subsequent scale until the fourth model contained all four scales. We conducted this analysis for Data Set 4 aggregated across all product designs as well as individually by design seen (Web Appendix B, Tables W9A–W9F). These results demonstrate that DCAS consistently explains significantly more variance, even when added last to the four-scale predictor model (a conservative test). Additionally, considering DCAS first and then adding in all three alternative scales at once in a second model, rather than one by one, increases the variance explained only to a limited extent, attesting to DCAS’s superior predictive performance.
Stage 2 Conclusion
Collectively, the work in Stage 2 establishes the validity and reliability of DCAS. In Stage 3, we demonstrate how DCAS can be utilized to improve design communications.
Stage 3: Application of DCAS
For this third stage, we assess the usability and generalizability of DCAS via two lab studies and two field studies. For space considerations, we present the lab studies (Web Appendix Lab Studies 1 and 2) in Web Appendix C. In the first field study (Study 1), we collaborated with an Amazon-based retailer for an in-market comparison of existing advertisements to DCAS-informed advertisements. In Study 2, a designer from Fiverr (a freelance services website) designed a product concept coffee maker and a subsequent DCAS-informed version to be advertised on Facebook. DCAS's input in both studies resulted in significant increases in marketing metrics (e.g., sales, orders, email signups). Both studies were conducted in three phases, and for ease of understanding and replicability, we present this process by phase. In Phase 1, initial designs were evaluated using DCAS; in Phase 2, feedback from DCAS was used to create new versions; and in Phase 3, both original and DCAS-informed versions were shown to consumers via Amazon (Study 1) or Facebook (Study 2). A summary of Stage 3 can be found in Table 3.
Study 1: Field Study—DCAS-Informed Ads Lead to Increased Amazon Sales
For Study 1, we illustrate the exploratory value of using DCAS in an applied real market context by partnering with an independent retailer selling products on Amazon. The objective of this partnership was to help bolster the retailer’s product design communications and sales for two products: silicone baby cups and reusable sous vide bags.
Product 1: Silicone Baby Cups
Phase 1
One hundred ten undergraduate students (51.8% male, 45.4% female, 0% nonbinary, 2.8% prefer not to say; Mage = 20.5 years) evaluated the original advertisement for the baby cups (provided to us by our partner retailer; see Figure A4) using DCAS, responded to the six behavioral intention questions, and provided demographics. 4 A regression analysis reveals that both usefulness (β = .413, SE = .154, t(103) = 2.682, p = .009) and style (β = .432, SE = .133, t(103) = 3.244, p = .002) had a significant, positive effect on the indexed purchase intention questions (r = .95). The same effect emerged for the indexed WOM activity (r = .92), WTP, and attitude measures (in Web Appendix C, see Table W1 for full regression results and Table W2 for structural equation model analysis). The dimensions found to have an influence (i.e., usefulness and style) were targeted for improvement. If usefulness and style could be further bolstered, then there should be a coinciding increase in the behavioral intentions. These findings were communicated to the retailer, which produced an adjusted, DCAS-informed advertisement (Figure A4). The adjustment involved adding milk to the ad to better convey the product's usefulness while also increasing its aesthetic appeal (i.e., style).
Phase 2
To determine whether the retailer's adjustments had the intended effect on evaluations, we recruited 86 undergraduate students (55.8% male, 44.2% female, 0% nonbinary, 0% prefer not to say; Mage = 20.2 years) to evaluate the DCAS-informed advertisement using the same measures as in Phase 1. To compare the data for both advertisements, we estimated a multivariate analysis of variance (MANOVA; F(11, 185) = 1.836, p = .051; Wilks' Λ = .90) with advertisement seen as the independent variable and the seven dimensions of DCAS and the behavioral intention measures as the dependent variables. We observed significantly higher results for the DCAS-informed advertisement. For full results, see Table 7.
Field Study: Baby Cup Advertisements.
Significant at .1. *Significant at .05. **Significant at .01. ***Significant at .001.
Notes: Impressions = number of times the ad is seen by a consumer.
Phase 3
To investigate whether the DCAS-informed advertisement would produce better behavioral outcomes than the original advertisement, the partner firm agreed to launch both ads on Amazon for a total of 52 days from November 23, 2020, to January 14, 2021 (a time frame that was determined based on the company's budget constraints and willingness to “test” the revised advertisements). To control for various issues and reduce the influence of higher shopping days and other potential variables, the advertisements displayed on Amazon were alternated each day, with only one advertisement being run per day. Consequently, each ad was shown for a total of 26 days. As this retailer is the exclusive seller of these products on Amazon, we are certain that these were the only ads running for these items during this time frame.
Whereas the original baby cup advertisement was seen by more people (i.e., impressions) than the DCAS-informed ad (Table 7), the DCAS-informed advertisement had better results over the 52-day period. Specifically, per 1,000 impressions (and assuming a normal distribution), the DCAS-informed ad was clicked on at a significantly higher rate (Moriginal = 6.59 vs. MDCAS-informed = 9.03; F(1, 50) = 36.493, p < .001), leading to a significantly higher number of orders (Moriginal = .39 vs. MDCAS-informed = .73; F(1, 50) = 36.172, p < .001) and a significantly higher average revenue per sale (USD; Moriginal = $4.71 vs. MDCAS-informed = $8.86; F(1, 50) = 36.296, p < .001). Consequently, per 1,000 impressions, the DCAS-informed ad generated a 37.03% increase in clicks, an 87.18% increase in orders, and an 88.11% increase in sales revenue.
Additional evidence for the effectiveness of the DCAS-informed intervention is that the baby cup featured in the DCAS-informed ad moved from ranking #9,047 in Baby and #281 in Toddler Cups to #975 in Baby and #42 in Toddler Cups on Amazon over a four-month span (the adjusted ad continued running for four months after the field study until a stockout occurred), indicating substantially improved sales versus the competition within the same product category.
Product 2: Reusable Sous Vide Bags
Phase 1
One hundred eleven undergraduate students (48.6% male, 48.6% female, 0% nonbinary, 2.8% prefer not to say; Mage = 20.7 years) evaluated the original advertisement for a reusable sous vide bag (Figure A5) using DCAS, responded to the behavioral intention questions, and provided demographics. A regression analysis revealed that function (β = .155, SE = .072, t(104) = 2.156, p = .033), solidity (β = .307, SE = .132, t(104) = 2.326, p = .022), usefulness (β = .375, SE = .129, t(104) = 2.907, p = .0047), and style (β = .628, SE = .120, t(104) = 5.235, p < .001) all had a significant, positive effect on purchase intention (r = .90). Similarly, we found significant results for WOM (r = .92), WTP, and attitude (Web Appendix C, Table W3). These findings were communicated to the retailer, which produced a DCAS-informed ad (Figure A5) that aimed to achieve an increase in these dimensions. Specifically, to increase the potential perceived functionality, solidity, and usefulness, a food item was pictured inside the sous vide bag, despite the expected attractiveness of seeing an unbagged food item.
Phase 2
Eighty undergraduate students (57.5% male, 41.3% female, 0% nonbinary, 1.2% prefer not to say; Mage = 20.3 years) evaluated the DCAS-informed advertisement using the same measures as in Phase 1. We compared the Phase 1 and 2 data sets using a MANOVA (F(11, 180) = 2.083, p = .024; Wilks' Λ = .89) with advertisement seen as the independent variable and the seven dimensions of the DCAS scale and the behavioral intention measures as dependent variables. Significantly higher results were observed for the DCAS-informed advertisement. For full results, see Table 8.
Field Study: Sous Vide Bag Advertisements.
Significant at .1. *Significant at .05. **Significant at .01. ***Significant at .001.
Notes: Impressions = number of times the ad is seen by a consumer.
Phase 3
The same pattern and time frame as utilized for the silicone baby cups was followed for the sous vide bags. As with the baby cups, significantly higher gains were observed for the DCAS-informed advertisement, even though it was seen less often (Table 8). Per 1,000 impressions (and assuming a normal distribution), the DCAS-informed sous vide ad was clicked on at a significantly higher rate (Moriginal = 6.19 vs. MDCAS-informed = 9.84; F(1, 50) = 72.345, p < .001), led to a significantly higher number of orders (Moriginal = 1.24 vs. MDCAS-informed = 1.88; F(1, 50) = 22.347, p < .001), and generated significantly higher sales revenue (USD; Moriginal = $20.77 vs. MDCAS-informed = $31.87; F(1, 50) = 23.215, p < .001). The DCAS-informed ad generated a 58.97% increase in clicks per 1,000, a 51.61% increase in orders per 1,000, and a 53.44% increase in average sales revenue per order.
Discussion
The participating Amazon retailer was able to increase click-through rates (CTRs; i.e., engagement), orders, and sales revenues by using DCAS to develop superior advertisements without altering the product design. In this case, the DCAS dimensions that were known to elicit a significant, positive response were emphasized with the expectation that doing so would maximize potential gains, and this is what occurred. This demonstrates that DCAS can indicate which design dimensions can be altered to increase consumer engagement, and that implementing these findings can also translate into meaningful, practical gains. Further, by integrating market research while tracking sales metrics, firms can ascertain over time which design elements most strongly drive sales, allowing for real-time innovations or design improvements.
Study 2: Field Study—DCAS-Informed Product Concept Generates More Email Signups
The context of our second study mirrors common practices utilized for crowdfunded new product launches on sites like Kickstarter and Indiegogo. Specifically, we recruited a highly rated graphic designer on Fiverr to create two product concepts (an original and a DCAS-informed ad) that we then promoted for a week on Facebook. Following frequently used marketing practices for product concept launches, consumers who clicked on the observed product concept advertisement on Facebook were directed to a landing page that included an email signup option to be notified of product launch updates. Consumers only saw either the original ad or the DCAS-informed ad. As with Study 1, we executed this field study in phases.
Product: Coffee Maker
Phase 1
For our concept product, we chose a coffee maker because it is a familiar product category that has been used in design scale research (Homburg, Schwemmle, and Kuehnl 2015). We recruited a graphic designer from Fiverr to create a prelaunch coffee maker product concept that would generate consumer interest. The designer selected a coffee maker product image from a collection of stock images and designed a product concept advertisement for prospective use on Facebook (Figure A6).
Two hundred Prolific panel participants (46.5% male, 51.5% female, 2% nonbinary, 0% prefer not to say; Mage = 34.9 years) evaluated this concept with DCAS, responded to the six previously utilized behavioral intention measures, and provided demographic information (for full demographic information, see Web Appendix C, Table W5). A regression analysis revealed that function (β = .181, SE = .077, t(192) = 2.351, p = .020), style (β = .549, SE = .113, t(192) = 4.883, p < .001), eco-consciousness (β = .196, SE = .085, t(192) = 2.297, p = .023), and uniqueness (β = .155, SE = .063, t(192) = 2.468, p = .014) had significant, positive effects on the indexed purchase intention questions (r = .90). Similar effects emerged for style, eco-consciousness, and uniqueness for WOM (r = .90), WTP, and attitude, but not for function (Web Appendix C, Table W6).
Consequently, we had our designer develop a new product concept that focused on style, eco-consciousness, and uniqueness, with particular emphasis on the latter two dimensions since they had more room for improvement (means of 3.64 and 3.88, respectively, compared with 4.57 for style). The designer selected a different stock image (i.e., form) for the coffee maker that they deemed as more stylish and unique and redesigned the ad background (introduction of leaves and color change to green) to cue eco-consciousness. Even though this redesign involved changing the product form, this change remains within the parameters of this type of campaign (i.e., the same functionality was retained and there were no finalized form restrictions as in Study 1). See Figure A6 for the revised concept.
Phase 2
A new set of 200 Prolific panelists (46.0% male, 51.5% female, 1.5% nonbinary, 1% prefer not to say; Mage = 20.3 years) evaluated the DCAS-informed concept using the same measures as in Phase 1. We compared the Phase 1 and 2 data sets using a MANOVA (F(11, 388) = 5.459, p < .001; Wilks' Λ = .87) with product concept seen as the independent variable and the seven dimensions of DCAS and the behavioral intention measures as dependent variables. This analysis revealed significantly higher results for style and uniqueness, and a directional improvement for eco-consciousness (Table 9). Lower evaluations for functionality also emerged, which aligns with our theorizing that a strong deviation from a product category prototype (i.e., higher uniqueness) leads to more difficult discernibility (i.e., lower function).
Field Study: Coffee Maker Product Concept.
*Significant at .05. **Significant at .01. ***Significant at .001.
Rather than impressions, this is the number of unique consumers exposed to a product concept.
The total number of unique clicks on the product concept leading to the associated website.
Phase 3
For Phase 3, both concepts were run on Facebook simultaneously for one week, with consumers seeing only one of the product concepts. When a consumer clicked on a concept, they were directed to a unique website landing page where they could input their email address for future product updates (Web Appendix C, Figures W4 and W5). Both product concepts were targeted to people residing in the United States over the age of 18 using Meta's A/B testing feature, which delivers the concept images to approximately equal audiences, allowing marketers to test communication effectiveness. In total, our concept images were shown to 24,128 people. The DCAS-informed concept significantly outperformed the initial concept even though it was shown to fewer people by Facebook (original concept's reach = 12,464 vs. DCAS-informed concept's reach = 11,664). A z-test shows significantly higher clicks (Moriginal = 329 vs. MDCAS-informed = 418; z = −4.231, p < .001) and email signups (Moriginal = 29 vs. MDCAS-informed = 51; z = −2.762, p = .006) for the DCAS-informed version. See Table 9 for full results. Note that this is a conservative test of DCAS’s ability, as the designer's original product concept generated a substantially higher CTR than is typically observed. Prior research has shown that CTRs are usually below 1% (e.g., Kupor and Laurin 2020; Tormala, Jia, and Norton 2012), yet the original product concept CTR was already well above the average, at 2.64%, and the DCAS-informed concept CTR was even higher, at 3.58%.
Discussion
An experienced designer initially created a high-quality product concept advertisement that performed substantially better than typical promoted content on Facebook, yet a new, DCAS-informed ad was able to further enhance performance by eliciting significantly more email signups. The recruitment of an independent designer, who used DCAS-based recommendations to create a superior product concept, demonstrates the unique ability of DCAS to facilitate designer–marketer interactions in the aid of enriching design communications.
Together, the studies in Stage 3 establish the ecological ability of DCAS to help marketers coordinate with designers in the improvement of design communications. As done here, practitioners should generally focus on adjusting no more than three dimensions at a time when using DCAS input. This adaptive improvement allows a designer to keep the core of a design intact, avoiding the time and cost associated with a complete redesign.
Stage 4: Validation of a Shortened Form of DCAS
Using DCAS in its full state (21 items, 3 items per dimension) is most appropriate for design testing, especially for academic research (Böckenholt and Lehmann 2015; Haws, Sample, and Hulland 2023; Netemeyer, Bearden, and Sharma 2003). However, practitioner settings may call for a shortened version of DCAS due to resource considerations. To provide for this, we examined the loadings of all items across data sets to determine the best individual item to capture each dimension (see Table 4 for the loadings across data sets and Table 10 for the chosen items).
Shortened DCAS Validation.
Significant at .1. *Significant at .05. **Significant at .01. ***Significant at .001.
Indicates reverse-coded.
Notes: Item order: form, function, usefulness, solidity, style, eco-consciousness, and uniqueness.
In alignment with prior research providing shortened scale validation (e.g., Rahman et al. 2022; Seiders et al. 2007), we used prior data sets to test this shortened version. We estimated a hierarchical linear regression comparing the performance of the selected seven items with the full version of DCAS using data from Data Set 4, Studies 1 and 2. In all cases, a significant amount of variance is explained by the condensed DCAS (all R2s > .35, p < .001; Table 10). These validation efforts indicate that the shortened version of DCAS can be relied on in resource-constrained settings.
General Discussion
Despite the relevance of design in today's marketplace, existing marketing research tools do not systematically and reliably assess the effectiveness of product design communications. We address this limitation by employing long-standing design theories to generate a conceptual framework for assessing design communications. In addition to the two traditional intrinsic dimensions (form and function), we incorporate five extrinsic dimensions (solidity, usefulness, style, eco-consciousness, and uniqueness) into our conceptual framework.
Based on this framework, we developed a measurement scale following recommended scale development procedures (Netemeyer, Bearden, and Sharma 2003; see Table 3 for this summary). A thorough literature review, interviews with world-renowned and award-winning designers, and empirical work generated seven dimensions and a list of 140 preliminary items subsequently winnowed to a 21-item scale (3 items per dimension) based on EFA, CFA, and face validity assessments (see Stage 1). In Stage 2, we tested DCAS and found satisfactory discriminant, nomological, and predictive validity as well as temporal stability, indicating that DCAS was ready for implementation.
In Stage 3, we applied DCAS in meaningful and practical studies. In Web Appendix C, Lab Studies 1 and 2, DCAS is shown to provide insights that can lead to positive changes for product designs and packaging. In the field studies, we demonstrate how DCAS can be used to effectively identify both the determinant dimensions of design communications and specific areas of potential weakness. This allows designers to make adjustments that increase managerially relevant outcomes such as CTRs (engagement), orders, and email signups. Finally, in Stage 4, we develop a shortened version of DCAS, primarily for resource-limited practitioner settings. In summary, we have generated a diagnostic, predictive, reliable, and valid scale to measure consumer response to design communications resulting in several theoretical and practical contributions.
Theoretical Contributions
Theoretically, we take a large step toward coalescing design and marketing thinking. Though important design elements have been considered in prior marketing literature (e.g., Bloch, Brunel, and Arnold 2003; Homburg, Schwemmle, and Kuehnl 2015), these perspectives have been inconsistent with or too distal from design theory, leaving a communication gap between marketers and designers. We address this issue by combining two centuries-old design theories into a cohesive framework. The current research is the first to connect the form follows function principle with the Vitruvian Triad. We do so to conceptually integrate the foci of designers and consumers with the objective of enhancing design-related communication effectiveness.
Additionally, while prior marketing research has touched on dimensions of design that are both intrinsic and extrinsic (Homburg, Schwemmle, and Kuehnl 2015), the present work is the first to delineate these and define the roles that they play in consumer evaluations. Whereas intrinsic and extrinsic dimensions are collectively used by consumers to assess designs, they are employed differentially. The intrinsic dimensions are fundamental in the assessments of product designs, as they inform whether the consumer accurately recognizes the product as a member of its product category and whether it meets consumer expectations. In contrast, the extrinsic dimensions are used in assessing the consumer-specific benefits of product designs, as they relate to comparable alternatives within the product category.
Managerial Implications
DCAS is a cost-effective and dynamic instrument equally suited for use by small businesses and large corporations. DCAS highlights communication strengths and/or deficiencies to marketing managers and facilitates better communication with designers by using familiar design language. In practice, managers or designers can use DCAS (following similar steps as outlined in the field studies) to identify which design dimensions have the most impact and room for improvement, and then adjust them appropriately. In general, we recommend adjusting no more than two or three DCAS dimensions at a time. This ensures greater control and predictability. Changing more than three dimensions at once will hamper comparisons across design iterations. Furthermore, as prior research has demonstrated (e.g., Liu et al. 2017; Luchs et al. 2010), attempts to strengthen some dimensions will inevitably weaken others, so numerous adjustments can be more detrimental than beneficial.
Future Research
The insights provided herein lay a fertile framework for future research. For instance, even though the intrinsic dimensions are more salient, there are most likely times when a product is too stylish or unique for a consumer to ignore. Thus, when or why will a specific dimension override the influence of the other dimensions for a consumer? Further, there are numerous other potential moderators, which can influence evaluations on the whole and potentially specific dimensions. Personality traits, emotions, perceived scarcity, or even luxuriousness could have distinct properties that influence and interact with various DCAS dimensions.
Extensions, reductions, or adjustments could be made to DCAS, depending on the design type of interest (e.g., retail, brands, logos, web). Being grounded in pervasive, time-tested thoughts from the design field, dimensions should overlap, but adjustments to items and dimensions may be needed. Furthermore, we have only demonstrated this scale with visual stimuli, but as noted, style is an appeal to any of the senses. Thus, how might DCAS be applied to evaluations of designs that appeal to other senses, such as perfumes or wines?
Conclusion
Our objective was to produce a scale that could assess consumer evaluations of product design communications, thereby enabling greater guidance regarding adjustments for product design and communication success. We accomplished this by integrating established design theories with marketing insights. Through the course of this manuscript, we demonstrate the reliability and validity of DCAS and its effectiveness in facilitating designer adjustments via consumer feedback. DCAS makes an important contribution by being the first design assessment tool to demonstrate the versatility and ability necessary to assess a wide range of design communications (product, package, advertisement) and boost positive outcomes such as orders, sales, CTRs, and email signups. We provide substantial theoretical and practical contributions to the marketing discipline while addressing the crucially important issue of further connecting the design and marketing practices (e.g., Dahl 2011; Luchs and Swan 2011). Finally, we provide evidence as to how DCAS can be effectively utilized by any firm to help increase the likelihood of success; for firms with resource restrictions, we provide a shortened version of DCAS. In summary, we hope that the introduction of our theoretically and technically rigorous scale, integrating insights from design theorists, design experts, and marketing researchers, can facilitate a more accurate, comprehensive, and profitable understanding of design.
Supplemental Material
sj-pdf-1-mrj-10.1177_00222437231166342 - Supplemental material for The Design Communication Assessment Scale (DCAS): Assessing and Adjusting the Effectiveness of Product Design Communications
Supplemental material, sj-pdf-1-mrj-10.1177_00222437231166342 for The Design Communication Assessment Scale (DCAS): Assessing and Adjusting the Effectiveness of Product Design Communications by Kevin L. Sample, John Hulland, Julio Sevilla and Lauren I. Labrecque in Journal of Marketing Research
Footnotes
Appendix
Coeditor
Vikas Mittal
Associate Editor
Dhruv Grewal
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
Notes
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
