Consumers’ responses to product design: Using a Semantic Priming Task to assess automatic understanding of product positioning

Abstract

Lack of understanding of new products’ positioning is one of the reasons proposed for their failure. Through a process of semantic transformation, product design can communicate a new product’s positioning to consumers. Drawing on the theoretical foundations of implicit cognition and the results of an empirical study, this article demonstrates that the Semantic Priming Task is a valuable tool to evaluate perceived product positioning conveyed by product design and to help practitioners in their decision-making with regard to product design.

Keywords

positioning product design Semantic Priming Task

Introduction

This article aims to investigate whether an implicit measure can be used to assess automatic understanding of product positioning that is conveyed by product design. Positioning is the act of building a credible, attractive, and distinctive place in the consumer’s mind (Fuchs & Diamantopoulos, 2012; Ries & Trout, 1986). It is an offshoot of differentiation, which is the approach by which a firm aims to develop unique products that are clearly distinguishable from similar offerings on the market (Porter, 1985; Wang, 2015). Effectively positioning a product in the minds of consumers has been identified as one of the strategies that favors market acceptance for a new product (Fuchs & Diamantopoulos, 2012). From a managerial point of view, positioning refers to showing or communicating how a product compares with other products (Fuchs & Diamantopoulos, 2012). From this perspective, decisions on how to communicate product positioning can be taken after product development has taken place, especially in cross-national contexts where the same product can be positioned in different ways because the importance of attributes may differ across countries (Aaker & Joachimsthaler, 1999; Malhotra & Bartels, 2002). However, product positioning may also play a key role in determining the overall product development strategy, especially with regard to product design choices (Creusen & Schoormans, 2005; Fuchs & Diamantopoulos, 2012).

Product design is the first element seen by consumers and is responsible for the product’s first assessment, as Hollins and Pugh (1990) clearly found: “whatever the product, the customers see it first . . .. The physical performance comes later, the visual always comes first” (p. 91). Product design is defined here at the final output of a multidisciplinary creative and manufacturing process, embodied in a product’s physical attributes and visible characteristics (Bloch, 1995; Veryzer, 1995). The past literature has suggested that product design can be used to gain a competitive advantage because product esthetics capture consumer attention, generate positive emotional reactions, and have a positive effect on perceived quality (Kreuzbauer & Malter, 2005; Radford & Bloch, 2011). In addition, and this is the focus of this research, product appearance may also be used to communicate the central consumer advantage to consumers (Creusen & Schoormans, 2005). Thus, the most critical positioning attributes should be the starting point in the design of product appearance (Creusen & Schoormans, 2005).

Whereas early works on product design have mostly focused on esthetic responses (Berkowitz, 1987; Holbrook, 1986), more recent work has aimed to understand the influence of design semantics on consumption behaviors (Dell’Era, Marchesi, Verganti, & Zurlo, 2008; Karjalainen & Snelders, 2010; Veryzer, 1993, 1995). Design semantics refers to the idea that product design conveys meaning through users’ interpretations of signs that are embodied in the products (Karjalainen & Snelders, 2010). Thus, using product design to position a product involves selecting the physical and formal attributes that physically translate the positioning that the marketing department aims to build in consumers’ minds. From that perspective, consumers’ response to product design should be understood as one stage in a communication process (Krippendorff & Butter, 1984).

However, this communication process may fail to deliver the appropriate meaning, and this failure may result from two major distortions: first, the designer may fail to embed proper meanings into the product and second, users may not succeed in correctly decoding the meanings embedded in the product design (Karjalainen, 2007). In other words, consumers’ perceived positioning may differ from marketers’ intended positioning. To limit the risks that product design fails to communicate the intended product positioning, it should be verified that product design leads to the consumer’s perception of product positioning in the direction intended by marketers.

Measures currently used to verify the alignment between product perceived positioning and product intended positioning (i.e., the absence of the distortions previously mentioned here) are qualitative and quantitative explicit measures. These measures have the features of typical self-reported measures: awareness of evaluating the product design, intention to evaluate, control over the evaluation, and deliberation in making the evaluation (Smith & Nosek, 2011). Even time pressure exerted on respondents during a free elicitation task does not totally eliminate reflection and verbalization biases. However, recent research has suggested the existence of dual evaluative processes by which two cognitive systems interact to produce evaluations: one system is automatic, associative, and nonconscious and the other is deliberative and controlled (Van Bavel, Xiao, & Cunningham, 2012).

In the specific case of product design, Veryzer, in his seminal 1999 paper, proposed a nonconscious processing explanation of a consumer’s response to product design and called for future research in this area. From that perspective, the use of explicit measures to control whether product design successfully communicates the intended product positioning is problematic because it only involves controlled reflection processes and overlooks the substantial nonconscious and automatic understanding of product design meaning. We suggest that such automatic understanding cannot be captured through explicit methods and that it requires implicit measures: implicit measures are indirect measures that do not inform the subject of what is being assessed and that tap memory associations in an automatic manner (Ackermann & Mathieu, 2015).

This article aims to adapt and apply existing implicit measures to assess the implicit understanding of product positioning that is conveyed by the design of the product. In the next section, we briefly review the literature on product design and we introduce implicit measures. After we explain how we develop a methodology from this review and we outline our results. Finally, we discuss the implications and contributions of this study and offer ideas for subsequent studies.

Theoretical background

How product design conveys product positioning

Design semantics refer to how meaning is mediated by the signs designers embed in products so the signs will be recognized and interpreted by consumers (Karjalainen & Snelders, 2010). Through a process of semantic transformation, designers aim to transform qualitative attributes into design features. For instance, designers purposely created the specific U or Y shapes of early Nokia phones to be interpreted as a friendly smile, which supports the brand identity of personalization and a human approach (Karjalainen & Snelders, 2010). In this way, the choice of colors, forms, and materials enables designers to translate product positioning into specific design features. It is at this stage that the first distortion (i.e., a failure to encode proper meanings) may occur (Karjalainen, 2007). Then, when exposed to the design of a product, consumers have the opportunity to perceive it (i.e., to attribute an interpretation to the formal attributes the designers embedded in the product to convey its positioning). The second distortion (i.e., a failure to attribute meaning to design features in the way expected by the designer) may occur (Karjalainen, 2007) at this point. Figure 1 illustrates how product positioning can be mediated by design features designers embed in products and consumers perceive when exposed to them.

Figure 1.

Semantic transformation and attribution.

A cognitive approach

Product design enables designers to deliver to consumers a message regarding the positioning of the product that consumers are able to understand (i.e., to decode product positioning on the sole basis of its design) only if consumers have stored in their memory some associations between specific design features, such as a specific shape or a given color, and qualitative attributes. For example, a consumer is able to interpret the use of stainless steel material or polished surfaces as signals of durability only if the consumer has learned to associate these specific design features with this attribute. In the same way of thinking, consumers have learned to associate angular forms with dynamism and masculinity, roundness with softness and femininity, and, in the automobile category, bright colors with aggression (Creusen & Schoormans, 2005).

Building on the Human Associative Memory model (Anderson, 1983), we suggest that consumers have developed a network of memory associations between design features and attributes as the result of a learning process. The Human Associative Memory model describes human memory as a network of interconnected memory nodes, which are basic elements that constitute a piece of information stored in a person’s mind. Such nodes can be viewed as memory traces of previous learning episodes (episodic memory models, Versace et al., 2014). According to the recent Act-In model, “memory traces reflect all the components of past experiences and, in particular, their sensory properties as captured by our sensory receptors, actions performed on the objects in the environment and the emotional and motivational states of individuals. . .. Memory traces are therefore distributed across multiple neuronal systems which code the multiple components of the experiences” (Versace et al., 2014, p. 282). During information retrieval, the emergence of specific knowledge comes from an interactive and integrated activation of these traces.

One of the main aspects of the Human Associative Memory model is spreading activation (Anderson & Pirolli, 1984; i.e., the fact that the activation of a specific node spreads throughout the network to nodes to which is connected). The activation of a memory node primes the activation of the entire memory network in which it is embedded, and this priming effect occurs outside of conscious awareness. In line with this, we suggest that the perception of a design feature automatically activates the conceptual attributes to which it is associated in memory without the individual being aware of it. Thus, the different design features of a given product—colors, shapes, materials, texture, and symbols—provide the basis for the perception of particular product characteristics. According to the Human Associative Memory model, this process occurs in an unintentional and nonconscious manner. As stressed by Veryzer (1999), “just as people may implicitly learn the rules of an artificial grammar and then apply them without conscious awareness, rules governing the processing . . . of product designs may be nonconsciously acquired and applied” (pp. 503–504). Recent models developed by economists provide theoretical arguments and empirical results that demonstrate the existence of two systems (Evans, 2003; Kahneman, 2011; Kahneman & Frederick, 2007). Two-system theories have been proposed to explain phenomena, such as perception (e.g., Goodale & Humphrey, 1998) and memory (e.g., Schacter & Tulving, 1994). They are based on the idea that there are different tracks of thought that constitute two distinct mental modes: System 1 is automatic, fast and unconscious, while System 2 is slow, effortful and conscious (Kahneman, 2011). Thus, the perception of particular product characteristics automatically activating conceptual attributes in memory would belong to System 1 processing, which implies specific protocols to be assessed. The purpose of this article is not to discuss the distinction between two-system or unique system (continuum) models (for a theoretical and empirical discussion, see Keren & Schul, 2009; Osman, 2004). Rather, building on Penn (2016), we argue that, as market researchers, we should be aware of the fact that implicit mind is something consumers are not aware of, but influence a lot their consumers’ decision-making, including product design evaluation. That is the reason why specific protocols should be used, not only to assess implicit attitudes, but also implicit semantic associations are automatically activated by being visually exposed to a product.

Explicit versus implicit measures

In light of the importance of consumers’ correct understanding of product positioning, product managers must evaluate the extent to which a new product is successfully positioned in the minds of consumers (Fuchs & Diamantopoulos, 2012). As suggested by the previously summarized literature, product design is a tool that enables marketers to communicate on the positioning of a product. Thus, it is critical that marketers control whether product design, in isolation of any other information about the product, enables consumers to position it correctly. However, and very surprisingly, little attention has been paid so far as to how the effectiveness of product positioning should be measured (Fuchs & Diamantopoulos, 2012). And the question of how product design enables consumers to automatically understand product positioning has received even less attention.

A review of the literature shows that two methods are currently used for measuring positioning success: (1) elicitation of positioning and (2) comparison with competing products among positioning dimensions (Wang, 2015). Elicitation techniques are qualitative techniques that aim to investigate which attributes come to consumers’ minds when exposed to a given product. Comparison techniques are perhaps the most popular approaches for measuring positioning success (Fuchs & Diamantopoulos, 2012) and involve the comparison of the positioning of a product against that of competing products on a number of a priori specified attributes in the attribute space (Fuchs & Diamantopoulos, 2012; Wang, 2015). These methods are derived from methods such as correspondence analysis, discriminant analysis, or multidimensional scaling (MDS; Fuchs & Diamantopoulos, 2012; Green, Carroll, & Goldberg, 1981; Wang, 2015) and use perceptual maps to reveal the relative positioning of the products in terms of attributes, preferences, or overall (dis)similarity to each other.

However, like all explicit measures, these qualitative and quantitative methods require respondents to deliberately engage in a reflection process. In a classic article, social psychologists Nisbett and Wilson (1977) reported from many experimental demonstrations that verbal reports on our own behavior often reflect reconstructive and interpretative processes, rather than genuine introspection. In market research, dissociations between conscious awareness, actual behavior, and memory have been reported (e.g., Calvert, Fulcher, Fulcher, Foster, & Rose, 2014; Penn, 2016; Rivière, Cuny, Allain, & Vereijken, 2013; Werle & Cuny, 2012). They demonstrated differences between memory that can be expressed explicitly as a conscious recollection and implicitly as automatic, unconscious influences on behavior. One underlying hypothesis to explain these differences builds on the idea that explicit measures are direct measurements (i.e., they make explicit reference to the relevant variable under study). Consequently, this will (1) place emphasis on the processes that are accessible (automatic processes are not accessible) and (2) be vulnerable to verbalization biases (e.g., positivity bias, socially desirable responding, and contextual cueing [Cunningham & Zelazo, 2007; Fazio & Olson, 2003; Gregg, Klymowsky, Owens, & Perryman, 2013]). In other words, the use of explicit measures cannot tap whether a semantic attribute, such as one of the dimensions of the positioning of the product, has been automatically activated in memory by exposure to the design of the product.

We argue that implicit measures may provide a more proximal estimate of how design features automatically activate meaning than is possible with explicit measures. As a consequence, implicit measures represent a more objective, although indirect, measurement of consumers’ perceived positioning and could be very effective to compare perceived and intended positioning. Implicit measures require participants to classify words or pictures into categories, and categorization speed is observed to assess the strength of memory associations. Thus, they block the control loop (Cunningham & Zelazo, 2007) that “prompts respondents to over-validate their responses” (Rivière et al., 2013, p. 377) in an explicit questionnaire and thereby gain access to automatic associations in memory.

Following Rivière et al. (2013), we build on Cunningham and Zelazo (2007) to argue that implicit and explicit measures assess the two endpoints of an iterative reflection process. Through their iterative reprocessing (IR) model, Cunningham and Zelazo (2007) proposed that evaluations are constructed through an iterative reprocessing of stimuli. When exposed to a given stimulus, associations existing in memory with that stimulus are automatically activated and provide the basis of an initial evaluation. Then, this initial evaluation is reprocessed iteratively during a process of reflection that produces more nuanced evaluations. However, the automatic process underlying the initial evaluation is active during the different iterations and influences the subsequent, more reflective evaluations. In other words, the process of reflection does not supplant the automatic process. It is just the opposite: automatic and reflective processes interact to generate evaluations that result from existing memory associations, but also from other processes that consumers use to process information, such as motivation, time pressure, or the presence of others (Rivière et al., 2013).

In our field of investigation, this implies that when a consumer is exposed to the design of a product, the meaning associated in memory with some specific design features is automatically activated. An explicit method, be it qualitative or quantitative, and requires consumers to reflect on the attributes they associate with the design of the product will prompt a reflective process similar to that described by Cunningham and Zelazo (2007). Thus, an explicit measure will reflect the outcome of the consumer’s reflection process prompted by the need to verbalize an answer; however, an implicit measure offers an immediate assessment of the ability of the design features to automatically activate meaning.

Semantic priming paradigm and the lexical decision task

Among the implicit measures, the implicit association test (IAT) has gained considerable support because of its validity and reliability properties and has been the most widely used in consumer behavior research (Ackermann & Mathieu, 2015; Dimofte, 2010; Gregg et al., 2013). The IAT measures differences in associative strength between two concept categories and two attribute categories. This makes it an appropriate tool in many situations in which there is an intent to compare two competing brands, a product and its challenger, or two opposite product categories. But this tool may not be optimal for marketing studies that aim to investigate associations with a single brand, product, or product concept (Rivière et al., 2013).

In contrast, priming tasks are implicit measures that aim to measure the strength of the relationship between a prime stimulus and a target. A Semantic Priming Task is a lexical decision task associated with a semantic priming paradigm: participants are first primed with a stimulus and then exposed to a target, word, or pseudo-word about which they have to make a lexical decision (Meyer & Schvaneveldt, 1971). A priming effect is observed if the time needed to categorize the target stimuli is shorter when primes and target stimuli are semantically related than when they are not. Although Semantic Priming Tasks are commonly used in research on memory as a means to explore automatic relations among concepts, their use in business research remains scarce (Rivière et al., 2013). However, their extensive use in other research fields provides support for their validity.

Our aim is to measure the extent to which exposure to a product design conveys meaning (i.e., the extent to which exposure to physical features activates semantic nodes in memory). This makes the Semantic Priming Task highly suitable because it may investigate the extent to which exposure to a prime (i.e., the design of the product) facilitates the categorization of attributes as words or pseudo-words. The ease with which an attribute will be recognized as a word depends on how strongly it is semantically associated in memory with some of the design features of the product.

Figure 2 illustrates the underlying mechanism, which is as follows: exposure to the design of the product (i.e., the prime) automatically activates in memory conceptual nodes to which some specific design features of the product are semantically associated; the activation level of these conceptual nodes is temporarily increased; the activation of these conceptual nodes facilitates the categorization as words or pseudo-words of attributes (i.e., the target) that are associated with them; and the degree of facilitation induced by the initial exposure to the product design is the measure of the associative strength between the product design and the attributes (Figure 3).

Figure 2.

Response Times as a way to measure the strength of association between design features and attributes.

Figure 3.

Semantic Priming Task: organization of sequences.

To conclude, we propose the following:

Traditional explicit measures cannot tap the automatic activation of attributes by exposure to product design;

Associative strength between attributes and product design can be measured by a Semantic Priming Task; and

Associations between attributes and product design that are measured with a Semantic Priming Task differ from self-reported associations due to verbalization biases.

Method

Participants and research approach

A total of 368 French native-speaking undergraduate students (62.2% females; 37.8% males) enrolled in a marketing course completed the experiment as part of a course requirement.The experiment was presented to them as a new product pretest, and it was preceded by a comprehension test. Indeed, implicit measures are indirect measures; this implies that respondents should not be informed of the purpose of the measurement (Greenwald & Banaji, 1995).

To test the validity of our theoretical propositions, we sought a product category with which participants would be somewhat familiar but in which they would be very unlikely to have expert knowledge. We selected the Oral B TriZone electric toothbrush. Participants first completed a Semantic Priming Task and then answered an explicit questionnaire. We measured the level of involvement in the category of electric toothbrushes by using Zaichkowsky’s (1994) Personal Involvement Inventory (PII). The PII level was inferior to 4(x̅ = 3.10, t (357) = −11.474, p < .001), indicating a low level of personal involvement in the category of electric toothbrushes.

We analyzed the Oral B website to identify the words that were used to describe the TriZone’s functional and symbolic attributes. We identified the attributes that were most used to describe TriZone, and we made the assumption that these attributes would correspond to the intended product positioning. The symbolic attributes we identified were designed,¹ esthetic, elegant, and sophisticated. The functional attributes we identified were performant, efficient, soft, dynamic, reliable, simple, and good quality. These attributes were used both in the Semantic Priming Task to measure implicit perceived product positioning and in the explicit questionnaire to measure explicit perceived product positioning.

Design and procedure

We used the E-Prime software to develop and administer the Semantic Priming Task. Participants were seated in front of a personal computer (PC) and required to indicate as quickly as possible whether the letter string that appeared on the screen represented a word or pseudo-word by pressing the appropriate key on the PC keyboard.

A semantic Priming Task consists of distractive sequences and test sequences. The aim of the distractive sequences is to prevent participants from understanding the real purpose of the test, whereas the aim of the test sequences is to measure the strength of the association between a focal stimulus and focal targets. Each sequence starts with the presentation of a stimulus (the prime), followed by the presentation of a target about which participants have to make a lexical decision.

The primes were pictures from three different Oral B electric toothbrushes, and one of them was the TriZone. The name and logo of the original brand were removed from the three stimuli, and the pictures were of the same size, quality, and shooting angle. Stimuli were presented on a neutral white background. The targets were 22 French words and 22 pseudo-words. The 22 French words included the 11 product attributes identified on the Oral B website that we believed capture the TriZone positioning and 11 other words. The 44 targets (i.e., letter strings corresponding to words or pseudo-words) and the 3 primes (i.e., electric toothbrushes pictures) were combined to form 88 couples (prime-targets), which combined in one block of 66 distractive sequences and one block of 22 test sequences. Each participant reviewed the 88 couples. Table 1 illustrates how primes and targets were combined to form the 88 prime-targets couples. The test sequences involved the presentation of a picture of the TriZone electric toothbrush followed by the 11 product attributes capturing TriZone positioning. Conversely, the distractive sequences used pictures of 1—the TriZone electric toothbrush followed by attributes not corresponding to TriZone positioning or pseudo-words and of 2—other Oral B electric toothbrushes followed by real words. The presentation orders of the distractive and test sequences were randomized for each participant.

Table 1.

Semantic Priming Task: test sequences and distractive sequences.

Prime	Target	Sequence type
TriZone	22 pseudo-words	Distractive
	11 product attributes capturing TriZone positioning	Test
	11 other product attributes	Distractive
Design 1	11 product attributes capturing TriZone positioning	Distractive
Design 1	11 other product attributes	Distractive
Design 2	11 product attributes capturing TriZone positioning	Distractive
Design 2	11 other product attributes	Distractive

Each sequence (1) started with the presentation of a fixation point in the center of the screen for a duration of 500 ms, (2) immediately followed by a 200 ms presentation of a priming stimulus, then a (3) 100 ms presentation of a mask composed of a row of five hash-masks (#####) was presented, and this was followed by (4) the presentation of a target letters-string that represented a word or pseudo-word. Participants had to answer as quickly as possible by pressing the “1” key if the string of letters was a real French word or the “2” key if it was not. The target disappeared as soon as the participant answered. The inter-sequence interval was set to 1500 ms. Response time (RT) was measured from the target onset until the participant’s response.

The 200 ms presentation of the prime guaranteed that its trace in memory would be relatively inaccessible to a conscious report. The stimulus onset asynchrony (SOA; i.e., the interval between the presentation of the prime and the presentation of the target) was set at 300 ms because this SOA level guarantees significant priming effects (e.g., Fazio, Sanbonmatsu, Powell, & Kardes, 1986).

After the Semantic Priming Task, participants completed a short questionnaire and provided demographic data, such as their ages and genders. Explicit product positioning was measured using semantic differential items anchored with 5-point scales, ranging from “not” to “very” on the 11 attributes.

Results

RTs shorter that 250 ms or longer than 1500 ms (Hermans, Houwer, & Eelen, 2001) were not retained for data analysis; this reduced the sample for the statistical analyses to 357 participants. We calculated the mean RT for each prime-target couple when the target was 1 of the 11 attributes identified as capturing the TriZone product positioning, though we included only correct responses in our statistical analyses. RT captures the design-attribute associative strength; a shorter RT indicated stronger associations between the prime (i.e., the design) and the target (i.e., the attribute). An attribute is suggested to be part of the TriZone implicit perceived positioning if a significantly shorter RT occurs in the test sequence (i.e., when the respondent indicates that an attribute is a real French word after seeing TriZone) compared with the distractive sequences (i.e., the attribute preceded by the design of another electronic toothbrush: design 1 or design 2).

We applied an analysis of variance (ANOVA) to the explicit and implicit data with product design as the within-subjects variable (see Table 2). The ANOVA revealed that the three prime designs led to a significantly different average RT to determine that each of the 11 target attributes is a real French word. The ANOVA for explicit answers also revealed significant differences between the three designs for 8 out of the 11 attributes.

Table 2.

Relationships between product design and attributes.

Attributes		Mean (SD)					Mean comparison Tukey test
Attributes		TriZone	Design 1	Design 2	F	p	Mean comparison Tukey test
Designed	Implicit	723 (255.21)	617 (241.05)	923 (150.50)	177.55	<0.001	Design 2 > TriZone > design 1
Designed	Explicit	3.28 (1.09)	4.52 (1.14)	2.97 (1.74)	2.25	<0.001	Design 1 >TriZone > design 2
Esthetic	Implicit	718 (228.98)	608 (248.78)	924 (150.50)	199.87	<0.001	Design 2 > TriZone > design 1
Esthetic	Explicit	3.08 (1.12)	4.80 (1.52)	1.75 (2.14)	1.72	<0.001	Design 1 > TriZone > design 2
Elegant	Implicit	715 (250.93)	600 (249.96)	851 (142.48)	116.12	<0.001	Design 2 > TriZone > design 1
Elegant	Explicit	2.93 (1.23)	3.43 (1.09)	2.43 (1.93)	3.81	<0.001	Design 1 > TriZone ns design 2
Sophisticated	Implicit	841 (189.15)	820 (83.07)	946 (125.30)	84.02	<0.001	Design 2 > TriZone ns design 1
Sophisticated	Explicit	3.94 (0.90)	3.89 (0.82)	3.91 (0.89)	12.16	0.70	TriZone ns design 1 ns design 2
Performant	Implicit	668 (260.98)	570 (252.65)	875 (191.02)	154.04	<0.001	Design 2 > TriZone > design 1
Performant	Explicit	3.98 (0.92)	4.06 (1.02)	3.26 (1.04)	3.20	<0.001	TriZone ns > design1 > design 2
Efficient	Implicit	643 (202.46)	557 (241.85)	801 (173.37)	126.66	<0.001	Design 2 > TriZone > design 1
Efficient	Explicit	3.98 (0.91)	3.96 (1.02)	2.43 (0.96)	1.14	<0.001	TriZone ns design 1 > design 2
Soft	Implicit	741 (645.17)	540 (279.29)	938 (716.06)	80.25	<0.001	Design 2 > TriZone > design 1
Soft	Explicit	3.28 (1.04)	3.89 (1.54)	1.78 (1.68)	1.79	<0.001	TriZone ns design 1 > design 2
Dynamic	Implicit	636 (223.70)	561 (245.05)	773 (197.69)	237.23	<0.001	Design 2 > TriZone > design 1
Dynamic	Explicit	4.08 (0.74)	3.21 (1.29)	1.54 (1.33)	6.01	<0.001	TriZone > design 1 > design 2
Reliable	Implicit	647 (223.80)	827 (188.86)	761 (182.23)	74.61	<0.001	Design 1 > design 2 > TriZone
Reliable	Explicit	3.75 (1.00)	3.66 (1.07)	3.78 (0.98)	15.27	0.34	TriZone ns design 1 ns design 2
Simple	Implicit	655 (199.05)	854 (61.83)	627 (210.43)	186.94	<0.001	Design 1 > design 2 ns TriZone
Simple	Explicit	3.57 (1.19)	3.41 (1.36)	3.53 (1.21)	13.98	0.15	TriZone ns design 1 ns design 2
Good Quality	Implicit	547 (197.97)	961 (84.76)	575 (195.05)	692.27	<0.001	Design 1 > design 2 ns TriZone
Good Quality	Explicit	4.07 (1.08)	3.45 (1.75)	1.13 (1.01)	3.98	<0.001	TriZone > design 1 > design 2

The results suggest that design 1 is the most differentiated design, with the attributes “designed,” “esthetic,” “elegant,” “performant,” “efficient,” “soft,” and “dynamic” being more implicitly associated to design 1 than to the TriZone design and design 2. At the explicit level, design 1 also generated stronger associations with the different attributes than the two other designs, but no significant differences were observed between design 1 and the TriZone regarding “performant,” “efficient,” and “soft.” Very interestingly, out of the 11 attributes we identified, only “good quality” appears as the attribute that is mostly associated with the TriZone design at both the explicit and implicit levels.

In addition, we performed correlation analyses (see Table 3) to identify attributes for which associations or dissociations between the TriZone explicit and implicit perceived product positioning are observed; dissociation is understood here as the signature of reflection and verbalization biases. The only attribute for which no dissociation is observed is the attribute “reliable”: the shorter the RT for “reliable” is (i.e., the stronger the implicit association), the stronger “reliable” is explicitly associated to the TriZone design. The TriZone is therefore clearly positioned on the attribute “reliable.” Conversely, an implicit explicit dissociation is observed for the attributes “designed,” “esthetic,” and “elegant”: for each of these attributes, the higher the RT (i.e., the weaker the implicit association), the stronger the attribute is explicitly associated to the TriZone design.

Table 3.

Correlation analyses.

Explicit measures
Implicit measures	1	2	3	4	5	6	7	8	9	10	11
1—Designed	0.157**
2—Esthetic		0.150**
3—Elegant			0.125*
4—Sophisticated				−0.084
5—Performant					−0.063
6—Efficient						−0.048
7—Soft							−0.018
8—Dynamic								−0.018
9—Reliable									−0.157**
10—Simple										0.005
11—Good quality											−0.006

p < .01.

p < .05.

Finally, to illustrate the difference between implicit and explicit perceived product positioning, we used a semantic differential method (SDM) to develop two perceptual maps of the TriZone positioning. First, principal components analyses with varimax rotation confirmed the distinction between functional and symbolic attributes at both the explicit and implicit levels, and two factors were extracted, one corresponding to the symbolic dimension of product positioning and the other corresponding to the functional dimension. Then, a factorial analysis of correspondence was performed on the RT and explicit answers to map the TriZone implicit and explicit perceived product positioning. This provided a visual illustration of differences between implicit and explicit perceived product positioning (see Figure 4).

Figure 4.

Perceptual maps of explicit and implicit product positioning.

Discussion

With this experiment, we attempted to develop and test a Semantic Priming Task that can be used in the context of product positioning and its materialization into a product design. We propose that (1) designers use formal codes to translate the positioning that has been defined by marketers, and (2) consumers can automatically understand marketers’ intended positioning because these formal codes are associated in memory with qualitative attributes. We propose that the results of the explicit measures traditionally used to control for consumers’ perceived positioning are biased by reflection and verbalization processes. Thus, explicit measures cannot tap the associations between some specific design features and qualitative attributes that have been automatically activated by the exposure to the product, and the sole utilization of explicit measures overlooks consumers’ nonconscious processing of product design.

Our aim was not to establish the superiority of one type of measure, implicit or explicit, over the other. It was to show that they offer different results that are complementary. As highlighted by Rivière et al. (2013), the radically different structures of implicit measures versus explicit self-reported measures enable the collecting of different data related to the same topic. When responding to an explicit questionnaire, respondents have ample time to scrutinize and to interpret the design of the product. In contrast, implicit measures tap associations that are automatically activated by the exposure to the design of the product. Thus, if the two measures are to offer similar results, this suggests that the measurement context, time allocated to the reflection process, and individual characteristics do not shift the deliberate evaluation away from the automatic evaluation. In contrast, a difference between the two types of measures suggests that consciously processing product design shifts the deliberate evaluation away from the automatic evaluation. A divergence between implicit and explicit results should not be understood as contradictory, which would question the validity of one of the two measures; rather, they should be seen as complementary in the sense that they highlight two different points within the iterative reprocessing of the stimulus (i.e., the design of the product) to which consumers are exposed. Differences between explicit and implicit measures suggest that the reflection process leads respondents to “review” their automatic interpretation of product design, the “revision” process being nonconscious.

Our research has clear implications for practitioners. We suggest that a combination of implicit and explicit measures offers a better measure of perceived product positioning than the measure provided by the sole use of an explicit measure. Combining implicit and explicit measures may help practitioners make decisions during the product development process because controlling perceived product positioning implicitly and explicitly may identify attributes that are explicitly, but not implicitly, associated to the design of a product and vice versa. Qualitative techniques may be used to further investigate the nature of this apparent “conflict” and thus provide a better understanding of design semantics (i.e., of how consumers perceive meaning conveyed by product design).

Implementing a Semantic Priming Task, although relatively easy, remains more time-consuming than developing and administering an explicit questionnaire, especially online. It requires a specific expertise in the protocol development, thus an additional cost if implicit and explicit testing have been both chosen to be used to provide complementary results. Even if some practitioners are already using adaptations of implicit tests for commercial purposes (e.g., investigation of semantic networks, Rivière et al., 2013; understanding of consumers’ emotional engagement with brands, Calvert et al., 2014; exploration of brand attitudes, Gregg et al., 2013; barriers to brand usage, Penn, 2016), these financial and time constraints may make marketers reluctant to use implicit methods. Nevertheless, in our field of investigation, that is, automatic understanding of product positioning, we suggest that there are some situations in which the use of a combination of implicit and explicit measures may be very useful. For example, low involvement often characterizes the purchase of fast-moving consumer goods because consumers do not dedicate much time to understanding the product, and this therefore stresses the importance of controlling the extent to which its design automatically conveys the product’s positioning. Similarly, processing new products also requires more cognitive efforts than processing existing products because unfamiliar stimuli are more difficult to handle and there is more to learn from them (Pieters, Warlop, & Wedel, 2002). Again, if consumers are not highly involved in the evaluation process, they may be reluctant to dedicate time and effort to understanding the new product. This, again, stresses how critical it is to control implicit perceived product positioning because implicit processes do not require cognitive efforts. Finally, any situation where traditional methods have shown their limitations; for example, when promising prelaunch evaluations fail to predict poor post-launch sales, this is an appropriate situation for the combined use of explicit and implicit measures.

Finally, our research only recruited student as the sample. The cognitive process at stake in a Semantic Priming Task, that is, categorization, is a process that is only slightly influenced by social variables such as occupation, incomes or marital status (Wu & Lin, 2006). Thus, the use of a student sample should neither critically reduce the external validity of our results, nor undermine the reliability of the methodology presented in our article. However, sampling is a matter of concern in any research related to product positioning, and it stands to reason that marketers who would like to use a Semantic Priming Task to measure consumer’s associations with product design should recruit a sample that would be representative of the targeted population.

Conclusion

By applying a Semantic Priming Task to measure consumer’s associations with product design, we aimed to demonstrate that (1) associations between attributes and product design that are activated by exposure to the product can be measured by a Semantic Priming Task and (2) associations between attributes and product design measured with a Semantic Priming Task differ from self-reported associations due to verbalization biases. The results support our views and contribute to further establishing the validity of the Semantic Priming Task in marketing studies by showing its relevance in the context of product design and perceived product positioning.

Indeed, of the various new implicit measures (see Dimofte, 2010 for a detailed review), the most widely used are the Implicit Association Task (Greenwald, McGhee, & Schwartz, 1998) and the Evaluative Priming Task. However, despite the relative ease of implementing the latter, there are few marketing studies that have used implicit measures other than the IAT, with the exception of one study using the Go-No Go Association Task (Spence & Townsend, 2007), and some have used the Evaluative Priming Task (e.g., Calvert et al., 2014; Dempsey & Mitchell, 2010; Penn, 2016; Rivière et al., 2013; Werle & Cuny, 2012). Table 4 presents the different implicit methods referenced in the article. From a conceptual standpoint, one of the main limitations of the IAT is that it is a relative measure (Ackermann & Mathieu, 2015), which makes sense in many managerial situations. However, there are also many cases where the researcher wishes to investigate a unique concept, and finding an “opposite” concept is a matter of a pure methodological constraint. From that perspective, the Evaluative Priming Task is a promising tool for market research because it enables measuring the strength of automatic associations with a single target, be it a brand, a product, a product category, or any concept relevant to the understanding of consumption behaviors. As suggested by Rivière et al. (2013), the “systematic use” of this method “could lead to insightful results that are both scientifically valid [as supported by extensive psychological research] and complementary to the findings of explicit tests” (p. 388).

Table 4.

Presentation of the different implicit methods referenced in the article.

Implicit measure	Description	Main references
Semantic Priming Task	Participants complete a series of sequences consisting of a prime presentation (that appears very quickly, sometimes subliminally) followed by the presentation of a target. The task requires participants to categorize the targets as words or pseudo-words as quickly as possible. A priming effect is observed if the time needed to categorize the target stimuli is shorter when primes and target stimuli are semantically related than when they are not. The degree of facilitation induced by the presentation of the prime is the measure of the semantic association between the prime and the target.	Meyer and Schvaneveldt (1971)
Affective Priming Task	In an Affective Priming Task, targets are words that are either positively or negatively valenced. A priming effect is observed if the time needed to categorize the target stimuli is shorter when primes and target stimuli share the same valence than when they are not. The degree of facilitation induced by the presentation of the prime is the measure of the attitude toward the prime.	Fazio et al. (1986)
Implicit Association Test	Participants complete a series of categorization tasks involving distinguishing stimuli belonging to two target categories, i.e., concept A and concept B, and to two contrasted attribute categories, e.g., “bad” and “good.” To perform these tasks, participants must press as quickly as possible a button to the left of a computer keyboard for stimuli belonging to one concept and one attribute category and a button to the right of the keyboard for stimuli belonging to the other concept and attribute categories. An implicit positive attitude toward concept A is observed if the task is completed more quickly when stimuli related to concept A and positive attributes share the same keyboard button than when stimuli related to concept A and negative attributes share the same keyboard button.	Greenwald et al. (1998)
Go-No Go Association Task	A GNAT (Go-No Go Association Task) requires participants to “go,” i.e., press the space bar when presenting to stimuli that belong to a target or evaluative attribute category (signal), and to “not go,” i.e., not press on any key, for stimuli that do not belong to the target or attribute category (noise). Strength of association is assessed by the degree to which stimuli belonging to the target and attribute category (signals) can be discriminate from items that do not belong to these categories (noise).	Nosek and Banaji (2001)

Footnotes

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Notes

References

Aaker

D. A.

Joachimsthaler

(1999). The lure of global branding. Harvard Business Review, 77, 137–144.

Ackermann

C. L.

Mathieu

J. P.

(2015). Implicit attitudes and their measurement: Theoretical foundations and use in consumer behaviour research. Recherche et Applications en Marketing, 30(2), 58–81.

Anderson

J. R.

(1983). A spreading activation theory of memory. Journal of Verbal Learning and Verbal Behavior, 2, 261–295.

Anderson

J. R.

Pirolli

P. L.

(1984). Spread of activation. Journal of Experimental Psychology: Learning, Memory, and Cognition, 10, 791–799.

Berkowitz

(1987). Product shape as a design innovation strategy. Journal of Product Innovation Management, 4, 274–283.

Bloch

P. H.

(1995). Seeking the ideal form: Product design and the consumer response. Journal of Marketing, 59(3), 16–29.

Calvert

Fulcher

Foster

Rose

(2014). Using implicit methods to develop an objective measure of media brand engagement. International Journal of Market Research, 56, 15–32.

Creusen

M. E. H.

Schoormans

J. P. L.

(2005). The different roles of product appearance in consumer choice. Journal of Product Innovation Management, 22, 63–81.

Cunningham

W. A.

Zelazo

P. D.

(2007). Attitudes and evaluations: A social cognitive neuroscience perspective. TRENDS in Cognitive Sciences, 11, 97–104.

10.

Dell’Era

Marchesi

Verganti

Zurlo

(2008). Language mining: Analysis of the innovation of dominant product languages in design-intensive industries. European Journal of Innovation Management, 11, 25–50.

11.

Dempsey

M. A.

Mitchell

A. A.

(2010). The influence of implicit attitudes on choice when consumers are confronted with conflicting attribute information. Journal of Consumer Research, 37, 614–625.

12.

Dimofte

C. V.

(2010). Implicit measures of consumer cognition: A review. Psychology & Marketing, 27, 921–937.

13.

Evans

J. S. B. T

. (2003). In two minds: Dual process accounts of reasoning. Trends in Cognitive Sciences, 10, 454–459.

14.

Fazio

R. H.

Olson

M. A.

(2003). Implicit measures in social cognition research: Their meaning and use. Annual Review of Psychology, 54, 297–327.

15.

Fazio

R. H.

Sanbonmatsu

D. M.

Powell

M. C.

Kardes

F. R.

(1986). On the automatic activation of attitudes. Journal of Personality and Social Psychology, 50, 229–238.

16.

Fuchs

Diamantopoulos

(2012). Customer-perceived positioning effectiveness: Conceptualization, operationalization, and implications for new product managers. Journal of Product Innovation Management, 29, 229–244.

17.

Goodale

M. A.

Humphrey

G. K.

(1998). The objects of action and perception. Cognition, 67, 181–207.

18.

Green

P. E.

Carroll

J. D.

Goldberg

S. M.

(1981). A general approach to product design optimization via conjoint analysis. Journal of Marketing, 45(3), 17–27.

19.

Greenwald

A. G.

Banaji

M. R.

(1995). Implicit social cognition: Attitudes, self-esteem and stereotypes. Psychological Review, 102, 4–27.

20.

Greenwald

A. G.

McGhee

D. E.

Schwartz

J. L. K.

(1998). Measuring individual differences in implicit cognition: The Implicit Association Test. Journal of Personality and Social Psychology, 74, 1464–1480.

21.

Gregg

A. P.

Klymowsky

Owens

Perryman

(2013). Let their fingers do the talking? Using the Implicit Association Test in market research. International Journal of Market Research, 55, 487–503.

22.

Hermans

Houwer

Eelen

(2001). A time course analysis of the affective priming effect. Cognition and Emotion, 15, 143–165.

23.

Holbrook

M. B.

(1986). Aims, concepts, and methods for the representation of individual differences in esthetic responses to design features. Journal of Consumer Research, 13, 337–347.

24.

Hollins

Pugh

(1990). Successful product design. London, England: Butterworths.

25.

Kahneman

(2011). Thinking, fast and slow. New York: Farrar, Strauss and Giroux.

26.

Kahneman

Frederick

(2007). Frames and brains: Elicitation and control of response tendencies. Trends in Cognitive Sciences, 11, 45–46.

27.

Karjalainen

T. M.

(2007). It looks like a Toyota: Educational approaches to designing for visual brand recognition. International Journal of Design, 1, 67–81.

28.

Karjalainen

T. M.

Snelders

(2010). Designing visual recognition for the brand. Journal of Product Innovation Management, 27, 6–22.

29.

Keren

Schul

(2009). Two is not always better than one: A critical evaluation of two-system theories. Perspectives on Psychological Science, 4, 533–550.

30.

Kreuzbauer

Malter

A. J.

(2005). Embodied cognition and new product design: Changing product form to influence brand categorization. Journal of Product Innovation Management, 22, 165–176.

31.

Krippendorff

Butter

(1984). Product semantics: Exploring the symbolic qualities of form. Innovation, 3(2), 4–9.

32.

Malhotra

N. K.

Bartels

B. C.

(2002). Overcoming the attribute prespecification bias in international marketing research by using nonattribute-based correspondence analysis. International Marketing Review, 19, 65–79.

33.

Meyer

D. E.

Schvaneveldt

R. W.

(1971). Facilitation in recognizing pairs of words: Evidence of dependence between retrieval operations. Journal of Experimental Psychology, 90, 227–234.

34.

Nisbett

R. E.

Wilson

T. D.

(1977). Telling more than we can do: Verbal reports on mental processes. Psychological Review, 84, 231–259.

35.

Nozek

B. A.

Banaji

M. R.

(2001). The Go/No-Go association task. Social Cognition, 19(6), 625–664.

36.

Osman

(2004). An evaluation of dual process theories of reasoning. Psychonomic Bulletin & Review, 11, 988–1010.

37.

Penn

(2016). True lies and true implicit: How priming reveals the hidden truth. International Journal of Market Research, 58, 175–199.

38.

Pieters

Warlop

Wedel

(2002). Breaking through the clutter: Benefits of advertisement originality and familiarity for brand attention and memory. Management Science, 48, 765–781.

39.

Porter

M. E.

(1985). The competitive advantage: Creating and sustaining superior performance. New York, NY: Free Press.

40.

Radford

S. K.

Bloch

P. H.

(2011). Linking innovation to design: Consumer responses to visual product newness. Journal of Product Innovation Management, 28, 208–220.

41.

Ries

Trout

(1986). Positioning: The battle for your mind. New York, NY: McGraw-Hill.

42.

Rivière

Cuny

Allain

Vereijken

(2013). Digging deeper: Using implicit tests to define consumers’ semantic network. International Journal of Market Research, 5, 375–390.

43.

Schacter

D. L.

Tulving

(1994). What are the memory systems of 1994? In Schacter

D. L.

Tulving

(Eds.), Memory systems (pp. 1–38). Cambridge, MA: MIT Press.

44.

Smith

C. T.

Nosek

B. A.

(2011). Affective focus increases the concordance between implicit and explicit attitudes. Social Psychology, 42, 300–313.

45.

Spence

Townsend

(2007). Predicting behaviour towards genetically modified (GM) food using implicit and explicit attitudes. British Journal of Social Psychology, 46, 437–457.

46.

Van Bavel

J. J.

Xiao

Y. J.

Cunningham

W. A

. (2012). Evaluation is a dynamic process: Moving beyond dual system models. Social & Personality Psychology Compass, 6, 438–454.

47.

Versace

Vallet

G. T.

Riou

Lesourd

Labeye

É.

Brunel

(2014). Act-in: An integrated view of memory mechanisms. Journal of Cognitive Psychology, 26, 280–306.

48.

Veryzer

R. W.

(1993). Esthetic response and the influence of design principles on product preferences. Advances in Consumer Research, 20, 224–228.

49.

Veryzer

R. W.

(1995). The place of product design and aesthetics in consumer research. Advances in Consumer Research, 22, 641–645.

50.

Veryzer

R. W.

(1999). A nonconscious processing explanation of consumer response to product design. Psychology & Marketing, 16, 497–522.

51.

Wang

H. J.

(2015). A new approach to network analysis for brand positioning. International Journal of Market Research, 57, 727–742.

52.

Werle

C. O. C.

Cuny

(2012). The boomerang effect of mandatory sanitary messages to prevent obesity. Marketing Letters, 23, 883–891.

53.

L. L.

Lin

J. Y.

(2006). The quality of consumers’ decision-making in the environment of e-commerce. Psychology & Marketing, 23, 297–311.

54.

Zaichkowsky

J. L.

(1994). Research notes—The personal involvement inventory: Reduction, revision, and application to advertising. Journal of Advertising, 23(4), 59–70.