You are on page 1of 27

Technometric Benchmarking: Identifying Sources of Superior Customer Value

Aviv Shoham Shlomo Maital Hariolf Grupp Sharon Lifshitz

SUMMARY. Multi-attribute models are popular aids in developing marketing strategies to enhance customer value. These models rely on self-reported measures of importance of product attributes and the scores of competing brands on these attributes. In this paper, we introduce technometrics as an extension of multi-attribute models. Technometrics is a benchmarked approach to measure the scores of competing brands on a set of attributes. It can be used as an aid in selecting appropriate strategic tools along the price and quality dimensions for existing and modified brands. Importantly, with rapid globalization of numerous product-markets, there is an increase in the importance of measuring competitiveness of international firms objectively. We provide an illustration for the use of this
Aviv Shoham is Senior Lecturer of Marketing, Graduate School of Business, University of Haifa. Shlomo Maital is Professor of Economics, Faculty of Industrial Engineering and Management, the Technion. Hariolf Grupp is affiliated with FraunhoferISI. Sharon Lifshitz is a graduate of the Master of Science in Industrial Management, the Technion. Address correspondence to: Aviv Shoham, Graduate School of Business, University of Haifa, Haifa 31905, Israel (E-mail: ashoham@research.haifa.ac.il).
[Haworth co-indexing entry note]: “Technometric Benchmarking: Identifying Sources of Superior Customer Value.” Shoham, Aviv et al. Co-published simultaneously in Journal of Global Marketing (International Business Press, an imprint of The Haworth Press, Inc.) Vol. 16, No. 1/2, 2002, pp. 1-26; and: Strategic Global Marketing: Issues and Trends (ed: Erdener Kaynak) International Business Press, an imprint of The Haworth Press, Inc., 2002, pp. 1-26. Single or multiple copies of this article are available for a fee from The Haworth Document Delivery Service [1-800-HAWORTH, 9:00 a.m. - 5:00 p.m. (EST). E-mail address: getinfo@haworthpressinc.com].

© 2002 by The Haworth Press, Inc. All rights reserved.

1

2

STRATEGIC GLOBAL MARKETING: ISSUES AND TRENDS

technique, which combines objective and subjective product assessments, in the executive jet industry.[Article copies available for a fee from
The Haworth Document Delivery Service: 1-800- HAWORTH. E-mail address: <getinfo@haworthpressinc.com> Website: <http://www.HaworthPress. com> © 2002 by The Haworth Press, Inc. All rights reserved.]

KEYWORDS. Technometric bench working, product strategy, customer value, international strategy

Customer value is an important positional advantage and can be a key success factor (Day and Wensley 1988). Measuring customer value is based, in many cases, on multi-attribute models. Such models explain attitude formation, which underlies brand preference (Fishbein 1963; Fishbein and Ajzen 1975). Fishbein’s model (1963) arrived at attitude scores based on the multiplication of brand’s performance on attributes and the same attribute’s evaluation. Bass and Talarzyk (1972) replaced the evaluation component with an assessment of the importance of each attribute. Empirical evidence for the superiority of either model is mixed (Bettman, Capon, and Lutz 1975; Mazis and Ahtola 1975). Similar models of consumer choice have been developed in economics (Lancaster 1991). Linear compensatory models have dominated (see Curry and Menasco [1983] and Wilkie and Pessemier [1973] for reviews and Curry and Faulds [1986] and Tversky, Sattah, and Slovic [1987] for exceptions). Given the importance of attribute-level information, it is not surprising that it has been used extensively. Many texts use such models and suggest a variety of strategies to improve a firm’s position in a perceptual map (e.g., Kotler 2000). Prior research has concentrated on the role of product attributes in determining perceived quality. For example, Chang and Wildt (1994) manipulated product attribute information and measured the impact of the quantity of attribute information on perceived value. They found that as more attribute information was given, the importance of price as a determinant of perceived value diminished. Zeithaml (1988) also recognized the importance of attributes and suggested that intrinsic attributes underlie an abstract quality dimension, which determines perceived product quality. Other attribute categorization schemes have also been used. Attributes can be categorized based on accessibility. Attribute accessibility is defined as the availability and usefulness of attribute information to poten-

Attributes at the first and second stage can be used to compare competing brands based on what Pessemier (1977) calls “determinant attributes. Many car models incorporated the cup holders during this phase.g. Many product attributes lose diagnosticity with the passage of time and the maturing of an industry. some product attributes are non-monotone. The second phase is competitive performance. Another categorization is based on attribute diagnosticity.Shoham et al. Accessibility is important because it suggests which attribute information is available to and useable by potential buyers. The first issue of choice of attributes involves their number (how many) and composition (which attributes).” Finally. In contrast. but color preferences cannot be compared across individuals. personal computer users may not know what the designation Pentium IV actually means. This loss is due to three phases of product design (Watson 1993). In a review of 42 empirical studies of multi-attribute models in marketing. choice of weights. an attribute that was innovative is expected by all buyers and loses diagnosticity altogether. They cannot be compared across consumers nor ordered on a utility function. clothing). where there exist truly innovative features that speak to real and. product color is very important to some consumers (e. Innovative performance is the first phase. Attributes at the final phase are not used to evaluate products because they are available in most existing brands (the situation in the car market for cup holders at this time). holders were designed into the Taurus model resulting in a diagnostic attribute for some buyers. The use of multi-attribute models in marketing has been subject to a number of critiques. Consumer Reports groups such attributes under the heading “All Have. unspoken customer needs. It may be inaccessible even to those who know what it means if they cannot assess the resulting computer performance. Wilkie and Pessemier (1973) show that the number of independent attributes as- . Wilkie and Pessemier (1973) show that most studies assume that their list only includes salient attributes without much substantiating evidence. at times. 3 tial customers and is similar to Kotler’s (2000) saliency of product attributes. during which customers compare features that are available in most brands.. For example. Before the launch of the Ford Taurus. and allocation of performance scores. such holders were unsafe add-ons. Earlier critical work falls into one of three major types: choice of attributes. For example. Watson (1993) uses the car coffee-cup holder as an example. Diagnosticity is defined as the extent to which customers can use an attribute to distinguish between competing alternatives. Most prior research included only monotonic product attributes for this reason.” In the third phase.

For example. Hjorth-Andersen 1984. Consumer Reports use an average of 6. Haulman. Another problem arises because not all users are aware of all brands and can only rate brands with which they are familiar. Camp 1989). Archibald. but they agree that the proper choice of attribute weights is extremely important (see also Hjorth-Andersen 1986). Sproles (1986) used Hjorth-Andersen’s data to argue that consumer markets contain a high proportion of inefficient brands that can lead to financial losses. Subjective assessments of evaluative criteria are important since consumers act based on their beliefs regarding performance of competing products (Wilkie and Pessemier 1973). Second. but they sometimes fail to provide managers with the information needed to change strategy. it is important to equalize the context of studies or to sub-set the population into context-similar segments. the major question is whether their ratings are justified. the firm may invest in improving product quality. Furthermore. Curry and Faulds (1986) fault Hjorth-Andersen’s summary methodology. say quality.8 attributes and the similar Danish Rad og Resultater an average of 9. 1986) argued that the process used by testing firms is flawed because it is based on subjective selection of attributes and. Such beliefs can be assessed using competitive benchmarking (Bowman and Faulkner 1994. low ratings are justified for a given product. For example.3 (Hjorth Andersen 1984). the importance of price and quality may vary by respondent’s income. and Moody (1983) voiced similar concerns in their analysis of the running shoes market. more importantly. if users rate a given brand poorly on some attribute. He argues that low correlations between price and quality may have been due to the use of summary.4 STRATEGIC GLOBAL MARKETING: ISSUES AND TRENDS sessed in previous research varied from 2 (of 9 original) to 8 (of 37 original). Thus. on the other hand. the firm may change its advertising strategy to convey a message of superior quality. Consumer Reports-based quality ratings. Hjorth-Andersen (1984. . If. it is well known that attribute importance can vary by context (Bearden and Woodside 1977). multi-attribute models are mostly subjective concerning brand performance scores on each attribute. Sproles 1986). If objective quality is high. Their ratings are limited to known brands and evaluations are comparative only for these brands. This is the basis for the use of multi-attribute models as segmentation tools in many marketing contexts (Wilkie and Pessemier 1973). on non-disclosed attribute importance weights. Third. 1986. the choice of attribute weights was the subject of much debate (Curry and Faulds 1986.

because it is based on readily available public sources. Benchmarked technometrics is designed to measure product attributes (Grupp 1990. because consumers may well be unaware of the full set of competing alternatives. rather than a combination of subjective attribute scores and weights as in present approaches. comparative attribute profiles and perceptual maps of competing brands. Thus. overall product evaluations are based on a combination of objective (attribute scores) and subjective (user weights) criteria. BENCHMARKED TECHNOMETRICS The Technique Grupp and his associates developed the technometric technique in Germany (Frenkel. forth- . or country levels. In the next section. which can provide profiles at brand. Harel. 1988. First. product-line. user-defined importance weights. a menu of possible attribute improvements. Porter 1986). we explain the technique and show why it provides an answer to some of the critiques of multi-attribute models. benchmarked technometrics requires that users only provide importance weights for the various attributes. Because of the more objective nature of benchmarked technometrics and its use of readily available data. firms can use the benchmarked data to improve product management relatively easily. Grupp and Hohmeyer 1986. and technological changes (Mansfield 1988. It is a quantitative. Maital.’ Benchmarked technometrics have two advantages over existing methodologies. Its output is performance data for a given brand.Shoham et al. industry. Second. Grupp. and a triangulation of manufacturerand user-based attribute importance weights. 5 In sum. product/brand attributes are measured objectively. Grupp 1990. Thus. forthcoming. These include the generation of objective. shorter product life cycles. Such issues are discussed in the “Application” section of this manuscript. As explained below. the technique has several practical advantages. compared to competitors. Other concerns are addressed directly by using objective. the technique’s usefulness is most important in international marketing. and Maital 1991). objective approach. any study of product attributes should be concerned with the issues discussed above. and Grupp 1993. Kostachsky. Existing methods require users to be knowledgeable about all available brands and their performance on the various attributes. These advantages are extremely important for internationally active firms because of competitive pressures such as global competition. Frankal. benchmarked attribute measures (Camp 1989) in combination with subjective.

k) is the technometric score for product i on attribute j in industry k. Grupp and Hohmeyer 1986. This theory (Lucas 1972. • Kmin(i. Grupp et al. The formula is: K * ( i. Rosen 1974) views products as utility-generating bundles of attributes. and Grupp in press). j. process. Product attributes are typically measured with differing scales. j . 1] scale.6 STRATEGIC GLOBAL MARKETING: ISSUES AND TRENDS coming. k) = K( i . ) − K min ( i . 1971. the first step in assigning a technometric score is to identify the best performing and worst performing products in an industry on . or industry (Grupp 1990. 1993. Reiss. 1988. Technometric measurement is based on relevant product attributes (we discuss this issue further when our study is presented). To sum. j . k) is the raw score of product i on attribute j in industry k. following the German Ministry of Science’s request for an assessment of the country’s competitiveness in high-technology industries. Saviotti 1985). k) is the raw score of the best performing product on attribute j for all products in industry k. Griliches 1961. Grupp and Hohmeyer 1986. Such studies use the hedonic price index approach (Frenkel et al. such as weight and length. Technometrics is defined as the quantitative measurement of the quality or technological sophistication of attributes of a product. k ) K max ( i . much like the approach used in marketing. • K(i. k ) − K min ( i . Theoretically. It has been applied in the lasers. Maital. j . 1988). k) is the raw score of the worst performing product on attribute j in industry k. j . The technometric technique circumvents these problems by normalizing product attribute-scores to a [0. • Kmax (i. j. which stabilize as products mature (Stankiewicz 1990). benchmarked technometrics builds up the foundations laid by Lancaster (1991) in his modern consumer theory. It has been used to assess the accuracy of consumer price indexes in accounting for changes in product quality from one period to another (Adelman and Griliches 1961). j. k ) where: • K*(i. k . industrial robotics. and bio-diagnostic kit’s industries (Frenkel. 1991). product line. Industry experts usually reach an agreement on such attributes (Grupp 1990). j. j.

. features. Technometrics result in measures that fit Garvin’s criteria. Then. Thus. K*(i. The use of benchmarked technometrics. Lutz and Bettman (1977) point out the significant impact of the existence of publicly available. 1983). Rosen 1982). Lancaster 1991. Garvin (1983.g. Six of these dimensions (objective performance. 7 each attribute. reliability. The approach so far yields technometric scores at the attribute level. Only two incorporate subjective criteria (aesthetics and perceived quality). Muellbauer 1974. j. even after trial. A simple summation of attribute-level scores ig- . Thus. objective data or discrete attributes (see Hauser and Simmie 1981) do not apply for the standard models. Furthermore. after ratings are published. The next step is to combine these scores to arrive at an overall technometric score. Archibald et al. as shown by Katz (1995). Furthermore. The use of benchmarked technometrics makes it possible to measure observable and non-observable attributes based on publicly available attribute information (and. She suggests that dependability and complaint-handling behavior fall into this category. k) is calculated for each product based on its relative performance for each attribute. compare them to user-based attribute assessments). In a “bicentennial” review of papers using multi-attribute models in marketing. popularized by companies such as Ford and Xerox (Zangwill 1993). 1987) argues that quality should be assessed on eight dimensions. while the argument about choice of weights raged. Because of these problems. and serviceability) should be measured objectively. at best. objective attribute information. ordinal and that the distances between points of these scales are not equal. some product attributes are not directly observable. In their view. many argue for the use of objective measures (e. results in cardinal measures of performance. conformance. As Hjorth-Andersen (1986) suggests. These are assigned values of one (for the best) and zero (for the worst). Lucas 1972. durability. firms adjust their advertising levels. Adelman and Griliches 1961. Hjorth-Andersen also argues that many such scales are. which are inherent in subjective quality assessments. at a later stage. The use of objective measures at the attribute level is an important departure from earlier research. Objective measures form the backbone of process benchmarking. (1983) argue further that advertising levels can help consumers locate good buys. Availability of public information has been shown to strengthen the positive relationship between advertising and quality (Archibald et al.Shoham et al. technometric scores are normalized and become benchmarked. no attention was paid to the choice of scales to measure each attribute.

Overall technometric score for product i is T. The region in which A operates represents a discounting strategy with competitors choosing to offer low technometric-score products at low prices. Changes in revenue can then be compared to the costs of the changed product. However. The region in which B operates represents middle-of-the-road competitors. a competitor. To calculate a weighted score. Assuming that firms set prices on the basis of a combination of benchmarked product attribute bundles. we allow a representation of a price-quality relationship. say B. This relationship is drawn with decreasing marginal utility. Figure 1 illustrates how a firm can use a perceptual map on the basis of price and overall technometric score. k = ∑w K * i j (i . we review potential applications of benchmarked technometric scores. or an expansive product for a given technometric score. but the implications are similar for other types of functions. Region C is held by competitors with “premium” offers-high . j . Like most perceptual maps. Point C represents the highest-price competitor with the highest overall technometric score. who charge medium prices for medium technometric-score products. and using Figure 1. k ) Conceptual Applications In this section.8 STRATEGIC GLOBAL MARKETING: ISSUES AND TRENDS nores attribute’s importance. unlike other maps. The region below and to the right of the curve represents a “high-value-for-money” strategy where competitors offer products with a higher quality at a similar price or a lower price for comparable products. When the best (or worst) do not carry the highest (or lowest) price two points may be required to represent each of the extreme regions of the graph. such that: Ti . Figure 1 is based on objective performance data. B can calculate by how revenues will decrease for a given reduction. These can be calculated from market surveys or from manufacturer’s estimates (discussed below). It entails either an inferior product at a given price. the region above and to the left of the curve represents an unfeasible strategy. Conversely. average weights ( w i ) for each attribute are needed. Point A represents the lowest-price competitor with the lowest overall technometric score. can calculate the additional revenues generated by higher prices after increasing attributes that underlie the technometric score. The location of a competitor indicates potential strategies. Long-range.

the firm may need to invest in improving the product. management may not know which of the two strategies is called for. The accuracy of user-based perceptions can then be gauged by comparing subjective and objective assessments of performance. This is important because. Figure 2 includes a combination of technometric scores and attribute’s importance scores. The focal firm scored highly on attributes A1 and A3. k)) are drawn on the other. For exposition purposes. Pmax) PMAX B(Tb. 9 FIGURE 1.Shoham et al. Prices and Technometric Scores: The Value of Attribute Change Price C(Tmax. j. Figure 2 can be used for this purpose. Pb) PB A(Tmin. to the extent that perceptions are justifiably poor. Figure 1 is indicative of feasible positioning. Pmin) PMIN Tmin Tb Tmax Technometric Score technometric scores with premium pricing. received a medium score on at- . Thus. Additionally. To the extent that the objective performance is high and is misperceived. the average weights (wi) for each attribute are drawn on one axis and the firm’s attribute-level technometric scores (K*(i. In Figure 2. subjective. a firm cannot use Figures 1 to select which attributes to improve. Stated differently. Without comparative performance data. user-based data can be super-imposed on Figure 1. managers need to “educate” the market through promotion so that customers adjust their perceptions. Figure 1 is based on each composite technometric score and does not provide attribute-level information (except for price on the second dimension). Feasible regions with few (or no) competitors represent market opportunities for modified or new products.

and. These high scores do little to enhance its overall technometric score. This graphical representation suggests changes in strategy. the firm scores poorly on one important attribute (A4) and only marginally better on another (A5). The firm scores highly on an unimportant attribute (A1) and on a medium-importance attribute (A3). enhancing its overall score implies an effort to improve performance on these two attributes. Thus. . Manufacturer-based wi’s can then be compared to the user-based. on attribute A4. and low scores on A2 and A4. manufacturer’s estimates of the importance of each attribute can also be assessed. self-reported weights. k) A3 A1 A5 A2 A4 wi tribute A5. However. most importantly. To the extent that differences between the two sets of wi’s are identified.10 STRATEGIC GLOBAL MARKETING: ISSUES AND TRENDS FIGURE 2. Another use for technometric scores is in triangulation of attribute weights (wi). On the other hand. Technometric Scores and Attributes’ Importance K*(i. managers in the focal firm need to reassess the marketing mix to emphasize and de-emphasize sub-sets of product attributes. j. These weights are usually generated by averaging user’s responses (within homogeneous sub-samples of the population).

Phillips. This market has several appealing characteristics. if a product is scored poorly by users. and heads of state. and Phillips 1994). billionaires.1 billion on the development of new airplanes . In fact. According to Business Week. extensions to consumer products remain an important topic for future research. Including a second full analysis would have made the current paper too long. the firm may need to adjust its advertising and promotion strategy to convey the message of superior quality. 11 Finally. the authors already have a second database for another industrial good. payload. First. Additionally. operation efficiency. it may cause future erosion in a firm’s market position. maintenance costs. this is just one possible illustration. For example. the study and its findings are described. To the extent that significant differences exist. and takeoff and landing requirements (Phillips et al. we introduce a study designed to illustrate applications of the approach in the executive airplanes market. speed. These are ideal settings for our study because the use of benchmarked technometrics is illustrated in a truly multi-attribute market environment. but not technometrically. Both durable consumer products and other durable industrial goods share many characteristics with the application to executive jets (below) in that objective data readily exists. An introductory discussion about the industry is provided. but not technometrically. the two firms have been spending an estimated $ 1. the products are complexly multidimensional (Phillips.Shoham et al. technometric scores can also be used to triangulate attribute scores generated from market surveys. Notably. Alternatively. The Market for Executive Airplanes Our study concerns the market for executive airplanes. Competition is especially fierce between the two market leaders (Gulfstream and Bombardier). Customers in the marketplace assess and compare performance on multiple attributes. Possible applications in consumer durable markets and other industrial goods also exist. such as range. the results of which are available from the authors. 1994). APPLICATION The study deals with executive airplanes. future strategy may need to be adapted. if a product is scored highly by users. the market is extremely competitive. In the next section. Business Week (Symonds and Greising 1995) estimated that the major aircraft producers are involved in heavy competition for a market that is estimated at 950 multinational corporations. Second. Then.

This group has six major brands: Bae 125-800. Additionally. These jets have a maximum takeoff weight of 18. Canadair 601-RJ. 3. Canadair. can be divided into four groups: Jets. Finally. and Rotorcraft (Forecast International 1992). Bombardier is spending $800 million and Gulfstream $300 million in their respective development projects (Business Week 1995). Thus.000 pounds and cost $ 14-25 millions. Learjet 35A. Third. Bae 125-1000. Prices range between $ 7 millions and $ 13 millions. also termed business aircraft. 2. and Dassault) with one smaller competitor (IAI). Pistons. information about the value of plane attributes should be very useful. Gulfstream.12 STRATEGIC GLOBAL MARKETING: ISSUES AND TRENDS (Symonds and Greising 1995). The business aircraft market provides close to ideal settings in this respect because such aircraft are used similarly by all customers. Light jets. Therefore.500 pounds and cost $ 3-6 millions. Executive airplanes. Heavy jets. Failing to respond to true customer needs may result in significant losses. Beech. and aeronautical engi- . Medium jets. Citation VI. These jets have takeoff limits of over 38. The group includes: Canadair 601-3A. Falcon 50. 1994). this market is entry-unstable. New entrants have been reported to displace first-movers through rapid technological improvements (Phillips et al. attribute-level information is critical even after the introduction of true innovations to the market. Falcon 900. Learjet 31A. Bae. There are six major brands: Citation 550. The jet’s group is divided into three categories: 1. Beechjet 400A. Canadair 601-3AER. pilots. and Learjet 36A. and Gulfstream IV. Data Collection and Technometric Profiles The first data collection stage involved personal interviews with managers in a local manufacturing plant. Maximum takeoff weight is 18. and Learjet 60. competition in this market requires major outlays for research and development.000 pounds. Cessna.501-28. This high level of competition increases the likelihood that competitors will seek to differentiate their products on multiple attributes. Citation 560. attribute weights can and do vary by context (Wilkie and Pessemier 1973). The risks involved in these projects are immense. Citation VII. The world market is dominated by seven large manufacturers (Learjet. as discussed earlier. Astra SP. Turboprops.

seating arrangement. rate of climb. These prices are for what Jane’s terms flyaway or standard versions. total cost per mile. have more comparable scores on all attributes. Wilkie and Pessemier 1973). Scores on attributes within a $ 5% difference between models were constrained to equality. attributes. Learjets have higher scores on physical attributes. Focusing on the benchmarked technometric attribute scores (K*). This phase resulted in a comprehensive list of 27 attributes.1] range. such as industry magazines and manufacturer’s brochures. cruise speed. The Cessna and Beech models. maximum useful load (the difference between non-fueled jet and the maximum landing weight). in the small jets sub-category. First. attribute measures. We note here that benchmarked technometrics were developed by non-marketing scholars. For example. and technometric attribute measures (K*). further discussions with managers.Shoham et al. some models have more balanced profiles than others. . Wilkie and Pessemier 1973). Each model’s scores (K*) are a combination of low and high scores. diagnosticity and contextual use situations in reducing the number of potential attributes (Hjorth-Andersen 1984. proper use of the technique in marketing applications implies that attention be paid to saliency. Price information for the various jets was based on a well-respected industry publishing firm (Jane’s 1993). and resale value. warranties. cruising-speed fuel consumption. However. or spare parts availability are comparable across the brands and offer no diagnostic information to potential buyers. in contrast. but not on comfort attributes. In this study. Scores were benchmarked at the sub-category level into the [0. mach number. and engineers resulted in the deletion of 14 of these attributes. Consequently. noise level. two comments about Table 1 are noteworthy. Second. pilots. cargo volume. 13 neers to identify a comprehensive list of jet attributes as they apply to business usage situations (Grupp 1990. the only guideline they offer for attribute generation (Grupp 1990) is that industry experts should be in agreement about the important attributes to include. Attributes such as engine options. documentation. Attributes were deleted either because they lack diagnosticity or because they may not be salient for all business jet users (Watson 1993). Table 1 lists the models. The final list of attributes included 13 attributes: maximum fuel-load range. Information on attribute scores was gathered from published sources. Attributes such as political considerations and product line width were deleted because they may only be salient to users from some countries or large firms respectively. no one model completely dominates or is dominated within the sub-categories. takeoff and landing distance. cabin volume.

81 5.14 TABLE 1a.301 .376 .543 2.9 5.467 305 57 88.000 1.7 5.534 .478 1.118 .762 .070 3.025 .340 4.98 415 .000 1.92 K* .9 5.924 2.000 1.000 .429 .468 .223 .413 227 34 83.81 4.6 62 3.687 Learjet 31A Value 1.45 335 .75 3.7 5.507 2.000 1.78 K* .972 .000 1.105 1.000 .71 3.972 .000 .000 .302 1.47 263 77 71.530 .000 .80 424 .6 4.140 .31 K* .140 .526 .70 350 .289 . Small Jets–Physical and Technometric Attribute Scores Citation 550 Value Range (Nautical Miles) Payload (’000 Lbs) Cruise Speed (Miles/Hour) Mach Number Climb Rate Takeoff Distance Fuel Consumption (Miles/Gallon) Cabin (Feet) Cargo Noise (EPN dB) Cost per Mile % Resale Value Price (Millions) 1.063 Beechjet 400A Value 1.480 2.000 .091 .000 1.000 .550 .000 .118 .000 1.626 .473 296 67 83.020 4.32 67 4.000 1.063 Learjet 36A Value 2.000 .528 268 40 81 5.301 .450 .688 .000 .000 .717 2.000 .81 4.000 .395 .000 1.683 1.530 1.31 52 4.78 4.1 51 4.418 1.577 1.461 1.428 268 40 83.290 .169 .340 4.000 .000 1.84 K* .000 Citation 560 Value 1.000 .526 .000 1.61 419 .000 1.100 3.98 424 .000 .280 .000 1.684 3.767 .840 .4 62 5.31 52 5.457 .000 .160 .000 Learjet 35A Value 1.12 K* 1.47 K* .000 .000 1.626 .000 .

58 63 9.852 2.000 .68 53 8.699 5.9 8.6 7.572 .95 K* .81 4.TABLE 1b.336 .000 .284 1.000 .69 65 7.030 .000 .000 1.529 1.30 K* .000 .461 .854 .70 402 .855 3.500 Learjet 60 Value 2.213 .800 3.000 .800 3.808 2.000 1.771 .714 1.182 1.560 .000 .000 .202 .35 77 12.19 420 .050 1.577 6.000 1.381 438 61 84.500 5.000 1.000 .700 5.398 .500 Bae 800 Value 2.404 453 64 83 7.000 .20 69 7.480 .417 Bae 1000 Value 3.000 1.000 .835 3.000 .000 .235 1.22 401 .000 .000 .000 .235 1.154 .400 .000 1.99 K* .77 412 .352 675 45 81 10.542 .1 8.000 1.54 K* .187 1.422 365 53 82.000 .867 Citation VI Value 1.427 2.000 1.305 1. Medium Jets–Physical and Technometric Attribute Scores Astra SP Value Range (Nautical Miles) Payload (’000 Lbs) Cruise Speed (Miles/Hour) Mach Number Climb Rate Takeoff Distance Fuel Consumption (Miles/Gallon) Cabin (Feet) Cargo Noise (EPN dB) Cost per Mile % Resale Value Price (Millions) 2.000 15 .844 .000 1.000 Citation VII Value 1.30 409 .000 .000 .000 .50 404 .740 .798 .034 .333 604 40 80.440 2.307 1.582 .835 3.000 5.250 .208 .787 1.690 .095 2.09 65 8.491 .717 .600 .351 438 61 77.000 .90 K* 1.000 1.921 4.481 .95 K* .000 .000 1.3 7.000 1.727 2.000 .

000 .000 .000 .58 78 22.000 .667 1.607 .408 1.930 .172 2.000 1.430 4.169 2.000 1.138 .067 Canadair 601 3AER Value 3.067 Canadair 601 RJ Value 1.000 1.288 5.000 .415 115 79.000 .000 1.20 424 .210 6.9 76 17.28 90 25.167 .286 1.000 .16 TABLE 1c. Large Jets–Physical and Technometric Attribute Scores Falcon 50 Value Range (Nautical Miles) Payload (’000 Lbs) Cruise Speed (Miles/Hour) Mach Number Climb Rate Takeoff Distance Fuel Consumption (Miles/Gallon) Cabin (Feet) Cargo Noise (EPN dB) Cost per Mile % Resale Value Price (Millions) 3.000 .95 K* .008 169 76.000 1.9 76 16.85 4.862 127 79.000 .4 12.141 3.227 1.000 .8 17.012 1.415 196 81 12.000 1.231 1.87 4.641 1.000 1.128 .98 K* .014 5.3 12.000 .00 424 .286 1.178 1.175 .000 4.648 .66 459 .278 845 115 84.363 .259 5.506 .000 .000 .88 4.067 Falcon 900 Value 3.56 430 .000 .509 .503 4.280 .415 115 79.863 .200 Gulfstream IV Value 4.85 4.86 3.85 3.126 .9 76 16.741 .845 3.000 .443 5.000 1.431 .440 1.000 1.875 .39 K* .000 .286 1.700 .000 .148 1.000 1.000 1.75 424 .000 .593 1.000 .125 .000 .400 .468 .000 Canadair 601-3A Value 3.363 .8 18.009 .8 12.000 .000 .000 1.973 12.64 410 .652 .071 3.000 1.000 .00 K* 1.000 .000 1.264 1.706 .75 K* .50 K* .73 75 14.

286 7.500 8.000 Pilots’ Weights 8.120 7. Notably.500 7.357 8.422 6.549 8. 17 Calculating Composite Technometric Scores–Jets Table 1 provides a profile at the attribute level (K*).718 6. Such weights can be used to combine attribute-level technometric scores into a composite score. manufacturers may have responded to the questionnaire with a “quality control” orientation whereas pilots may have been “usage” oriented.493 8. None of the seven firms agreed to provide user’s lists because such lists were trade secrets. because of these differences. However.286 8.786 8.929 8. Table 2 lists average attribute weights for this group. Forty pilots of a national airline and forty military TABLE 2. Therefore.000 6. Earlier. both groups were sampled.912 7.422 .214 7.408 8. Attribute Weights–Jets Manufacturers’ Weights Range (Nautical Miles) Payload (’000 Lbs) Cruise Speed (Miles/Hour) Mach Number Climb Rate Takeoff Distance Fuel Consumption (Miles/Gallon) Cabin (Feet) Cargo Noise (EPN dB) Cost per Mile % Resale Value 7. Two groups of pilots were used to represent users since they are knowledgeable about the industry and approximate actual user’s profiles. a comparison of both group’s weights may serve to identify manufacturer’s production myopia.930 7.056 7.667 7.887 7.338 6. Additionally. Four questionnaires were mailed to each manufacturer in the three sub-categories with instructions to distribute each copy to knowledgeable individuals in the firm. However.667 8. Fifteen questionnaires from seven firms were returned. decision weights are necessary to use the tools in Figures 1-2. each manufacturer was asked to provide a list of 20 customers for the user survey. This represents a response rate of 34% at the individual level and 67% at the firm level.Shoham et al. it was argued that it might be useful to compare manufacturer’s and user’s attribute weights.

547 .585 .576 .513 .515 . TABLE 3. providing managers with insights on how well their products do at the attribute level.695 .294 .499 .18 STRATEGIC GLOBAL MARKETING: ISSUES AND TRENDS pilots were asked to complete the questionnaire.457 .545 .541 . Applying Benchmarked Technometric Tools Figure 3 is an illustration of technometric scores for range in the light jets category.722 Pilots' Weighted .510 .731 .506 .594 .452 . Similar figures can be plotted for all attributes.545 .507 . They were averaged within the manufacturer’s and pilot’s groups to arrive at overall scores (Table 3).558 .517 .509 . The differences between the two groups were not significant and they were combined. Weighted overall technometric scores were then calculated for each model.516 .525 .569 . Overall Averaged Technometric Scores–Jets Manufacturers' Weighted Citation 550 Learjet 31A Citation 560 Learjet 35A Learjet 36A Beechjet 400A Astra SP Citation VI Learjet 60 Citation VII Bae 800 Bae 1000 Falcon 50 Canadair 601-3A Canadair 601-RJ Canadair 601-3AER Falcon 900 Gulfstream IV .498 .478 .299 .509 .686 . Table 2 lists their average attribute weights.569 . Values range from 0 (for Beechjet 400A) to 1 (for Learjet 36A).449 .454 . Thirty-four of the former (85%) and 39 of the latter (98%) returned complete questionnaires for an overall response rate of 91%.527 .500 .533 .

4 0. Two competitors (Learjet 60 at the low end and Citation VI at the medium level) are close to optimal positioning.3 0. and noise) are low or medium.5 Scores 0. Light Jets–Range Technometric Scores 1 0. but not otherwise. Learjet 60’s position can be improved. The last two (Bae 800 and Bae 1000) are in an unfeasible long-range position in that their prices are too high for their quality.Shoham et al. Its scores on important attributes (cabin. Moving from the overall technometric score to attribute level scores.7 0. Figure 5 depicts how well Learjet 60 does. As is evident in Figure 5. FIGURE 3.8 0. .6 Range Technometric 0. range. It provides an overview of how well each competitor is doing in relation to all others (since the overall scores are benchmarked). It does well on two unimportant attributes (climb rate and cargo) and two medium-importance attributes (cruise speed and cost per mile) and has low scores on two unimportant attributes (pay load and resale value). These two models may have a viable market if their higher ranked attributes are important to a sub-sample in the market. Attributes are presented from the least (climb rate) to the most (noise) important as judged by the pilots. Figure 5 can be used to assess whether Learjet 60 does well on important versus unimportant attributes.1 0 Citation 550 Learjet 31A Citation 560 Learjet 35A Learjet 36A Beechjet 400A 19 Figure 4 maps the six medium-level jets into a price and overall pilot’s weighted technometric score. fuel consumption.2 0.9 0. Improving the product’s specifications on these attributes will improve its overall score. Two competitors provide value-for-money (Astra SP and Citation VII) because they could have charged higher prices for their levels of technometrically assessed quality.

8 Pilot’s Weighted Overall Technometric Score FIGURE 5. Prices and Overall Technometric Score for Medium Jets P 14 r i 13 c e 12 Bae 1000 11 10 Bae 800 9 Citation VII 8 7 .8 0.4 0.2 0.6 Attribute Scores 0.9 0.4 .20 STRATEGIC GLOBAL MARKETING: ISSUES AND TRENDS FIGURE 4.5 0.7 0.7 .5 Learjet 60 Citation VI Astra SP .6 .1 0 Climb Mach Pay Cargo % Resale Takeoff Cruise Cost Cabin Range Rate Number Load Value Distance Speed per Mile Fuel Noise .3 0. Technometric Scores and Attribute Importance–Learjet 60 1 0.

Manufacturers’ versus Users’ Based Attribute Weights 9 Manufacturers’ Weight Users’ Weights 8.056 respectively). LIMITATIONS AND DIRECTIONS FOR FUTURE RESEARCH Limitations.5 7 6. First. Reducing performance on the two will harm overall positioning by less. none of the responding firms agreed to FIGURE 6. two weights stand out in that the differences between the two groups are large. Decisions to change performance on these two attributes should be made with caution. manufacturer’s weights are close to user’s weights.912 and 6. 21 Another important use of the market surveys of both manufacturers and users is for graphing attribute weights. However. The findings of our study should be interpreted in light of several limitations. Differences in attribute importance weights also lead to the observed differences in overall technometric scores (Table 3). but differences in overall scores are small.500) compared to pilots (6.Shoham et al. observing their causes necessitates an analysis such as in Figure 6. Figure 6 depicts the average attribute weights from both samples. Manufacturers assign a much higher weight to payload (8.5 8 Attribute Importance 7. Improving performance on the two will do less to improve overall positioning than what manufacturers probably think. As evident from Figure 6.667) and climb rate (7.5 6 Climb Mach Rate Number Pay Cargo % Resale Takeoff Cruise Cost per Cabin Range Load Value Distance Speed Mile Fuel Noise .

Objective and comparative profiles of competing brands at the attribute level. and costs into one framework. providers of “ideal” price-for-value (around the curve . prices. attribute-score component to such models. However. This profile can be derived from readily available public information for either domestic or international competitors. Thus. as noted previously. in which competitors may be more vulnerable to an aggressive market entry strategy. it can also be used to suggest new international markets. Many of these uses are important for either domestic firms facing competition from multi-national firms in their home markets. This should be relatively easy for manufacturers because they can query existing and potential customers. the technique could be useful for international firms competing in rapidly globalizing world markets. we believe that the biases are minimal and do not harm the exposition of the method and its use. For example. The uses we highlighted earlier include: 1. Benchmarked technometrics adds an objective. In real life applications. A conceptual map of competing brands. can firms set higher (lower) prices for improved (reduced) performance on a given attribute? By how much? Would the additional (lower) costs be lower (higher) than the additional (reduced revenues)? This limitation suggests that future research is needed to combine benchmarked technometrics. they were asked to play the role of potential users.22 STRATEGIC GLOBAL MARKETING: ISSUES AND TRENDS identify user lists. firms should base the analysis on the answers of real customers. While the data could be used to create profiles in markets the firm already competes in. Furthermore. Price-scaled perceptual maps (Hauser and Simmie 1981) may be a promising approach to this issue. Managerial Implications. objective data about competitor’s offerings can be garnered at international tradeshows. The map separates competitors into providers of added value-for-money (below the curve in Figure 1). Potential managerial uses of this approach were suggested conceptually and illustrated. In other words. All that is required is a visit to competitor’s booths and asking for their promotional literature. Second. This paper introduced benchmarked technometrics as an extension of the multi-attribute models. Consequently. Additionally. 2. It may be that army and commercial pilots do not represent accurately the perceptions of the target market. benchmarked technometrics do not provide an answer to the question of improvement’s payoff. we base our user’s data on a pilot sample. sampled pilots are experienced and knowledgeable.

W. On an Index of Quality Change. 1972. and Published Quality Ratings. it could strive to maintain a different positioning strategy that would emphasize different technometrically based advantages in each market according to these differing needs. Alternatively. Journal of Marketing Research 9 (February). 167-77. Moody. Capon. and C. and those that hold unfeasible positions (above the curve in Figure 1). to the extent that the set of competitors and product offerings is similar across national markets. Advertising. 4. in: A. and Z. B. American Statistical Association Journal (September). Talarzyk. if needs differ. the approach proposed here could be useful in the context of the debate about standardization versus adaptation of international marketing strategies (Shoham 1995. Should differences be market specific. G. user-based data can be misleading.. E. A. C. Quality. 1983. Bettman. O. Bass. Price. Reading. I. Technometric scores. 3. Additionally. especially when the set of competitors differs across such markets.. Differences between the two sets of weights (Table 2) suggest that inside. M. R. and A. Jr. F. Haulman. NY: North Holland. Situational Influences on Consumer Purchase Intentions. N. similar maps can be created for specific national markets. Bearden. Here. (eds. R. MA: Addison-Wesley. An Attitude Model for the Study of Brand Preference. too. and J. 1996). Woodside. that account for these differences.and user-based attribute importance weights. A menu for improvements in a given product. J. Griliches. different menus can be identified in different international markets. REFERENCES Adelman. W. 23 in Figure 1). 1977. 1961. 535-48. Woodside et al. . can be used to identify attributes that can be improved and those in which improvements will pay off most handsomely (Figures 2 and 5). 93-96. a firm should consider standardized marketing strategies if customer needs are homogenous across the same markets.Shoham et al. in combination with attribute’s importance. 347-56.R.) Consumer and Industrial Buyer Behavior. Lutz. G. varying international marketing strategies may be required. Archibald. Journal of Consumer Research 9 (March). An Information Processing Theory of Consumer Choice. 1975. For example. Finally. and W. A triangulation of manufacturer.

J. ______ and O.24 STRATEGIC GLOBAL MARKETING: ISSUES AND TRENDS Bowman. Grupp. ______ and ______. D. Long Range Planning 27 (1). Assessing Advantage: A Framework for Diagnosing Competitive Superiority. 1990. Price.). M. Journal of Consumer Research 10 (June). Amsterdam. ASQC Quality Press. S. Menasco. Cambridge. K. Maital. NY: N. Germany: Fraunhofer Institute. Israel: Technion-Israel Institute of Technology and Karlsruhe. 1983. Technological Forecasting and Social Change 30. Indexing Product Quality: Issues. Reiss. B. Neaman Institute. Van Rann (Ed. Handbook of Quantitative Studies of Science and Technology. S. A. S. R. ______. S. 1994. T. and H. R. 1971. and K. in: Sigurdson (ed. 65-75. Human Relations 16. Technometrics as a Missing Link in Science and Technology Indicators. 1-20. 1961. London. Journal of Consumer Research 13 (June). Theory and Results. 134-45. Z. Wensley. Wildt. forthcoming. Hohmeyer. An Investigation of Relationship between Beliefs about an Object and the Attitude toward the Object. Intention. G. Holland: Elsevier. and R. Maital. Chang. 1993. J. Research Policy. Benchmarking: The Search for Industry Best Practices that Lead to Superior Performance. H. Journal of Marketing 52 (April). and S. 1991. Measuring Product Advantage Using Competitive Benchmarking and Customer Perceptions. Koschatsky. A Paper Presented at the MIT Workshop for Research in Management of Technology and Innovation. MA: Addison-Wesley. Ajzen. Grupp. A Technometric Model for the Assessment of Technological Standards. Faulds. ______ and M. 1992.B. Competing on the Eight Dimensions of Quality. and D. T. 1988. 16-27. Quality on the Line. 123-37. Curry. Product Information. ______. In: A.E. Harvard Business Review 65. ______. and A. J. Camp. Koschatsky. A. C. Reading. Faulkner. Maital.. Research Policy. 1983. MA: Harvard University Press. How Good Are Our Odds?–Technometric Evaluation of Product Quality. Price Indexes and Quality Change. Frenkel. 1989. Day. ______ and L. Technological Standards for Research-Intensive Product Groups. Haifa. . Forecast International. The Price Statistics of the Federal Government. On the Separability of Weights and Brand Values: Issues and Empirical Results. Griliches. Garvin. UK: Pinter. Technometric Evaluation and Technology Policy: The Case of Biodiagnostic Kits. 57-76. 173. Harvard Business Review 61.R. Identifying the Sources of Market Value for Science-Based Products. ______ 1987. and D. 1994. 1986. 233-40. 1988. A. 1986. and Purchase Intention: An Empirical Study. D. F. 1963. Civil Aircraft Forecast. E.) Measuring the Dynamics of Technological Change. and Behavior. H. a report. 101-9. Frenkel. Journal of the Academy of Marketing Science 22 (1). Fishbein. 83-95. ______. Sloan School of Management. Hedonic Price Indexes for Automobiles: An Econometric Analysis of Quality Change. forthcoming. Harel. Grupp. Belief. Attitude. The Measurement of Technical Performance of Innovations by Technometrics and Its Impact on Established Technology Indicators. 1975. A Working Paper. 119-32. Massachusetts Institute of Technology.

91-119. and P. and O.” American Economic Review 64 (December). Wage-Rates. and J. The Concept of Quality and the Efficiency of Markets: Issues and Comments. 1990. Modern Consumer Theory. NY: John Wiley and Sons. in: A. 53-73. Mazis. Dordrecht. More on Multidimensional Quality: A Reply. An Approach to the Measurement of Technology Based on the Hedonic Price Method and Related Methods. B. Journal of Consumer Research 13 (June). Lucas. Pessemier. Aldershot. E. 1975. 34-55. A. C. Porter. 1995. Journal of Political Economy 82 (1). The Netherlands: Kluwer Academic Publishers. The Speed and Cost of Industrial Innovation in Japan and the United States: External Versus Internal Technology. 309-34. Sproles. E. R. Rosen. 1986. Muellbauer. Englewood Cliffs: NJ. Shoham. ______. B. 1988. 1984. and T. S. 1972. Hedonic Prices and Implicit Markets: Product Differentiation in Pure Competition. A. Measuring the Dynamics of Technological Change. UK: Elgar. J. E. Journal of Consumer Research 11 (September). J. Phillips.) Consumer and Industrial Buyer Behavior. . J. 1996. Basic Technologies and the Innovation Process. NY: North Holland. T. What Is Strategy? Harvard Business Review 74. 137-150. Simmie. 1977. Jane’s All the World’s Aircraft 1993-1994. Journal of Consumer Research 13 (June). Management Science 27 (1). Paper presented at the Conference of the European Association for Consumer Research. Ahtola. VA: Jane’s Information Group. Biz Jets. 13-38. and Human Capital: A Hedonic Study. Working Conditions. 33-56. Technological Forecasting and Social Change 29.). London. 1981. Product Management. R. Stankiewicz. Mansfield. R. M. M. Phillips. P. 977-94. Quality. Katz..Shoham et al. 61-78. Phillips. 1986. Household Production Theory. 1991. 1995. 1977. Woodside (Ed. ______. Alexandria. Marketing Management. Kotler. A Comparison of Four Multi-Attribute Models in the Prediction of Consumer Attitudes. H. A. The Impact of Brand Attitude and New Information on Product Evaluation. P. Marketing Mix Standardization: Determinants of Export Performance. 1993. Global Marketing Standardization. Journal of Global Marketing 10 (2). 1974. 38-52. 1985. R. K. UK: Pinter.T. Copenhagen (June).I. An unpublished doctoral dissertation. in: Sigurdson (Ed. 1994. and the “Hedonic Technique. P. Journal of Global Marketing 9 (1/2). R. G. 149-54. Journal of Consumer Research 2 (June). Hjorth-Andersen. 146-8. 1974. 708-18. Multiattribute Models in Marketing: A Bicentennial Review. 2000. R. Lutz. Saviotti. J. 25 Hauser. Prentice-Hall. A. B. M. Bettman. The Concept of Quality and the Efficiency of Markets for Consumer Products. P. 1986. Lancaster. Profit Maximizing Perceptual Positions: An Integrated Theory for the Selection of Product Features and Price. Management Science 34: 1157-68.

1988. Watson. A. 1993. Zeithaml. Sattah. Slovic. Lightning Strategies for Innovation. W. NY: Lexington. Journal of Marketing Research 10 (November). A. 2-22. A Dogfight over 950 Customers. Contingent Weighting in Judgment and Choice. C. Business Week (February 6). Journal of Marketing 52 (3). I. A. Greising. H. Quality. G. Psychological Review 95. L. W. W.. Consumer Perceptions of Price. and Value: A Means-End Model and Synthesis of Evidence. . Strategic Benchmarking. Issues in Marketing’s Use of Multi-attribute Models. and E.26 STRATEGIC GLOBAL MARKETING: ISSUES AND TRENDS Symonds. 1987. S. Wilkie. Tversky. 371-84. NY: John Wiley & Sons. Pessemier. 1993. 1973. Zangwill. 1995. 428-41. V. and D. 39. and P.