Mark Lett DOI 10.1007/s11002-015-9392-8
A longitudinal examination of the impact of quality perception gap on brand performance in the US Automotive Industry M. Billur Akdeniz 1 & Roger J. Calantone 2
# Springer Science+Business Media New York 2015
Abstract A quality perception gap, defined as the difference between perceived and objective quality, indicates either consumers’ overappreciation or underappreciation of product or brand quality and can have critical effects on performance. The purpose of this research is to examine the impact of a quality perception gap on brand performance and its moderating role in the relationship between marketing-mix signals and performance. Analyses based on a longitudinal dataset from the US automotive industry reveal that the relationship between the quality perception gap and brand performance has an inverted U-shape. Findings also demonstrate that, except for advertising, the impact of marketing signals on performance is higher when the quality of a brand is perceived as higher than its actual quality. Finally, over an 18-year period, the average gap between perceived and objective quality demonstrates a decreasing trend, indicating that the nature of demand in the automotive industry has become more utilitarian. Keywords Quality perception gap . Objective quality . Perceived quality . Marketing signals . Panel data econometrics . Automotive industry
1 Introduction Quality is a key element of marketing strategy and performance. The extant marketing literature lists various strategic benefits of increased quality for performance such as its
* M. Billur Akdeniz
[email protected] Roger J. Calantone
[email protected] 1
Department of Marketing, Peter T. Paul College of Business and Economics, University of New Hampshire, Durham, NH 03824, USA
2
Department of Marketing, Eli Broad College of Business, Michigan State University, East Lansing, MI 48824, USA
Mark Lett
positive impact on customers’ brand attitudes, willingness to pay, sales, profitability, stock market returns, and return on investment (e.g., Aaker and Jacobson 1994; Tellis and Johnson 2007; Zhao 2000). Being a complex and multidimensional concept, quality has been examined from various perspectives. Drawing on information economics, two of the more frequently examined dimensions of quality in marketing are objective and perceived quality. Objective quality refers to the features of a product whether or not anybody realizes their presence whereas perceived quality refers to a consumer’s subjective judgment of those features (Mitra and Golder 2006). Perceived quality usually differs from, yet is influenced by, the actual 1 quality of a product (Golder et al. 2012; Morgan and Vorhies 2001). Relatively scarce research, with the exception of the service quality literature (Zeithaml 1988), has examined the concept of a discrepancy between the two quality measures. In the manufactured goods or brand management areas, a limited number of studies has examined the impact of a gap between perceived and objective quality of a product or brand on its market performance (Kopalle and Lehmann 2006; Steenkamp et al. 2010). However, a gap indicates either an overappreciation or underappreciation of product quality by its consumers and can have important long-lasting effects on performance. To address this void in the literature, first, this study conceptualizes the quality perception gap as the difference between perceived and objective quality and empirically tests whether the quality perception gap of a brand affects its market performance. Second, it investigates the role of market signaling and whether the relationship between various signals (i.e., price, advertising, warranty, and distribution network) and brand performance is influenced by consumers’ having a higher or lower perceived quality than the actual quality. We choose the US automotive industry as the context of our study since quality is a highly relevant attribute to the industry and there is a large variance in terms of perceived and objective quality of car brands. We compiled data from a variety of wellestablished secondary sources, and the final dataset comprises of 354 annual brand-level observations of 33 automotive brands between 1990 and 2007. Key findings reveal that the relationship between the quality perception gap and brand sales has a non-monotonic relationship implying that only up to a certain level will the gap favoring a higher perceived than objective quality have a positive impact on brand performance. Furthermore, except advertising, an increase in marketingmix signals has a positive impact on brand performance when the quality perception gap increases. Finally, we discover that the quality perception gap has a decreasing trend over time. This study contributes to the marketing literature in specific ways. First, the study is one of the very few studies that empirically examine the impact of a quality perception gap on brand performance in the context of manufactured goods using longitudinal real-industry data. It is common for manufacturers or sellers to intentionally overstate or understate the quality of their products as part of their marketing strategy. For instance, when introducing the Lexus brand, Toyota understated the quality level to frame customers to expect less and to pleasantly surprise them when their expectations
1
BActual^ and Bobjective^ are used interchangeably in the manuscript.
Mark Lett
were exceeded with their experience of the brand (Mannering and Winston 1991). Second, this research contributes to the previous literature in the area of marketing and quality strategy alignment by addressing the potential interaction between marketingmix signals and a brand’s quality perception gap in affecting its market performance.
2 Conceptual background 2.1 Definitions of perceived versus objective quality and the quality perception gap A review of the literature suggests that the quality concept is highly elusive, complex, and multidimensional; and, various disciplines have their own approach to studying quality. For instance, in engineering, quality means conformance to design standards and manufacturing processes, whereas in operations management, quality is defined with reference to organizational processes such as total quality management (e.g., Juran 1974). In the marketing discipline, quality has been defined in various ways and examined as a key metric of firm strategy for more than 30 years. Garvin (1984) summarizes quality in five categories as employing transcendent, product-based, userbased, manufacturing-based, and value-based approaches. Parasuraman et al. (1985) examine the differences among perceived quality, actual quality, and consumer expectations. Golder et al. (2012) present the three quality processes of production, evaluation, and experience in one integrative framework. This research hones in on quality production and evaluation processes with particular emphasis on the two dimensions of quality: objective and perceived. We follow Mitra and Golder (2006) and define objective quality as the actual superiority or excellence of a product and perceived quality as consumers’ subjective evaluation with respect to a product’s overall superiority or excellence. Objective quality consists of the features that are present in a product whether or not anyone realizes or acknowledges them (e.g., the engine or transmission of a car), and these features can be objectively measured most of the time. On the other hand, perceived quality consists of the idiosyncratic judgment of those features from the customer’s point of view, and trial or experience is not always necessary to form a perception about product quality. Zeithaml (1988) is among the first to propose that perceived quality, not as a specific attribute but rather a higher level abstraction, usually differs from objective quality. A significant body of research has examined objective and perceived quality as distinct entities and suggested that a consumer’s quality perception typically differs from the actual quality of a product or service (e.g., Kopalle and Lehmann 1995; Morgan and Vorhies 2001). Researchers have also demonstrated that objective quality has a significant impact on consumer perceptions and the variability in perceived quality is lower than that in objective quality (Boulding et al. 1993; Mitra and Golder 2006). Building on the literature examining perceived versus objective quality, we propose a new construct named Bquality perception gap^ to emphasize the discrepancy between perceived and objective quality. The quality perception gap can be either positive or negative depending on whether consumers have higher or lower perceptions than the actual level of quality. With the exception of previous research in the service quality
Mark Lett
literature, the marketing literature has scarcely investigated the concept of a gap between two different conceptualizations of quality for manufactured goods and brands. 2.2 Signaling theory and brand quality management In the marketing strategy literature, signaling theory has been influential in the study of firm-to-consumer relationships. Drawing on information economics, the main premise of signaling theory is that sellers and buyers usually have asymmetric levels of information about complex or experiential products when the quality is not readily observable (Akerlof 1970; Spence 1973). Since firms possess more information about the quality of such products, consumers often have imperfect information, uncertainty in their evaluation, and high levels of risk in their purchase behavior (Rao et al. 1999). In this situation, a firm uses observable signals to deliver information about the characteristics of products and services to alleviate information asymmetry, and customers try to assess the actual quality of a product using these signals. The main purpose of signaling theory is to determine whether a signal conveys credible information and reduces the information asymmetry between firms and customers (Kirmani and Rao 2000). Credible signals help consumers distinguish a high-quality seller from a lowquality seller since the former has an incentive (e.g., higher future revenues due to signaling) and the latter has a disincentive (e.g., lower future revenues compared to not signaling) to signal the market. In contrast, if market forces are not strong enough to encourage firms to choose different strategies, a pooling equilibrium occurs where customers cannot rely on signals to differentiate between high- and low-quality sellers (Kirmani and Rao 2000; Basuroy et al. 2006). Brands are often considered credible signals of unobservable quality (Erdem and Swait 1998). Being one of the most valuable intangible assets for a firm, brands create a Bbonding mechanism^ between the firm and its customers (Boulding and Kirmani 1993). Signaling theory posits that a rational consumer expects a branded seller to stick to its claims and the implicit commitment behind the signals; a brand’s false claims or failure to keep its promise results in a loss of profits, a damaged reputation, and other unattractive outcomes (Aiken and Boush 2006; Boulding and Kirmani 1993). Also, a seller with a high brand reputation has less incentive to overstate quality or make a false claim because its credibility is based on the brand name embodying the cumulative effect of all past marketing strategies and actions (Wernerfelt 1988). When quality is difficult to assess, a branded seller can use its reputation as a credible marketing signal to emphasize the consistent quality of its products and alleviate consumer’s uncertainty in decision making. 2.3 Relationship between quality perception gap and brand performance The impact of perceived or objective quality on consumer attitudes and performance has been examined in previous research (e.g., Aaker and Jacobson 1994; Bolton and Drew 1991; Rust et al. 1999), but the impact of a discrepancy between perceived and objective quality of a brand on its market performance has scarcely been addressed
Mark Lett
(e.g., Steenkamp et al. 2010). Nonetheless, this impact might be critical since the presence of a gap indicates either consumers’ overappreciation or underappreciation of brand quality and can have long-lasting effects on brand performance. Some earlier findings have suggested that deviations in customer perceptions from actual quality can indicate a firm’s lack of understanding of perceived quality drivers or its overrepresentation of actual quality through exaggerated claims. In that case, while performance increases initially as a result of a quality differential in favor of customer perceptions, it can deteriorate in later periods (e.g., Kopalle and Assuncao 2000; Kopalle and Lehmann 2006). Furthermore, in the case of a misalignment between a firm’s and a customer’s understanding of quality, the performance can be negatively affected because of ineffective translation of quality management practices with respect to customer perceptions. Prior research has also demonstrated that firms are less likely to overstate quality through inaccurate claims when the actual quality is high. Furthermore, when perceived and actual quality are better aligned, firms have better success rates in the long-term and are likely to allocate resources between quality improvement and marketing efforts more effectively (e.g., Morgan and Vorhies 2001). Thus, first, we investigate whether a significant relationship exists between the quality perception gap of a brand and its market performance. Specifically, we argue that when there is a discrepancy between perceived and objective quality of a brand, it is likely to have a significant effect on brand performance because either consumers have higher opinions of a brand than its true quality levels justify or they underestimate the actual quality. As consumer perceptions are not likely to change over time, the relative position of perceived quality with respect to objective quality of a brand is expected to have a long-term impact on consumers’ purchase decisions. Second, we investigate how the impact of various marketing-mix signals on brand performance is influenced by the quality perception gap. We specifically focus on four marketing signals: price, advertising expenditure, warranty, and distribution network. An increase in price usually hampers performance (e.g., Tellis 1988) whereas the impact of an increase in advertising expenditures is expected to be positive on a brand’s performance (e.g., Assmus et al. 1984; Vakratsas and Ambler 1999). Furthermore, an increase in warranty offering or expansion in the distribution network contributes positively to the value proposition of a brand and, thus, is expected to affect its performance positively (Padmanabhan and Rao 1993; Purohit 1997). Additionally, prior research in marketing has suggested that a firm’s marketing and quality strategy must be aligned for the firm’s quality improvement efforts to be righteously communicated to its customers (Kordupleski et al. 1993). Thus, by investigating the interaction effects between a quality perception gap and marketing-mix signals, we aim to understand how the impact of signals on brand performance is influenced by consumers’ overappreciation or underappreciation of brand quality.
3 Methodology 3.1 Data source and measurement The dataset comprised annual observations of car brands in the US automotive industry between 1990 and 2007. The automotive industry is highly relevant for the purposes of
Mark Lett
this study because quality is an indispensable attribute of cars yet a large discrepancy exists in the objective and perceived quality among different brands. We compiled the dataset depicted in Table 1 from multiple secondary sources, which are well known and frequently used in marketing research. The dependent variable is brand performance measured as the annual unit sales of a car brand (sales). We obtained annual unit sales for each brand from the Automotive News Market Data Book. We identified four marketing-mix signals as independent variables in our model. We collected the manufacturer’s suggested retail price (price) from Consumer Reports and National Automobile Dealers Association’s yearly guides. Advertising expenditure (advertising) was collected from TNS Media Intelligence through Ad$pender and Advertising Age. For warranty (warranty), the duration of the basic warranty in terms of months and mileage, and a warranty index were obtained from the Car Book by Jack Gillis (Douglas et al. 1993). In the analyses, the warranty index was used as a more comprehensive variable to provide more variance across brands than basic month and mileage data. The index is a weighted average of the basic, power train, and corrosion warranties where higher numbers indicate better warranties. To check the appropriateness of the index, a factor analysis was conducted to understand which measure was a better proxy for the warranty variable. Results showed that only one component score with an eigenvalue greater than 1.0 explained 84 % of the variation and the warranty index had the highest loading to this component. Finally, to assess the distribution network, data for the number of dealerships (dealer) of an automotive brand in the US market were collected from the Automotive News Data Center.
Table 1 Measures and sources of the variables Variable
Notation
Measure
Source
Sales
Sales
Number of units sold
Automotive News Market Data Book
Quality perception gap
Gap
Difference between Bstandardized perceived Calculated quality (PQ)^ and Bstandardized objective quality (OQ)^ measures (PQ-OQ) • PQ from Equitrend by Harris Interactive: scale of 0–10 (0, poor quality, and 10, excellent quality) • OQ from Consumer Reports: scale of 1–5 (1, far below average, and 5, far above average)
Retail price
Price
Average MSRP of all models under each brand
Advertising expenditure
Advertising Dollars (in millions) spent on media
TNS Media Intelligence and AdAg
Warranty
Warranty
Composite index of basic, powertrain, corrosion warranties (larger number=better warranty)
The Car Book by Jack Gillis
Distribution network
Dealer
Number of dealerships of a car brand
Automotive News Data Center
Consumer Reports and NADA
Mark Lett
We calculated the quality perception gap (gap) by subtracting scores for objective quality from those for perceived quality. Perceived quality scores, measured on a scale of zero (0) to ten (10), zero meaning poor quality and ten meaning excellent, were collected from Harris Interactive’s Equitrend dataset. Equitrend is based on online surveys conducted since 1989 with 20,000 to 45,000 consumers aged 15 years or older to determine their perceptions of more than 1000 brands across 35 product categories. Providing longitudinal information on various brand dimensions, this dataset has been widely used in marketing research (e.g., Olsen et al. 2014; Bharadwaj et al. 2011). We gathered objective quality ratings from Consumer Reports (CR) quality ratings, which are widely used in quality research and measured on a scale of one to five: one meaning far below average and five meaning far above average (e.g., Bronnenberg and Wathieu 1996; Rhee and Haunschild 2006). CR provides one of the most trusted quality ratings in the marketplace because (i) the organization does not have any relationship with businesses and strongly discourages the use of its ratings in company advertisements and (ii) its ratings are the outcome of a series of comprehensive laboratory tests conducted by product experts on products they purchase (Mitra and Golder 2006). To calculate the gap, we adjusted perceived and objective quality measurement scales and standardized each variable. The final dataset consisted of 354 annual observations for 33 brands over 18 years. 3.2 Model estimation To answer the research questions, Eq. 1 examines the effects of marketing-mix signals, the quality perception gap, and the interaction between signals and gap on the unit sales of a car brand: Salesitþn ¼ β 0 þ β 1 Priceit þ β 2 Advertising it þ β 3 Warrantyit þ β 4 Dealerit þ β 5 Gapit þ β 6 Gap2it þ β 7 ðGap*PriceÞit þ β 8 ðGap*Advertising Þit ð1Þ þ β 9 ðGap*WarrantyÞit þ β 10 ðGap*DealerÞit þ γdY t þ θdBi þ uit
where Bi^ denotes a brand, Bt^ is the time period, Yt represents a set of year dummies, and βi represents a set of brand dummies. Sales, price, advertising, warranty, and dealer enter the equation in natural log forms. While estimating the model, we needed to address several issues as (1) panel data structure, (2) normality, autocorrelation, and heteroskedasticity, (3) endogeneity, and (4) multicollinearity. Panel data structure The structure of our dataset presents brands across years in an unbalanced form. To select between a fixed- or random-effects model estimation, we used a Hausman test (χ2 =31.08, ρ
