Product quality : an investigation into the concept and how it is perceived by consumers

Steenkamp, J.B.E.M.

The objectives of the present work are (1) to review the literature on product quality from different perspectives, (2) to develop a model that describes the way consumers form judgments about product quality, and (3) to investigate the proposed model empirically. In addition, price-perceived quality tradeoffs and the relationship between price and product quality in the marketplace are investigated. The main conclusions are summarized in this chapter.The concept of product qualityThe concept of product quality has received attention in various disciplines. Four major approaches to product quality are identified: the metaphysical approach, the production management approach, the economic approach, and the perceived quality approach. To enhance our insight into the product quality concept, each of these approaches is discussed, the most attention being given to the economic and the perceived quality approaches.The metaphysical approach concentrates on the being of quality. Quality is regarded by many authors as being synonymous with innate excellence. In essence, it is an unanalyzable property that an individual can learn to recognize only through experience. People differ about quality not because quality is different but because people are different in terms of experience.The production management approach differs from the metaphysical approach in that it regards quality as a concept that is objectively measurable. In the production management approach, quality is described in technical specifications. Quality is conformity with these technical specifications. This approach concentrates on producing a product of a given predetermined quality level. This level of quality is achieved by quality of design, quality of production, continuity of service, and customer service after sale. The most important recent development in the production management approach is the emergence of the ‘zero defects philosophy’. In the traditional view, very few or no defects are not economical because such perfectionism would cause the costs of preventing defects to exceed the costs of having the defects. However, it has been shown that quality improvement need not lead to higher costs, and increasingly, companies, especially Japanese firms, adopt the zero defects philosophy.The economic approach to quality studies the role of product quality in the market. Important areas of research are quality as a competitive weapon of the firm, the hedonic approach, and the role of quality in consumer behavior, both when consumers are perfectly informed about the market and when they are imperfectly informed. It is remarkable that these areas of research have developed rather independently from each other. Notable exceptions are the widespread adoption of Lancaster’s conceptualization of the product as a bundle of characteristics, and the increasing attention in economic theory in general for the notion that consumers are imperfectly informed.The production management approach and the economic approach have in common that they both define quality in terms of objectively measurable product characteristics. The three approaches to quality hitherto mentioned share the notion that quality is an objective property inherent in the product. The perceived quality differs from the other approaches in that it regards quality neither as absolute nor as objective. Quality is a subjective concept, dependent on the perceptions, needs, and goals of the individual consumer. One of its tenets is that people are imperfectly informed about on quality. need cues to form quality perceptions. Since these cues are usually imperfect indicators of quality, quality risk will be experienced. A meta-analysis of studies dealing with the effects of quality cues on perceived quality shows that the cues price, brand name, store name, country of origin, and the physical product (i.e., the physical characteristics of the brand) influence quality perceptions. This does not imply, however, that all of these cues have a significant effect on perceived quality for all products and in all situations. Presently, not much is known about the mediating role of personal and situational characteristics on cue effects. It is important to note that few models of the quality perception process have been proposed in the literature.A development in consumer behavior research that is relevant to the perceived quality approach is the increasing use of descriptively realistic but more limited models. These models usually are empirically testable, which is not the case for the large integrated models, and also stand a better chance of being applied outside the academic setting.The confusion surrounding the concept of quality is largely due to differences in perspective taken by different authors. A consumer behavior researcher and a philosopher have something quite different in mind when they communicate about quality. It is therefore important that a researcher clearly states the view on quality taken by him/her.It cannot be said that any one approach to quality is superior to any other. All four approaches are valuable in their own right. The usefulness of a certain approach is dependent on the issues to be investigated.The four approaches to product quality have developed largely independent from each other. However, it is possible to develop relations between the approaches as was done in this work. Briefly, the results of a study employing the perceived quality approach can be used to develop the technical quality specifications employed in the production management approach. This integration of the perceived quality approach and the production management approach has been called ‘quality guidance’, the theoretical underpinnings of which are discussed more fully in Steenkamp and Van Trijp (1988a, 1989b). The economic approach can be used, in conjunction with the production management approach and the perceived quality approach to select the quality level that yields the highest profit or market share. The metaphysical approach draws attention to the role of esthetic product aspects in consumer evaluations. A model of the quality perception processIn the remainder of the study, we adopt the perceived quality approach (with the exception of Chapter 12). A model of the quality perception process is developed that integrates concepts developed in information processing, social and cognitive psychology, and economics. Die model is based on a definition of perceived quality developed within the broader context of value. The main elements of this definition are that perceived quality (1) involves preference, (2) involves a subject-object interaction, (3) is relativistic in that it is comparative, personal, and situational, and (4) resides in the consumption of the product.A distinction is made between quality cues and quality attributes. Quality cues can be ascertained by the senses prior to consumption. Quality attributes are benefit-generating product aspects and cannot be observed prior to consumption. It is posited that overall quality judgments are based on quality attribute perceptions. Quality cues are valued because they predict quality attributes. A consumer uses quality cues since direct information about the quality attributes is usually not available to consumers at the point of purchase.Quality cues are categorized as either intrinsic or extrinsic. Intrinsic cues are part of the physical product. Extrinsic cues are related to the product, but are physically not part of it. A distinction is further made between experience quality attributes and credence quality attributes. Experience attributes can be ascertained on the basis of the actual experience of the product, whereas credence attributes cannot be ascertained even after normal use for a long time. The most important process governing the use of quality cues in the formation of (experience and credence) quality attribute perceptions is the process of inferential belief formation. People infer quality attribute beliefs on the basis of cues that are acquired and categorized. It is hypothesized that the magnitude of the effect of a certain cue in inferential belief formation with respect to a certain attribute is (1) positively affected by the strength of the perceived relationship between the cue and the attribute in question, i.e., by the predictive value of the cue with respect to that attribute, (2) positively affected by the confidence an individual has in his/her ability to accurately perceive and categorize the cue, i.e., by the confidence value of the cue, and (3) is usually greater for an intrinsic cue than for an extrinsic cue. Cue confidence value and cue intrinsicness or extrinsicness are independent of the inferred attribute. Further, it is assumed that consumers are homogeneous with respect to their perceptions of a cue’s intrinsicness or extrinsicness.Personal and situational variables are hypothesized to influence the quality perception process. The personal variables prior experience, level of education, quality-consciousness, and perceived quality risk, and the situational variables usage goal for which the product is purchased, physical surroundings, social surroundings, and time pressure appear especially relevant. Qualityconsciousness is a new concept developed in this work. It is defined as ‘A mental predisposition to respond in a consistent way to quality-related aspects, which is organized through learning and influences behavior’.An empirical investigation into the proposed model of the quality perception processThe proposed model of the quality perception process is tested empirically for two meat products, saveloy and gammon. For each meat product, two usage goals were specified: for saveloy, use on sandwiches and use as snack, for gammon, use on sandwiches and use at dinner. Thus, the empirical investigation into the model of the quality perception process involves four different combinations of products and usage goals. An experimental design was developed to parameterize the relationships between quality cues, quality attributes, and overall perceived quality for each productlusage goal combination separately. Real samples of saveloy and gammon were used instead of verbal descriptions. The products were factorially composed from a set of intrinsic and extrinsic quality cues, using highly fractionated designs, and had been produced by a meat products firm.From the consumer panel of a large Dutch market research agency 600 subjects were sampled. In each of the four experimental conditions (i.e., saveloy for sandwiches or snack, gammon for sandwiches or at dinner) 120 subjects participated. The remaining 120 subjects participated in an information integration experiment concerning price-perceived quality tradeoffs (see below).The most important proposition of the model: ‘Quality attributes act as intervening variables mediating the effects of quality cues on perceived quality judgments’ receives strong support. For most quality cues, indirect effects exceed direct effects. In line with this result, as hypothesized, perceived quality judgments appear to be predominantly based on the quality attributes. Cues add little to the variance explained in perceived quality judgments when the effect of the quality attributes is controlled for. Cues are valued not for their own sake but because they predict quality attributes.For each product, the results are compared across usage goals to explore the effect of the usage goal for which the product is purchased on the quality perception process. Although some effect of the usage goal on the quality perception process is found, the differences are, in general, modest. It suggests that the quality perception process is not strongly dependent on the usage goal. A number of other hypotheses concerning specific aspects of the model of the quality perception process and the influence of certain consumer characteristics thereupon are tested.Experience attributes taken as a whole are more important in the formation of perceived quality judgments than credence attributes, but this need not be true for every pair of experience and credence attributes. Some credence attributes (e.g., attributes related to health issues) are so important to consumers that this overrules the rather large uncertainty inherent in this type of attributes.Consumers are found to be more able to use quality cues in inference processes with respect to experience attributes than in inference processes with respect to credence attributes. The reason is that credence attributes cannot be ascertained after consumption and that, therefore, the consumer has less opportunity to learn/modify beliefs about the inferential relationships between those attributes and quality cues.The relevance of the predictive value/confidence valuelintrinsic-extrinsic framework for explaining cue importance in the formation of perceived quality judgments was explored. The results are not unequivocal. Partial support is found for the hypothesis that the higher the predictive value of a cue, the more important that cue is in the formation of perceived quality judgments. The hypothesis stating that the importance of a cue in the formation of perceived quality judgments is positively influenced by the confidence value of that cue is rejected. It appears that the measures proposed in the literature to estimate cue predictive value and cue confidence value are of questionable validity (see also below). The hypothesis that intrinsic cues are more important than extrinsic cues in the formation of perceived quality judgments is largely supported. In some cases, however, an extrinsic cue will be more important.No support is found for the hypothesis that the intervening role of quality attributes is greater for consumers having much experience with the product in question than for less experienced consumers. An explanation might be that subjects did not differ enough in experience. This is caused by the sampling procedure adopted. The hypothesis that the intervening role of the quality attributes is greater for quality-conscious consumers than for consumers who are less quality-conscious is supported.It was hypothesized that consumers experiencing high risk in evaluating the quality of the product alternatives have used fewer cues in the quality perception process than consumers experiencing low quality risk. The hypothesis is weakly supported. This result is probably due to the unreliability of the quality risk measure used. The hypotheses stating that higher-educated consumers use more cues and exhibit more cue interactions than lower- educated consumers are supported for saveloy, but not for gammon.Implications of the proposed model of the quality perception processThe model of the quality perception process is of theoretical as well as of managerial importance. It also has implications for public policy.The model gives a deeper insight into the quality perception process. The model allows the researcher to explain cue effects on perceived quality in terms of the intervening role of the quality attributes. Previous studies were not well able to explain the cue effects found. For example, some studies reported a significant effect of packaging on overall perceived quality but failed to explain why this effect occurred. With the present approach, we are able to explain the effect of packaging through the intervening role of the quality attributes sensory perception and keepability. In this way it is possible to go beyond the simple cue effects found and to explore the causes of these effects. The distinction between quality cues and quality attributes, and the further distinction between intrinsic and extrinsic cues, and between experience and credence attributes is important because it enhances our understanding of the way quality perceptions are formed. The model also highlights the influence of variables such as qualityconsciousness and perceived quality risk on the quality perception process.The model serves as a frame of reference in which future research could be integrated. In this way the results of a particular study are more easily generalized to other products and situations. The model has considerable heuristic potential because it is based upon theories and concepts from the literature and integrates these theoretical elements. Numerous hypotheses can be developed on the basis of the model. Many hypotheses are proposed and some are tested in the present study.This distinction between quality cues and quality attributes is also relevant to the marketing manager. It is instrumental in closing the quality perception gap between the company/marketing manager’s perspective and the consumer’s view on quality. Further, it assists the marketing manager in developing a marketing strategy based on quality. The model can be used to investigate which quality cues predict which benefits or attributes to consumers. Advertising could concentrate on those cues on which the brand rates favorably and that predict important quality attributes. The message content should be developed in terms of the favorable cues and the attribute(s) they predict. An empirical investigation based on this model can also provide information that is relevant to product development. The results of such a study can be used to identify the quality cues that are prime candidates for modification to enhance the quality image of the brand. It is also possible to identify those quality cues that are not used by consumers, and hence can be modified to reduce costs. Further, the results need not only be used as input for product and communication strategies. The importance of the place of purchase as quality cue in the formation of quality perceptions can be quantified and the results can be used for developing a distribution strategy. It assists the marketing manager in developing a pricing strategy, especially when information is also obtained about price-perceived quality tradeoffs in consumer choices.The potential of the model is illustrated by the results obtained in our empirical study. Let us consider the possibilities for a firm that produces ovate gammon of a variegated pink and red color with little gloss. Shape is the most important quality cue for sensory perception. Ovate gammon rates much higher on sensory perception than rectangular gammon. The color of the gammon is an important quality cue for perceived fatness. Variegated pink and red gammon rates considerably higher on fatness than pink gammon. Gloss has no sigificant effect on any of the quality attributes. Given these results the firm, in view of the quality of its product, should stress the shape of its gammon. In advertisements, the ovate shape should be linked to superior sensory characteristics. Since sensory perception plays a prominent role in perceived quality judgments, such an advertising campaign will enhance the quality image of the product.The quality image can be further enhanced by product modification. Variegated pink and red gammon is perceived to be fatter than pink gammon and rates lower on sensory perception as well. Since perceived fatness has a negative effect on perceived quality, it might be advantageous to modify this cue. Ovate gammon of a pink color has a better quality image. Such a product modification should be communicated to consumers.Possibilities for cost reduction also exist. The quality cue gloss is hardly used by consumers in the quality perception process. The extent of gloss is positively related to the percentage of brine in gammon. Brine is inversely related to costs since more brine means less meat per unit of weight. The firm could save costs by using more brine in its gammon. This recommendation is supported by the results of a sensory experiment. About 50% of the subjects prefer gammon with a relatively large percentage of brine (30%) because of its hearty taste.The results of the empirical study show that the place of purchase is of limited importance in the quality perception process for gammon. The butcher’s shop rates only slightly higher than the supermarket. This implies that the firm can sell its gammon through both types of outlets without serious loss of quality image. Such a broad distribution will have a positive effect on sales.As might be expected, a high price contributes to the quality image of gammon. In addition, research on the relative importance of perceived quality vis-à-vis price in consumer decision-making indicates that quality is considerably more important than price (see below). This suggests that the firm can sell its (modified) gammon at a high price (more specifically, in our study, at about Dfl. 2.70 per 100 grams).This example illustrates the potential of the proposed model of the quality perception process to provide marketing managers with concrete recommendations about which marketing actions could be taken. The model also draws attention to the influence of consumer characteristics on the quality perception process. Variables such as level of education and quality- consciousness (these were found to influence the quality perception process; see above) can be used for market segmentation. Currently, quality-consciousness is being used together with other variables to segment the market for meat.The model also has implications for public policy. Credence attribute perceptions are more uncertain than experience attribute perceptions. Further, it was found that people have relatively much difficulty in inferring credence attribute perceptions. Thus, consumers have difficulties in forming ‘complete’ quality judgments, i.e., judgments based on both short-term and long-term benefits. This situation is undesirable from a public policy point of view. Ideally, it could be remedied by making available cues that provide direct information about the credence attributes. An example is nutritional information. A problem with this kind of cues is that they frequently have little impact on the quality perception process because they are often pallid and ill-understood, or perceived to be untrustworthy in case the information is provided by the manufacturer. Another possibility is to impose minimum quality standards with respect to credence attributes for some products. The question emerges for which products minimum quality standards would be socially desirable. Economic theory (Leland 1979) suggests that minimum quality standards increase social welfare for products exibiting high demand sensitivity to quality variation, low price elasticity of demand, low marginal cost of providing quality, and/or a low willingness to pay for low-quality brands.Price-perceived quality tradeoffsIn a study about product quality, it is in place to give some attention to the importance of quality in consumer decision-making. Information about the importance of perceived quality in the formation of purchase intentions or choice is relevant to marketing managers since it assists them in developing a marketing strategy. Is a high-quality strategy more profitable or is it advantageous to offer a low-quality/low-priced brand?In this study, the attention is limited to the tradeoff between perceived quality and price. Two investigations were carried out employing different methodologies to explore the way consumers trade off perceived quality against price and to quantify the relative importance of perceived quality vis-à- vis price. The dependent variable in both studies is purchase intention. The first study investigates price-perceived quality tradeoffs for saveloy and gammon, using the correlational approach. The second study explores price- perceived quality tradeoffs for gammon using the information integration approach.It was found that price and perceived quality are integrated in an additive way in the formation of purchase intentions. As might be expected, perceived quality has a positive effect on purchase intention and price has a negative effect. Perceived quality is considerably more important than price. This finding supports the importance of perceived quality as a decision variable in consumer behavior. A disproportionately large increase in purchase intention is found when quality increases from below average to above average. This suggests that it might be particularly profitable to a firm to market a product of aboveaverage quality. This finding supports the drive for quality that can be observed in many markets.The influence of five possible determinants of the relative importance of perceived quality vis-à-vis price were investigated: budget of the household, quality-consciousness, perceived quality risk, the product in question (gammon or saveloy), and the usage goal for which the product is purchased. Qualityconsciousness is the only variable with a significant effect on the relative importance of perceived quality. Subjects who are more quality- conscious attach more importance to perceived quality than subjects who are less qualityconscious. The budget of the household has an indirect effect on the relative importance of perceived quality. Subjects with a larger budget are more quality-conscious, and hence attach more importance to quality.The relationship between price and quality in the marketplaceThe economic approach to quality assumes that price and quality are strongly correlated, and research in the perceived quality approach has shown that consumers frequently use price as quality cue. Thus, it seems appropriate to devote some attention to the veridicality of price as market signal of quality. Are price and quality strongly correlated as suggested by economic theory, and can consumers hence use price as quality cue? To obtain insight into this normative issue we need a measure of objective quality, i.e., quality as measured by standardized techniques and experts. Such a measure is provided by the product test data published by consumer magazines. The data set consisted of 413 product tests involving 6580 brands published in the Dutch consumer magazine Consumentengids over the ten year period 1977-1986.It is found that for 81.4% of the products price and objective quality are positively correlated. The relationship however is in general rather weak. Mean Spearman’s rho between price and quality is only .29 (median .32).Price-quality correlations vary over time. Correlations are lower when the inflation rate is rather high. Further, correlations are higher for durable products than for nondurables. The correlation between price and quality in a product test is positively related to the degree of price variation between the brands for that product. The hypothesized curvilinear relationship between the number of available brands and the strength of the price-quality relationships is not found.In sum, price is a poor market signal of quality for most products. Although consumers’ evaluations of product quality might differ from objective quality, even if consumers would be perfectly informed, due to differences in importance attached to the quality attributes etc., these findings indicate that it will often be a suboptimal strategy to evaluate quality on the basis of price.Suggestions for future researchThis work is concluded with some suggestions for future research in the area of perceived quality.To obtain more knowledge of the quality perception process, we should move from exploratory studies, whose generalizability is limited, to testing of explicitly formulated hypotheses. In this work, a number of such hypotheses are proposed, and others could be developed. It is important that hypotheses are tested within the context of an explicitly formulated model of the quality perception process, such as the one proposed in this work, so that research findings from different studies can be more easily integrated.In our empirical study, all subjects were presented the same list of quality attributes. It is implicitly assumed that all these attributes are relevant to the quality perception of each individual, that the set of attributes is exhaustive for each person (i.e., each person only uses these attributes), and that an subjects attach the same meaning to a particular attribute.The first assumption can be relaxed by allowing the subject to chose from a list of attributes the attributes that are relevant to him/her. Subsequently, s/he rates the product alternatives only on these attributes. Computer interactive techniques such as Adaptive Perceptual Mapping (Sawtooth 1987) can be very helpful in data collection and analysis.One can even go further by relaxing all three assumptions. In that case the subject formulates his/her own attributes and rates the product alternatives on these attributes only. Generalized Procrustes analysis (Gower 1975) can be used to construct a common perceptual space on the basis of the completely individualized data. Individual differences are retained in this technique. This procedure has been employed successfully by Steenkamp and Van Trijp (1988b).Another issue for future research is the development of better measures for cue predictive value and cue confidence value. Previous research as well as the present study have, for the most part, yielded disappointing results with respect to the ability of these concepts to explain cue importance in the quality perception process. It appears that the questions concerning cue predictive value and cue confidence value are ill-understood by subjects. A possible solution might be to let subjects rate the predictive value and the confidence value of each cue level employed in the study separately. The difference between the highest and the lowest predictive value/confidence value rating could be used as a measure of the predictive value/confidence value of that cue. Such a procedure makes these concepts more vivid for the subject. Research on this issue is urgently needed, given the theoretical potential of cue predictive value and cue confidence value in understanding the quality perception process.In this study two measures of perceived quality risk are used, quality risk measured by a modified version of Cunningham’s model and quality risk measured by Bettman’s model. These are the two most influential models of risk in the marketing literature. It is found that the reliability of both models was unacceptably low. The quality risk ratings as computed on the basis of each of the models are weakly correlated. It is important to investigate whether this is a coincidence or that new or modified models of quality risk must be developed.Future research could also expand on the role perceived quality plays in consumer decision-making. The effect of other variables, along with price and perceived quality, in the formation of purchase intention and choice might be explored. An issue especially interesting for consumer research is the relationship between perceived quality and attitude.Another research issue is ‘quality guidance’, i.e., the integration of the perceived quality approach and the production management approach. Companies are increasingly making adjustments in the physical product and marketing strategies to improve the quality image of the product. A key issue is the selection of the appropriate physical product characteristics. For which physical product characteristics is the consumers’ tolerance for deviations from the optimal level the lowest and which ones should be modified to enhance consumers’ perceptions of quality? This requires insight into the linkage of physical product characteristics to quality perceptions. Only when the company has knowledge about these relationships it is able to select the appropriate features. Otherwise, it might modify features that have no linkage to quality perceptions and hence will incur costs without obtaining the desired improvement in the quality image. We believe that in the future quality guidance will become an important issue in marketing.

2020/NL/NL2020_0.rdf