consumer perceptions of sales pressure

J. of the Acad. Mark. Sci. DOI 10.1007/s11747-015-0468-z

METHODOLOGICAL PAPER

An offer you can’t refuse: consumer perceptions of sales pressure James J. Zboja 1 & Ronald A. Clark 2 & Diana L. Haytko 3

Received: 16 January 2014 / Accepted: 22 May 2015 # Academy of Marketing Science 2015

Abstract High-pressure sales tactics may result in short-term increases in sales; however, the long-term effect of employing these tactics on the relationships among the customer, the salesperson, and the firm is generally negative. Despite this widely held view, research on sales pressure has been limited to examining the selling orientation of the salesperson and/or the firm. As such, consumer perceptions of sales pressure have been neglected, and how the consumer experiences and reacts to pressure is not well understood. A structural model of antecedents and outcomes of perceptions of sales pressure is tested using a quota sample of consumers. To facilitate this test, the authors develop and validate a scale to measure consumer perceptions of sales pressure. The resulting scale consists of two factors, aggressive sales pressure and directive sales pressure. While aggressive sales pressure results in negative satisfaction and trust outcomes, directive sales pressure does not. These results underscore the assertion that some perceived pressure on the part of the buyer, assuming there is a directive element present, may not result in negative results for the salesperson and firm.

* Diana L. Haytko [email protected] James J. Zboja [email protected] Ronald A. Clark [email protected] 1

Department of Management & Marketing, Collins College of Business, University of Tulsa, Tulsa, OK 74104-9700, USA

2

Department of Marketing, College of Business, Missouri State University, Springfield, MO 65897, USA

3

Department of Marketing, Lutgert College of Business, Florida Gulf Goast University, Fort Myers, FL 33965, USA

Keywords Sales pressure . Coercion . Influence . Persuasion . Scale development . Structural equation modeling

Marketing practitioners and researchers alike understand the salesperson’s power to influence the consumer’s decision of whether, what, and when to buy (Olshavsky 1973). The very act of closing a sale, or the selling profession in general, necessitates that the salesperson “exercise some degree of influence over the buyer” (Brown 1990, p. 18). Indeed, the argument has been made that if “low pressure” selling means that there is no attempt to influence the buyer at all, this method, “unless supplemented by positive action of some kind, would not be selling at all” (Bursk 2006, p. 3). Given today’s climate of consultative selling, the salesperson’s objective should be to achieve an outcome that is positive for both the seller and the buyer. However, a salesperson that employs selling pressure as part of his or her persuasion tactics risks changing the nature of the sales message from persuasive to coercive (Barron and Staten 1995). In the sales context, a persuasive sales message is viewed as more positive, focusing on the product’s ability to meet the goals of the purchaser with credible arguments, whereas a coercive sales message is more negative, using aggressive and manipulative arguments to compel the purchase. Although the short-term benefits of high-pressure tactics may often lead to quick sales for the selling firm, the results of employing pressure tactics can be injurious to the consumer, salesperson, and retailer (Chu et al. 1995). Chu et al. define “hard-sell” as a “situation in which to an average consumer the annoyance is greater than the expected benefit” (1995, p. 97) and provide examples of aggressive techniques such as bait and switch selling, manipulative presentations, creating a false sense of urgency, and overstating the product’s merits. Sales training guru Nick Moreno at the National Sales Center

J. of the Acad. Mark. Sci.

(2009) echoes these sentiments, stating that sales pressure will only lead to failure. A study of retail shoppers found that, among 38 cited shopping irritants, high-pressure selling was rated as most irritating among the consumers surveyed (d’ Astous 2000). Despite the opinions of many selling practitioners and researchers that a high-pressure approach is bad for all involved parties, there is relatively little research on the factors that influence the customer’s perceptions that pressure tactics are being employed and the consumer response to such pressure. In fact, according to Chu et al. (1995, p. 102), “hard-sell may be in the eye of the beholder; one person’s hard-sell is another person’s friendly customer service”. As such, defining and measuring “hard-sell” is difficult, since each consumer experiences it differently. In addition, the literature on interpersonal influence in psychology, communications, management and marketing (including channels research) focuses on the strategies used by persuasion agents to influence targets (see Kellerman and Cole 1994 for a review), with little attention paid to the target of persuasion. Interest in relationship marketing (and CRM) in the channels literature has led to studies of key constructs from the customer perspective such as trust, commitment, and satisfaction. However, these constructs have been mostly studied from the firm perspective rather than the interpersonal/individual perspective (with the exception of boundary spanners) and in longer-term relationships. Cialdini and Trost concluded, “we have much to learn by viewing the influence process from the perspective of the target” (1998, p. 151). This is particularly true of the individual salesperson– consumer relationships found in retail settings characterized by one-time transactions. Though some research suggests that the hard-sell is detrimental to all consumers, Chu et al. (1995) found that, from a societal perspective, the benefits of improved product matching to consumers outweigh the negative impact of consumer annoyance. The purpose of our study, then, is twofold. First, using the theoretical context of Friestad and Wright’s (1994) Persuasion Knowledge Model and subsequent work by Campbell and Kirmani (2000; 2004), this paper tests a structural model of potential antecedents and outcomes of consumer perceptions of sales pressure. Our data reveal that there are instances where sales pressure does not necessarily have a negative effect on outcomes, contradicting past research suggesting that all sales pressure is negative. Second, in order to test the hypotheses, we conceptualize, develop, and validate a scale for the measurement of consumer perceptions of sales pressure, using an updated procedure based on Churchill’s (1979) scale development paradigm. We address the issue of past research ignoring the perspective of the target in the persuasion process, as suggested by Cialdini and Trost (1998), Kirmani and Campbell (2004), and Arndt et al. (2014). The target is treated as an active rather than passive participant in the selling interaction. Developing a way to

measure felt pressure will be a step toward developing other measures of the target’s experience as an active member of the dyad. Contributions of this research include the following: (1) hypothesizing and empirically testing a structural model of consumer perceptions of sales pressure, including antecedents and outcomes of the construct, (2) through this process, creating the first measure of consumer perceptions of sales pressure (POSP), comprised of two dimensions: aggressive and directive, which treat the consumer as an active participant in the dyad (extending the work of Kirmani and Campbell 2004 and Arndt et al. 2014), and (3) updating and applying a revised Churchill (1979) paradigm to include recommendations by leading scale development researchers. We conclude the paper with theoretical and managerial implications, limitations and directions for future research.

Background High-pressure sales tactics, often referred to as the “hard-sell,” are sometimes used by salespeople to coerce the consumer into making a purchase (Barron and Staten 1995). Most consumers have likely felt pressure in a buying situation at some point in time. A recent survey by AutoTrader magazine surveyed 1657 Millennials, 993 Gen Xers and 1062 Baby Boomers who had bought a vehicle within the last 3 months. They found that Millennials are painstaking when it comes to online and in-store vehicle research, but they dislike highpressure sales tactics. Millennials said they were much more likely than their older counterparts to avoid dealership salespeople in the future. Fifty-six percent said so, compared with 49% for Gen X and 37% for Boomers. Clearly, no matter what the consumers’ age, car salespeople are not providing an enjoyable buying experience (Automotive News 2013). It’s not just automotive salespeople; senior citizens are often contacted with offers for financial management “free lunches,” college students are constantly offered credit cards with special promotional rates, and nearly everyone has been cold-called by timeshare and vacation plan companies. Among the various definitions of high-pressure selling in the literature are “persuading the prospect to buy something he can’t afford, that is not suited to his needs, that he does not want, and that will leave him dissatisfied after the salesperson leaves” (Bursk 2006, p. 2). High-pressure tactics have been criticized for a multitude of reasons. First, the average consumer feels these tactics cause discomfort with the purchase experience (Chu et al. 1995). When the customer perceives that high-pressure tactics are being used, the practice is viewed as unfair, manipulative or unethical (Brown 1990; Star 1989; Trawick and Swan 1988). The consumer’s perception of the employment of high-pressure tactics can even result in the consumer doing the opposite of what the salesperson intended


because of a perceived threat to their freedom of choice (Clee and Wickland 1980). Ultimately, the benefits of using highpressure tactics to salespeople and their firms are usually outweighed by the costs (Chu et al. 1995). Indeed, empirical evidence suggests that coercive tactics can only result in a slight increase in short-term sales (Barron and Staten 1995). Hartman (2005, p. 74), in a comprehensive study of influence strategies, concluded that “pressure is consistently an effective strategy for producing managerially undesirable consequences”. Therefore, it is imperative for salespeople and sales managers to understand what increases the consumer’s perception that they are being pressured and the outcomes of this perception. It is also vital to improve our understanding of consumerperceived sales pressure to determine if there are any potential positive outcomes, situations that warrant its application or optimal levels of sales pressure and, if so, where the line between positive pressure and negative pressure lies. Scheer and Stern (1992, p. 139) found that favorable outcomes from a dominating influence can “sometimes be more critical to the improvement, maintenance, or decline of a relationship than the way in which behavior is motivated”. This would suggest that some applied pressure is acceptable from the perspective of the consumer, so long as the product ultimately purchased as the result of the influence episode met needs and expectations. The Persuasion Knowledge Model (PKM) was introduced by Friestad and Wright (1994, p. 1) as a means to examine how consumers “develop and use persuasion knowledge to cope with persuasion attempts”. The PKM posits that persuasion attempts of the salesperson (i.e., agent) on the consumer (i.e., target) and the consumer’s corresponding coping behaviors are critical in determining the outcome of the persuasion episode. The model indicates that both the target and agent enter a persuasion attempt with different levels of knowledge in the following areas: topic knowledge, agent/target knowledge, and persuasion knowledge. It is important to note that persuasion in this model is represented by an individual’s global interpretation, including, as people do tend to oscillate between the two, both target and agent roles. Thus, this consumer behavior theory examines both sides of the dyad in terms of motives, and it takes into account the salesperson’s tactics versus the awareness of these tactics by the consumer. The use of the term coping behaviors implies that persuasion targets are “resourceful participants who pursue their own goals and have the ability to select response tactics from their own repertoire, akin to the way agents select persuasion tactics” (Friestad and Wright 1994, p. 3). Coping in this sense is neutral in that it does not specify a positive or negative response, and includes cognitive behavior leading up to a persuasion episode. The PKM has been used as a guiding framework in subsequent studies covering a wide range of topics, such as the

effects of pricing tactics (Hardesty et al. 2007; Kachersky 2011; Kachersky and Kim 2011; Pillai and Kumar 2012),the effects of charitable giving (Hibbert et al. 2007), and the impact of rhetorical questions (Ahluwalia and Burnkrant 2004). Recently, much research has focused on the use of persuasion knowledge among the millennial generation (Henrie and Taylor 2009) and among children (McAlister and Cornwell 2009; Wright et al. 2005). The application of PKM to the sales context has most notably been accomplished by Campbell and Kirmani (2000) and Kirmani and Campbell (2004). These authors assert that identifying factors that either inhibit or encourage the use of persuasion knowledge is a critical step in its further development. Kirmani and Campbell (2000) manipulated persuasion motives to determine when consumers would use their persuasion knowledge. Their findings suggest that when persuasion motives are less accessible, the use of persuasion knowledge involves higher order reasoning that requires cognitive capacity. They conclude the following: “in sales interactions that are commonly believed to involve high-pressure persuasion, such as negotiating the purchase of a car, persuasion knowledge may be highly accessible before the encounter, thereby helping consumers cope with the expected persuasion attempt” (p. 81). The experiments reported in this study were conducted in lab settings, using one type of influence agent, a role-playing situation, and student subjects. Yet, the suggested conclusions provide motivation and support for the current study to examine a consumer’s experience of felt pressure and the resulting outcomes. Mallalieu (2006, p. 257) claims that “understanding consumer perceptions is an important precursor to understanding consumer behavior during retail sales encounters”. The most recent use of the PKM in sales (Arndt et al. 2014) treats the consumer as an active participant and examines consumer responses to the use of credibility-building statements across the stages of the encounter through a multi-stage qualitative analysis. This work extends other work that has been done that studies consumer issues, such as Mallalieu (2006), a study of how the consumers’ goals affect consumer perceptions of salesperson influence strategies with data collected from three experiments. Gremler and Gwinner (2008) examined rapport-building behaviors used by retail employees using the critical incident technique to collect data from consumers about specific incidents. Each of these studies examines sales encounters from the consumer side of the dyad using very different methodologies, which all support the PKM assertion that the consumer is an active (rather than passive) participant. In its application to this research, the elements of the PKM lend themselves well. In the case of a buying situation, the target (consumer) and the agent (salesperson) both enter into the episode with their own personal levels in the areas of topic (product) knowledge, agent/target knowledge (each of the other), and persuasion knowledge. The agent enters into the


persuasion attempt with predetermined expectations regarding some combination of the level of persuasion to be attained, the amount of pressure to exert, and the resultant desirable outcome(s). The target, as a result of the persuasion attempt, preconceived notions prior to the event, and level of pressure perceived, will then (to a varying degree) apply the knowledge they hold to cope with the attempt in what they feel to be an appropriate manner. One of the key objectives of this research is to ascertain conditions that can impact a consumer’s level of perceived sales pressure (e.g., the degree of perceived salesperson knowledge, the consumer’s persuasion knowledge self-confidence, and time pressure) and, perhaps more importantly, how outcomes (e.g., satisfaction and trust) are impacted by these perceptions.

Consumer perceptions of sales pressure In an effort to begin the process of conceptualizing the construct of perceptions of sales pressure, we conducted an exploratory study. The data were collected as a project for an upper-level undergraduate marketing course at a U.S. university in the Midwest. Similar to other recent scale development data collection approaches used (Brocato et al. 2012; Walsh and Beatty 2007), each student completed an open-ended survey about a shopping incident in which they felt pressured and recruited additional participants via quota sampling instructions. A total of 118 surveys were completed (eight were removed because the subject could not recall a high-pressure incident). The key demographics of the sample are as follows: mean age 36, 61% male, 49% some college (28% college degree), and 74% Caucasian (19% African American). Similar to the critical incident technique (CIT) process (Flanagan 1954), the participants were asked the following: Please describe in detail a recent (within the last year) experience when you, as a retail consumer, felt pressured in a buying situation. In particular, please describe events leading up to the experience in which you felt pressured, and describe exactly what happened in the interaction. Be sure to also include specifics such as what type of business/product for which you were shopping, etc. Finally, describe your reactions, as well as those of the other party, to the encounter. This qualitative process was intended to simultaneously provide input for a conceptualization of the construct and provide additional insight relative to the potential antecedents and outcomes associated with sales pressure. Consistent with the PKM, the common themes found were assigned to the following categories: the persuasion episode agent, the persuasion episode target, and the persuasion episode situation.

Among the primary themes found in the qualitative data was consumers’ sense of felt pressure based on “an inability to browse” and feeling as if they were not given enough space (25% of respondents), resulting in discomfort with the heightened amount of attention they were receiving during the episode. Recognition of tactics being utilized by the salesperson, consistent with the PKM, was cited by 22% of the respondents. Chief among these tactics was a high-pressure/early close attempt. The third most frequently mentioned theme (cited by 19%) revolved around the common sales practice of up-selling and cross-selling. These respondents reported being pushed toward products and/or additional services they did not want. Interestingly, 6% of respondents reported feeling pressure due to helpful, quality service provided by the salesperson. That is, they felt an obligation to buy (or buy even more than planned). Adding to the potential import of this finding is the fact that a surprisingly high percentage of respondents (32%) reported that they purchased the product for which they were shopping despite the perceived pressure. Based on both the theoretical background provided by the PKM and the exploratory qualitative study, we conceptualize the construct of consumer perceptions of sales pressure as follows: A consumer’s perception of being pressured by a sales agent to purchase a product within an artificially urgent time frame and/or being directed by the sales agent towards the purchase of different or additional product alternatives. In concert with the findings of the qualitative portion of the study, the theoretical justification for the dimensionality of the perceptions of sales pressure construct is derived from Kirmani and Campbell’s (2004) extension of the PKM. In this research, the consumer was theorized to have two distinct roles in a persuasion episode. The first role is that of a “persuasion sentry,” “one who guards against unwanted marketing persuasion” (p. 573). This role is represented in the first part of the conceptual definition; consumers universally do not approve of being rushed or pressured into a purchase before they are ready, usually resulting in negative outcomes for the marketer. We term this type of sales pressure “aggressive”. The second role is that of a “goal seeker,” using the agent to achieve the consumer’s own goals and objectives; in this case, buying the product that best meets his or her need. Despite the fact that pressure may still be felt, it is believed to be in the interest of finding the best product match to the need of the consumer (which may or may not have been what the consumer had originally thought). We term this type of sales pressure as “directive”. The target is believed to easily move between these two roles; so too, the consumer can therefore feel varying degrees of both types of pressure at the same time. An important distinction in our conceptualization of sales


pressure is that we seek to measure a consumer’s perception of sales pressure. It is the consumer’s perception that determines outcomes, not the salesperson’s intent (Brown 1990; Friestad and Wright 1994; Yoho 1998). That said; intentional highpressure tactics employed by the salesperson are certainly one possible means by which the consumer could feel pressured. While the construct to be defined exists within the consumer domain; it may be noted here, that the construct will likely be adaptable to organizational buyers as well.

Theoretical model The theoretical basis for the model to be tested is the PKM, which was first introduced by Friestad and Wright (1994). The topic knowledge and agent knowledge elements of the PKM are represented by the consumer-perceived product knowledge of their salesperson. Persuasion knowledge is represented by the consumer’s self-reported self-confidence in their persuasion knowledge. Added to the original elements of the PKM is the situational variable, time pressure, which would likely impact a consumer’s perceptions of sales pressure. Finally, the model’s two outcomes (see Fig. 1) include satisfaction with the salesperson and trust in the salesperson. Friestad and Wright (1994) indicate that the relative allocation of mental resources devoted to each of the three knowledge structures (persuasion, topic, agent) will vary for the target across persuasion episodes; as such, we expect that each knowledge structure’s impact on both types of perceived sales pressure will also vary. Given that the outcome (i.e., product purchased and ultimate assessment of suitability of the purchased

product) was already known to consumers prior to being surveyed, an underlying assumption is that the consumer perceived that salespeople generally had their best interest in mind with the recommendations that were offered.

Hypotheses According to Campbell and Kirmani (2000), the motive of salespeople to influence the consumer to make a sale or gain a commission is more accessible than other potential motives. Meanwhile, they found that, when this motive is highly accessible, consumers are likely to use persuasion knowledge to evaluate the salesperson. This use of persuasion knowledge takes place, not only during or after, but also before the persuasion episode. This prior activation of persuasion knowledge and perceived motive of the salesperson, and with it a degree of suspicion (Hilton et al. 1990) can impact a consumer’s perception of the salesperson’s credibility. A key component of credibility is expertise (e.g., Hovland et al. 1953) or competence (e.g., Sirdeshmukh et al. 2002). A consumer’s perception of their salesperson’s product knowledge is a proxy for competence or expertise, and it would likely have a negative impact on their perceptions of sales pressure. That is, ceteris paribus, the consumer should perceive less pressure from a salesperson who has a great deal of product knowledge. A salesperson who demonstrates expertise during the persuasion episode may be perceived positively by a consumer even if that consumer is suspicious of an underlying persuasion motive (Campbell and Kirmani 2000). Consistent with Kirmani and Campbell’s (2004)

Fig. 1 Conceptual model of perceptions of sales pressure

Perceptions of Sales Pressure (Directive)


“goal seeker” role, the salesperson perceived to be highly knowledgeable can be used by the customer to achieve his or her own goals. Furthermore, Clopton et al. (2001, p. 135) found, in a retail context, that “highly knowledgeable salespeople were judged to be more friendly, expert, helpful, and concerned”. Given these positive connections with salesperson product knowledge, we posit: H1a: Salesperson product knowledge has a negative influence on the consumer’s perception of aggressive sales pressure. H1b: Salesperson product knowledge has a negative influence on the consumer’s perception of directive sales pressure. Beatty and Ferrell (1998) found that time available for a shopping episode had a negative impact on a consumer’s negative affect; which, according to Watson et al. (1988) involves feelings of distress and other nonpleasurable mood states (e.g., anger, disgust, and fear). Maule and Svenson (1993) found that time pressure can cause worrisome thoughts, which could interfere with attention to relevant information, making the buying decision more stressful. This more difficult situation, in concert with its negative affect, would surely result in increased feelings of sales pressure for the consumer. H2a: A consumer’s level of time pressure has a positive influence on their perception of aggressive sales pressure. H2b: A consumer’s level of time pressure has a positive influence on their perception of directive sales pressure. Persuasion knowledge represents one of the six dimensions, and quite obviously the most vital one to this research, of Bearden et al. (2001) consumer selfconfidence scale. Within the dimensions of their scale, persuasion knowledge fits under the broad category of “protection”. Persuasion knowledge represents an individual’s confidence in his or her knowledge of tactics that can be used in a purchase episode and, ultimately, aids in coping and protecting the individual from harm that could be brought on by the perceived pressure applied by the salesperson. As theorized by the PKM, and further underscored by Campbell and Kirmani (2000), persuasion knowledge is often highly accessible even before the persuasion episode, particularly in situations where high-pressure tactics are stereotypically expected, thereby helping consumers cope with the expected persuasion attempt. Persuasion knowledge, in this case, plays the role of “persuasion sentry” (Kirmani and Campbell 2004), guarding against unwanted persuasion attempts. As a result of this higher level of coping ability, consumers are less likely to feel pressure during the interaction.

H3a: A consumer’s level of persuasion knowledge selfconfidence has a negative influence on their perception of aggressive sales pressure. H3b: A consumer’s level of persuasion knowledge selfconfidence has a negative influence on their perception of directive sales pressure. Satisfaction has been defined formally by Oliver (1997, p. 13) thusly: “Satisfaction is the consumer’s fulfillment response. It is a judgment that a product or service feature, or the product or service itself, provided (or is providing) a pleasurable level of consumption-related fulfillment, including levels of under- or over-fulfillment.” The importance of customer satisfaction, due to its impact on outcomes such as loyalty and repurchase intentions, is well documented in the literature (Brady and Cronin 2001; Mittal and Kamakura 2001). The importance of salesperson knowledge to customer satisfaction has been cited in the marketing literature (e.g., Agnihotri et al. 2009). Furthermore, Palmatier et al. (2006) cited seller expertise as having the greatest positive influence on customer-focused variables such as satisfaction. In an effort to replicate this widely reported relationship: H4a: A consumer’s perception of their salesperson’s knowledge has a positive influence on their satisfaction with the salesperson. Meanwhile, Beatty and Ferrell’s (1998) finding regarding the negative affect experienced by time-pressured consumers certainly impacts their level of satisfaction as well. Specifically, these authors state: “possessing limited time to shop, browse, or accomplish planned tasks could produce frustration and a negative reaction to the environment” (p. 176). This negative reaction could result in lower levels of satisfaction regardless of other factors at play in a given situation. Additionally, Haynes (2009) found that consumers with limited time were less satisfied with their purchase. These negative reactions could lead to similar attitudes toward the salesperson. H4b: A consumer’s level of time pressure has a negative influence on their satisfaction with the salesperson. It has been found that emotional reactions to a sales interaction (Babin et al. 1995) and fairness assessments (Oliver and Swan 1989) can affect customers’ satisfaction levels. Perceptions of sales pressure can certainly be seen as both sources of emotion and fairness concerns. Finally, Scheer and Stern (1992) found that satisfaction was strongly affected by influence types; more dominating influence types resulted in less positive attitudes. Additionally, perceiving aggressive sales pressure puts the


consumer firmly into their role of persuasion sentry, resulting in negative outcomes. H4c: A consumer’s perception of aggressive sales pressure has a negative influence on their satisfaction with the salesperson. In the case of directive sales pressure, however, the impact on satisfaction is likely less negative. Despite finding negative impacts on consumer attitudes from more dominating influence efforts, Scheer and Stern (1992) also found that favorable outcomes serve to mitigate the impact of influence type on negative attitudes. Consumer reactions to being directed to products other than the one for which they had originally planned to shop certainly can be negative. However, if they are potentially directed to a product that better suits their needs, the end result of the persuasion episode can be quite positive. The consumer takes on the role of goal seeker. That is, rather than coping in a negative manner to the directive persuasion attempt, the consumer uses the sales agent in an effort to accomplish his or her immediate goal (i.e., finding the right product). H4d: A consumer’s perception of directive sales pressure has a positive influence on their satisfaction with the salesperson. Along with customer satisfaction, trust has been found to be a key driver of consumer behavioral intentions (e.g., Doney and Cannon 1997; Sirdeshmukh et al. 2002). One of the more widely adopted definitions of trust in the literature is “a willingness to rely on an exchange partner in whom one has confidence” (Moorman et al. 1993, p. 82). It has been argued that, barring the experiences gained by being a repeat customer, individuals are innately able to assess trustworthiness of others during social exchanges (Buss and Kenrick 1998). These resultant observations of salesperson cues can then lead to trust of a salesperson via likeability and perceived expertise of the salesperson. Likewise, Belonax et al. (2007, p. 249) stated that “the research on expertise has generally been conducted in conjunction with the trust component of credibility”. Finally, in reporting generalizations from the marketing literature, the first common theme related by Swan et al. (1999) is that trust is supported by salesperson competence. H5a: A consumer’s perception of their salesperson’s knowledge has a positive influence on their trust in the salesperson. Once again, the negative affect that can come from time pressure (Beatty and Ferrell 1998) has been linked to many adverse mood states for the consumer. Among those mentioned are anger, disgust, guilt, and fear. Furthermore, these

negative mood states may decrease approach behaviors. Avoidance can often be attributed to a lack of trust; hence, a certain level of mistrust in the situation, and especially, the salesperson would appear likely. H5: A consumer’s level of time pressure has a negative influence on their trust in the salesperson. Scheer and Stern (1992) also found that more dominating influence types resulted in lower trust. Since perceived high-pressure sales tactics are certainly dominating in their approach, a consumer’s perception of aggressive sales pressure can negatively impact trust in the salesperson. H5c: A consumer’s perception of aggressive sales pressure has a negative influence on their trust in the salesperson. As with satisfaction, however, trust can be enhanced through the feeling that the salesperson potentially directed the consumer to a more appropriate product for his or her specific needs, despite initial feelings of pressure on the part of the consumer. H5d: A consumer’s perception of directive sales pressure has a positive influence on their trust in the salesperson.

Method Perceptions of sales pressure scale development Marketers need a way to measure the consumer’s perception of the pressure that can potentially result from a salesperson’s hard-sell tactics. It is, after all, the consumer’s perception that determines outcomes, not the salesperson’s intent (Brown 1990; Friestad and Wright 1994; Yoho 1998). While nearly every consumer can articulate what hard-sell tactics are and how they feel when they encounter them, to date there is no accepted measure of this concept (from the consumer’s perspective) in the marketing literature. Therefore, we must have a valid and reliable measure of our construct before we can learn more about consumer perceptions of sales pressure and its impact. The following briefly outlines the process undertaken to create this measure. The scale development procedure used for this study is based on the Churchill (1979) approach, with the inclusion of several updates and improvements as recommended by experts in scale development (i.e., Anderson and Gerbing 1988; DeVellis 1991; Gerbing and Anderson 1988; Hardesty and Bearden 2004; Netemeyer et al. 2003; and Spector 1992). The purpose of updating the Churchill (1979)


procedure was to ensure that the scale development procedure employed for this study reflects the current state of the art in the development of marketing measures. While maintaining the framework of the widely accepted Churchill (1979) approach to developing marketing measures, the additional analyses recommended by modern scale development experts resulted in a more thorough vetting of the items and thus a measure with excellent psychometric properties. To our knowledge, this is the first paper to update the scale development procedures outlined by Churchill to include recommendations by subsequent scale development researchers. The updated procedure can be examined in Fig. 2. Additional steps not included in Churchill’s (1979) procedure are annotated accordingly. Following the qualitative exploration (completed to provide insight into the construct definition), Step 2 requires determining response categories and respondent instructions. Based on accepted practices and standards in our field, we chose to measure the construct via a seven point Likert-type response format ranging from “strongly disagree” to “strongly agree” with the following instructions accompanying the scale: Please think about your most recent shopping episode for a specific product that required the help of a salesperson and at least a moderate level of thought and effort

Step 1 CONSTRUCT DEFINITION Churchill 79 (Step 1) Step 2 SCALE DESIGN DeVellis 91, Spector 92 Step 3 ITEM GENERATION Churchill 79 (Step 2) Step 4 EXPERT ITEM JUDGING DeVellis 91, Netemeyer et al. 03 Step 5 PILOT TEST Spector 92, Netemeyer et al. 03 Step 6 DEVELOPMENT SAMPLE Churchill 79 (Steps 3 & 4)

Step 7 INITIAL VALIDATION Churchill 79 (Steps 5 through 8)

Step 8 FINAL VALIDATION Gerbing & Anderson 88

• Literature Review • Theoretical Domain • Theoretical Definition • Theoretical Dimensionality • Response format type • Number of response categories • Scale Instructions • Item Sources - Scale Developer - Experts - Consumers • Item Content • Item Wording • Face validity • Small Number of Respondents • Debrief Respondents • Revise Based on Feedback • Social desirability bias check • Item statistics • Exploratory factor analysis • Coefficient alpha (preliminary) • Confirmatory factor analysis - Dimensionality - Consistency - Convergent validity - Discriminant validity • Coefficient alpha - Reliability • Norms (initial) • Confirmatory factor analysis - Measurement invariance • Coefficient alpha - Reliability • Norms (initial) • Structural Equation Modeling - Nomological Validity

Fig. 2 Updated scale development procedure

on your part. Based on your memory of this buying situation, please circle the number indicating your level of agreement with each of the following statements (with 1 meaning “strongly disagree,” 7 meaning “strongly agree”). A large pool of potential scale items was generated. Using both the marketing literature and personal experience as inspiration, potential items were generated by the authors (104 items), an academic expert in the field (11 items), an expert practitioner (5 items), and a sample of undergraduate student consumers (38 items). The external participants in the process were provided with the conceptual definition of the construct. Four items were added during the expert judging phase bringing the total number of items to 162. The resultant pool of 162 items was compiled and presented to eight individuals deemed to be experts in the theoretical domain (e.g., consumer psychology, and/or scale development). Each item was judged to be either “clearly representative,” “somewhat representative,” or “not representative”. Following the recommendations of Hardesty and Bearden (2004), items were retained if at least 60% of the judges rated them “clearly representative”. Also, if any one judge rated an item “not representative,” then the item would only be retained if 80% of the remaining judges rated it to be “clearly representative”. Following this process, which exhibited an inter-rate reliability of .808, 31 of the original 162 items were retained for further scale refinement. The 31 items retained in the expert judging phase of the study were then administered, along with demographics, to a group of 34 business students for pilot testing. The participants were debriefed upon completion of the questionnaire. Other than the common concerns with redundancy among questionnaire items and scale length (to be expected at this stage in the development process), there were no substantive issues uncovered during this test. Interestingly, although the perceived pressure reported was moderate (mean=3.09), 82% of respondents (28/34) reported that they purchased the product for which they were shopping. The process continued with the first full-scale data collection. The primary objective of the development sample is to examine the factor structure of the perceptions of sales pressure scale items, and to begin the scale purification process by removing items, based on guidelines described herein. The study respondents were recruited by undergraduate students at two universities (in the Midwest and Southeastern regions of the United States), following detailed instructions from the researchers. Responses were exposed to the following quality checks, as suggested by Dollinger and DiLalla (2005): “I have tried to answer all of these questions honestly and accurately” and “If you read this item, do not respond to it”. The demographics of the sample, as well as other results from this phase of the study (including EFAs and item statistics), can be found in Table 1.

J. of the Acad. Mark. Sci. Table 1

Steps 6 & 7 of scale development

Respondents: 434

DEVELOPMENT SAMPLE Respondents Removed Due to Failed Quality Checks: 73 (16.8%)

Age (mean): 35 yrs | Range: 18-65 yrs

Age and Gender Male: 46%

N= 360

Female: 54%

Education 4-year Degree: 24% Some Graduate Work: 5% Graduate Degree: 12% Ethnicity Caucasian: 83% Black: 10% Asian: 4% Native American: Hispanic: 1% Other: 1% 1% Exploratory Factor Analysis Extraction: Principle Axis Factoring Rotation: Direct Oblimin Criteria for Deleting Items (used on all EFA's performed in subsequent studies as well) 1 - Failure to load significantly on a single factor (.40) 2 - Cross-loading on multiple factors. 3 - Communalities of at least .50 Initial No. Items: 31 No. of Iterations No. of Factors: 1 Remaining Items: 29 (EFA's): 3 Item Statistics Range of Item- to-total Correlations: .73-.90* Range of Interitem Correlations: .50 - .85* *Suggested thresholds for item-to-total and interitem correlations is .50 and .30 respectively (Hair et al. 2009, Robinson Shaver and Wrightsman 1991) Range of Item Means: 2.44-3.49 INITIAL VALIDATION Respondents: 370 Respondents Removed Due to Failed Quality Checks: 66 (17.8%) N= 304 Age and Gender Age (mean): 48 yrs | Range: 20Male: 37% Female: 63% 74 yrs Education Some College: 31% 4-year Degree: 37% Some Graduate Work: 5% Graduate Degree: 19% Ethnicity Caucasian: 87% Black: 5% Asian: 4% Hispanic: 2% Other: 2% Exploratory Factor Analysis Initial No. Items: 29 No. of Iterations No. of Factors: 2 Remaining Items: 27 (EFA's): 3 Confirmatory Factor Analysis Parameter Estimates: Maximum Likelihood Software: LISREL 8.80 (Jöreskog and Sörbom 2001) Initial No. Items: 27 No. of Factors Specified: Factor 1: 18 items Factor 2: 9 items 2 NOTE: Items removed iteratively (i.e. 1 at a time) based on completely standardized loadings, significance, standardized residuals, modification indices, and error variances. Resulting No. Items: 8 No. of Factors Specified: Factor 1: 5 items Factor 2: 3 items 2 CFI: .99 TLI: .99 IFI: .99 SRMR: RMSEA: Model Fit: Χ2 (df = 19, n = 304) = 42.60, p = 0.0015 .025 .064 Discriminant Validity Confirmed (AVE > Factor 1 Factor 2 Shared Variance (Squared Correlation): SV) AVE: AVE: .59 (via Fornell & Larcker, 1981) .64 .67 Convergent and discriminant validity was also confirmed using the method suggested by Anderson and Gerbing (1988). The coefficient of each item was significant and greater than twice its standard error (convergent validity) and CFAs run to compare an unconstrained model to one in which the factors’ correlation was constrained to 1 showed the chi-square value of the constrained model to be significantly higher (78.06 vs. 29.41, ∆ df = 1) suggesting that discriminant validity had been achieved. Some College: 50%

Item Statistics Range of Item- to-total Correlations: .69-.81* Range of Interitem Correlations: .48 - .73* Range of Item Means: 2.20-2.65 Scale Norms Factor One Summated Range: 5.0-35.0 Mean: 12.68 SD: 6.86 Scale Factor Two Summated Range: 3.0-21.0 Mean: 7.87 SD: 4.28 Scale


The primary objectives of the second, initial validation study were to examine dimensionality, reliability, and both convergent and discriminant validity. To achieve these objectives, a second sample of data was collected in a similar manner as the first, at a different university in the Midwestern United States. The demographics of the sample and other results are summarized in Table 1. The data analysis executed during this phase consisted of the following: EFA (per Netemeyer et al. 2003), CFA (Gerbing and Hamilton 1996), convergent and discriminant validity tests on scale factors (Fornell and Larcker 1981; Anderson and Gerbing 1988), an assessment of face validity conducted by the research team, item statistics, and scale norms. Face validity would compel a more specific discussion of the two factors that have been confirmed by the scale development process. The first factor, consisting of four items and coinciding with Kirmani and Campbell’s (2004) target role of “persuasion sentry,” could best be described as “aggressive sales pressure”. Its items specifically mention “high-pressure” tactics, being rushed into a decision, and feeling “played” by the salesperson. The salesperson is controlling the transaction, manipulating the customer, and acting aggressively and/or pushy. Factor two, consisting of three items and coinciding with Kirmani and Campbell’s (2004) target role of “goal seeker,” taps into the feeling on the part of the customer that the salesperson is pushing the sale of a product other than what the customer intended to buy. Since it would appear that the perception was that the salesperson was directing the sale, we have termed this factor “directive sales pressure”. Research in the marketing channels area has examined the use of influence strategies by boundary spanning personnel on the interfirm relationship for many years. However, the concepts studied were all reported at the firm level instead of the individual level and most often with respect Table 2

to longer term relationships as opposed to a single interaction. Thus, they are helpful in the retail context but not the same. That said; the second factor in this scale is similar in nature to the noncoercive influence strategy of “recommendations” used in previous influence research in marketing channels (Frazier and Summers 1984, 1986; Frazier and Rody 1991). Consistent with this comparison, the recommendation strategy is directive in nature (Angelmar and Stern 1978) and “is more closely related empirically to the coercive strategies” (Frazier and Rody 1991, p. 58). Hence, while not to the same degree as the aggressive factor, the directive factor is still a source of perceived pressure to the consumer. Given that both factors are indications of felt pressure on the part of the consumer, this assertion applies here as well. Furthermore, Frazier and Summers (1984) indicate “recommendations” as the preferred influence strategy in channels characterized by low interdependence and infrequent contact (indicative, in this case, of the retail channel). The final scale items and their loadings can be found in Table 2. To examine convergent and discriminant validity evidence, outside of the factors within the measure, an additional closely related variable, a three-item measure of “source pressuring” (Swasy and Munch 1985), was included for this phase. The end points for the three-item, seven-point, semantic differential scale were: low/high pressure, not pushy/pushy, and unaggressive/aggressive. The source pressuring construct is strongly positively correlated to both the aggressive (r= 0.74) and directive (r=0.59) factors. These strong correlations call for an examination of discriminant validity. Again, using Fornell and Larcker’s (1981) method, the shared variance between each factor of the perceptions of sales pressure and the source pressuring scale is 0.54 and .35, each less than the average variance extracted for the newly developed scale

The consumer perceptions of sales pressure scale (final version)

Completely standardized Instructions - Please think about your most recent shopping episode for a specific product that required the help of a salesperson and at least a moderate level of thought and effort on your part. Based on your memory of this buying situation, loading please indicate your level of agreement with each of the following statements. Factor one- aggressive sales pressure 1. The salesperson used high pressure sales tactics. 2. The salesperson forced me to make a decision before I was ready. 3. The salesperson attempted to close the sale before all of my concerns were addressed. 4. I felt like I had been “played” by the salesperson. Factor two- directive sales pressure 5. The salesperson kept pushing me toward one product when I was interested in another. 6. I felt pressure to buy what the salesperson insisted on, as opposed to what I originally wanted. 7. The salesperson was more interested in selling me his or her recommendation than the product I wanted.

0.83 0.75 0.74 0.77 0.83 0.89 0.89

• Items are measured using a seven-point Likert-type response format with anchors ranging from 1=Strongly Disagree to 7=Strongly Agree • All items are worded in the positive direction. Higher scores on the scale are indicative of higher levels of perceived sales pressure


factors (0.64 and 0.67). These results indicate sufficient discriminant validity. Sample and measures The full structural model test, in addition to testing the hypothesized study model, serves the secondary purpose of confirming dimensionality, reliability, and validity of the POSP scale. Evidence of nomological validity is found when a newly developed construct, empirically tested as part of a theoretical model, exhibits distinct relationships with antecedents and consequences (Netemeyer et al. 2003). We recruited a quota sample of 347 consumers to complete our online survey instrument. After the removal of cases for nonresponse and quality check failure, the final sample size was n=275. The demographic breakdown of the sample included 52% females, ranged from 22 to 70 years of age, with an overall mean age of 43. With respect to ethnicity, 91% were Caucasian, 3% were African American, with 2% each being reported as Hispanic, Native American, and other. While 89% of the sample reported at least some college, 25% held a four-year college degree, 5% reported some graduate school, and 26% held a graduate degree. The measures used in the study have all been shown to be reliable and valid in the literature.

Results We adopted Anderson and Gerbing’s (1988) two-step approach, whereby a measurement model (CFA) was estimated, followed by the simultaneous estimation of the measurement model and the full structural model, testing the hypothesized relationships among latent variables. The results of the measurement model were favorable in that all items loaded significantly on their intended construct (25 total items over 7 latent constructs). Furthermore, the resultant fit statistics suggested that the measurement model provided an adequate fit to the data: Χ2 (df=254, n=275)=428.004; CFI, TLI, and IFI all at 0.99, and RMSEA at 0.0499. To further confirm the appropriateness of the two-factor solution, these results compared significantly favorably to the same for a CFA run with a unidimensional POSP measure: Χ2 (df=260, n=275)=524.764; CFI, TLI, and IFI all at 0.98, and RMSEA at 0.0609. To assess validity of scales used in the final validation, we followed the recommendations of Fornell and Larcker (1981). Convergent validity was confirmed through the calculation of average variance extracted (AVE) by each construct, relative to error. All of the study constructs met the requirements of this test, with AVEs over 50% (range=60–88%). The criteria for discriminant validity were also met for all constructs. These results, along with correlations, descriptive statistics, and construct reliabilities can be found in Table 3. One item from the aggressive factor of the POSP scale was removed for

its excessive error variance relative to its loading (1.48/1.17), its relatively low R-square value (0.48), and content validity concerns. Additionally, convergent and discriminant validity was confirmed for the POSP scale using the method suggested by Anderson and Gerbing (1988). The coefficient of each item was significant and greater than twice its standard error (convergent validity), and CFAs run to compare an unconstrained model to one in which the factors’ correlation was constrained to 1 showed the chi-square value of the constrained model to be significantly higher (100.31 vs. 16.70, Δ df=1), suggesting that discriminant validity had been achieved. The hypothesized structural model was then tested for path estimates and overall model fit. The tested model reflected good overall fit: Χ2 (df=258, n=275)=675.49; CFI, TLI, and IFI all at 0.97, and RMSEA at 0.077. The complete results of the structural model can be found in Table 4. In total, 10 of 14 hypotheses were supported by the data. All hypotheses related to antecedents to the aggressive factor of perceived sales pressure found support. Specifically, the consumer’s perception of salesperson knowledge negatively impacted sales pressure perceived (H1a), as did the consumer’s perceived self-confidence in their persuasion knowledge (H3a). The consumer’s perceived degree of time pressure was found to positively impact perceptions of both types of sales pressure, supporting H2a and H2b. Meanwhile, H1b and H3b were not supported; that is, the impacts of perceived salesperson knowledge (H1b) and self-confidence in persuasion knowledge (H3b) were not significantly related to the consumer’s perception of directive sales pressure. While the relationships between self-confidence in persuasion knowledge with satisfaction and trust were found to be fully mediated by both factors of perceived sales pressure (Baron and Kenny 1986), direct paths were hypothesized from antecedents (salesperson knowledge and time pressure) to the outcomes of satisfaction and trust. Support was found for all of these hypotheses. Specifically, salesperson knowledge was positively related to both satisfaction (H4a) and trust (H5a), and time pressure was negatively related to the same (H4b, H5b). The results concerning the impact of perceived sales pressure on outcomes of note are mixed. The negative relationships hypothesized between perceived aggressive sales pressure and satisfaction with the salesperson (H4c) and trust in the salesperson (H5c) were both supported strongly (with path coefficients of 0.52 and 0.5, respectively). This finding certainly underscores the potential negative impact that perceived sales pressure can have, given the myriad of potentially negative consequences that can result from diminished customer satisfaction and trust. On the other hand, the relationships between perceived directive sales pressure and both satisfaction with (H4d) and trust in (H5d) the salesperson were not found

J. of the Acad. Mark. Sci. Table 3

Means, standard deviations, correlations, reliabilities, and convergent/discriminant validity assessment of the measurement model

Variable

M

SD

AVE

1

2

3

4

5

6

7

1. POSP (A)

2.47

1.30

0.60

0.86

0.55

0.05

0.15

0.04

0.36

0.28

2. POSP (D) 3. SKNOW

2.51 5.44

1.45 1.51

0.75 0.88

−0.74** −0.22**

0.90 −0.19**

0.04 0.98

0.17 0.03

0.01 0.00

0.29 0.26

0.21 0.24

4. TIME 5. SCPK

3.44 5.93

1.63 0.66

0.66 0.63

−0.39** −0.19**

−0.41** −0.12**

−0.17** −0.06**

0.85 −0.13**

0.02 0.83

0.15 0.02

0.15 0.02

6. SAT 7. TRUST

5.09 4.89

1.41 1.26

0.77 0.75

−0.60** −0.53**

−0.54** −0.46**

−0.51** −0.49**

−0.39** −0.39**

−0.13** −0.13**

0.94 −0.79**

0.62 0.90

• Listwise n=275, AVE = Average variance extracted • Construct reliabilities are provided on the diagonal in bold • Shared variances are reported above the diagonal • Pearson correlations are below the diagonal ** Correlation is significant at the 0.01 level (2-tailed) * Correlation is significant at the 0.05 level (2-tailed) Key: POSP (A, D) Perceptions of Sales Pressure (Aggressive, Directive), SKNOW Salesperson Product Knowledge, TIME Time Pressure, SCPK SelfConfidence, Persuasion Knowledge, SAT Satisfaction with Salesperson, TRUST Trust in Salesperson

to be significant. These findings suggest that aggressive sales pressure may reduce trust and satisfaction; however, directive sa le s pressure may be perce iv ed as a salespersons attempt to assist the consumer in finding a product that better addresses the consumer’s needs. While not necessarily a driver of positive assessments of satisfaction and trust, the directive factor of perceived sales pressure can certainly be described as considerably less damaging to desirable marketer outcomes, when compared to aggressive sales pressure.

Table 4 Structural model results maximum likelihood estimates

Discussion Within a theoretical context of PKM, we tested a structural model of potential antecedents and outcomes of consumer perceptions of sales pressure. The results of this research bring to light the existence of two factors of consumer-perceived sales pressure that behave differently in terms of their antecedents and outcomes. In the process, we conceptualized, developed, and validated a scale for the measurement of consumer perceptions of sales pressure (POSP). The resultant

Hypothesized path

Path coefficient

R2

Support

1a. Salesperson Knowledge → POSP (A) 1b. Salesperson Knowledge → POSP (D) 2a. Time Pressure → POSP (A) 2b. Time Pressure → POSP (D) 3a. Self-Confidence (PK) → POSP (A)

−0.14* −0.11* −0.47* −0.49* −0.14*

0.31 0.29

Supported Not Supported Supported Supported Supported

3b. Self-Confidence (PK) → POSP (D) 4a. Salesperson Knowledge → Satisfaction 4b. Time Pressure → Satisfaction 4c. POSP (A) → Satisfaction 4d. POSP (D) → Satisfaction 5a. Salesperson Knowledge → Trust 5b. Time Pressure → Trust 5c. POSP (A) → Trust 5d. POSP (D) → Trust Structural model results: Χ2 (df=258, n=275)=675.49, p