product attributes model: a tool for evaluating brand positioning

Product Attributes Model: A Tool for Evaluating Brand PositioningAuthor(s): Carol F. Gwin and Carl R. GwinSource: Journal of Marketing Theory and Practice, Vol. 11, No. 2 (Spring, 2003), pp. 30-42Published by: M.E. Sharpe, Inc.Stable URL: http://www.jstor.org/stable/40470109 .

Accessed: 30/09/2013 02:57

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp

.JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact [email protected].

.

M.E. Sharpe, Inc. is collaborating with JSTOR to digitize, preserve and extend access to Journal of MarketingTheory and Practice.

http://www.jstor.org

This content downloaded from 148.61.13.133 on Mon, 30 Sep 2013 02:57:42 AMAll use subject to JSTOR Terms and Conditions

http://www.jstor.org/action/showPublisher?publisherCode=mes

http://www.jstor.org/stable/40470109?origin=JSTOR-pdf

http://www.jstor.org/page/info/about/policies/terms.jsp


PRODUCT ATTRIBUTES MODEL: A TOOL FOR EVALUATING BRAND POSITIONING

Carol F. Gwin Baylor University

Carl R. Gwin Baylor University

A brand's positioning is designed to develop a sustainable competitive advantage on product attribute(s) in the consumer's mind. Perceptual maps are commonly used to evaluate a brand's positioning. Another tool from the economics literature, the product attributes model (based on Lancaster 1966, 1979), utilizes three components: attribute ratings, budget constraint, and indifference curves. The key advantage of this analysis over perceptual maps is the ability to incorporate the impact of price into assessment of the brand's positioning. This paper discusses how the product attributes model helps managers understand the strategic implications of positioning decisions and provides an example of its use.

INTRODUCTION

Positioning "refers to how customers think about proposed and/or present brands in a market (Perreault and McCarthy 1999)." Through a brand's positioning, a company tries to build a sustainable competitive advantage on product attribute(s) - tangible or intangible - in the mind of the consumer. This advantage is designed to appeal to one or more segments in that product category. Developing a successful positioning strategy is not easy; in fact, Schnedler (1996) states: "Positioning products in a complex market can be one of a company's most difficult decisions."

To effectively position (or reposition) a brand, the company must know how this brand is perceived in relationship to other brands in the product category. Several tools are available for evaluating a brand's positioning: multidimensional scaling, factor analysis, discriminant analysis, and multiattribute compositional models are the most commonly used. These

tools have relative advantages and disadvantages that make them more suitable to specific applications as summarized in this paper. Based on these tools, the marketer can evaluate whether:

• The brand has a positioning that is differentiated from other brands in the market.

• Potential opportunities exist for the introduction of new products or repositioning an existing brand.

• Certain segments are underserved by existing brands in the category.

One tool that has not achieved common use is Lancaster's model of consumer demand (1966, 1979), referred to in this paper as the product attributes model. This model incorporates price and the consumer budget into the analysis of the competitive positioning for brands. As such, it has advantages over other methods when price or value is a key issue in brand positioning. This paper provides an overview of the theoretical

30 Journal of Marketing THEORY AND PRACTICE



basis for the product attributes model. Our primary contribution is a discussion of how the product attributes model can be used to make strategic decisions in brand positioning or repositioning. We further show the results of an application to the automobile industry. This model allows the marketer to better understand the customer-perceived value for the attributes in light of the price of the brand. From a marketing strategy standpoint, this understanding has specific implications for product design/re-design and pricing.

THE IMPORTANCE OF POSITIONING

Urban and Hauser (1993) state, "Positioning is critical for a new product. Not only must a new product deliver the benefits the customer needs, but it must do so better than competition (p. 202)." In developing a positioning, the marketer must consider four things:

• The target market. • How the product is different or better than

competitors. • The value of this difference to the target market. • The ability to demonstrate or communicate this

difference to the target market.

These elements roughly relate to the components of a brand position as described by Aaker (1996): target audience, subset of identity/value proposition, create advantage, and actively communicate. The brand identity and positioning are central to developing a strong customer base and brand equity.

The target market and the perceived differentiation from competitors are core concepts of positioning. Rao and Steckel (1998) define a brand's positioning as "how it is perceived relative to its competitors by a relevant group of customers. Segmentation and positioning are too often treated as independent concepts in practice and in the literature. However, positioning has no value unless it is appropriate for a target segment (p. 36)."

The concept of value to the customer is central to an effective positioning. Further, the brand must be able to communicate this value effectively to at least one segment of customers to get customers to switch to the brand. Aaker (1998) states, "A differentiation strategy must add value for the customer, and the added value must be perceived by the customer (pp. 164- 165)."

Differentiation for new and established products typically takes the form of attributes that can be built into the product design to deliver benefits to the customers. The differentiation can be in quality (higher or lower), additional features, packaging innovation, to name a few. Or if the firm has a cost advantage relative to competitors, the product may be positioned on value by offering a lower price.

One of the key tasks of a new product manager is to define a brand's positioning, and managers of existing brands must

continue to monitor the effectiveness of the brand positioning and reposition when needed. Several methods have been developed to help marketers evaluate a brand's positioning. These methods are described in the next section.

Methods for Evaluating a Brand's Positioning

Keon (1983) describes four primary methods for evaluating a brand's current or potential positioning. The methods are: multidimensional scaling, factor analysis, discriminant analysis, and multiattribute compositional models. Table 1 provides a summary of the input data, analysis, advantages, limitations, and best use for each method as drawn from the literature. The summary is primarily based on Keon (1983) and Hauser and Koppelman ( 1 979) with additions and updates from other papers as shown in the table. For each method, a paper or two using or discussing this method is provided as a reference.

According to DeSarbo, Young and Rangaswamy (1997), multidimensional scaling is the primary method used for positioning. For further information on specific analytical techniques for multidimensional scaling, refer to Bujmolt and Wedel (1999) and Carroll and Green (1997) for a detailed comparison.

Huber and Holbrook (1979) compare and contrast factor analysis and discriminant analysis with the conclusions presented in Table 1 . According to this analysis, discriminant analysis is more likely to yield objective dimensions while factor analysis is more likely to yield affective dimensions. These findings lead to the conclusion about the best use for each method.

Conjoint analysis is the most common approach in terms of multiattribute compositional models and is described in the table. For a taxonomy of conjoint methods, refer to Carroll and Green (1995). Green and Krieger (1989) conclude that conjoint analysis provides practical advantages over multidimensional scaling in terms of measurement and data collection for evaluating brand positioning. These authors suggest that the combination of conjoint for analytical purposes and multidimensional scaling for display purposes offers an attractive combination for researchers.

Hauser and Koppelman (1979) conclude that attribute-based techniques such as factor analysis and discriminant analysis provide better measures of consumer perceptions than similarity techniques such as multidimensional scaling if the set of attributes is reasonably complete. In addition, Hauser and Koppelman (1979) show that factor analysis is typically better than discriminant analysis. In our table, we provide a description of the "best use" for each method, recognizing that no one approach is best in all circumstances. Each method has strengths and limitations that make it more or less appropriate in specific decision making situations.

Spring 2003 31



One limitation across all of the methods (with the exception of the Lancaster model described later in this paper) is the failure to consider price in the analysis, except as an additional attribute or dimension. One exception is a study by Hauser and Gaskin (1984) that used price in developing perceptual maps with factor analysis. Hauser and Gaskin (1984) use attributes per dollar as variables. In this analysis, they find that using attributes per dollar provides a better prediction of share than traditional methods or than the use of price as an additional dimension. Urban and Hauser (1993) discuss value maps and show that these maps provide guidance to new product teams in setting an initial price for a new product.

In looking at choices across product categories, Hauser and Urban (1986, p. 448) conclude that price is appropriate as a dimension when "price itself carries utility, such as in conspicuous consumption or when perceptions of quality are based on price." Otherwise, consumers prioritize choices based on the surplus of utility over price.

The purpose of this paper is to demonstrate the applicability, and advantages and limitations, of the Lancaster model (1966, 1971) in evaluating brand positioning. The Lancaster model would be categorized as a multiattribute compositional model in the methods described in the table. This model has been discussed and extended in the marketing literature (Taylor and Haines 1975; Hendler 1975; Ratchford 1975, 1979; Ladd and Zober 1977; Hauser and Simrnie 1981). The key advantage of this method over other methods is that the model explicitly considers price in evaluating the positioning of brands. Our paper adds to previous research on the Lancaster model in the marketing and economics literature in two ways. First, we provide an overview of the model and discuss its usefulness to the strategic issues surrounding brand positioning. Second, we provide an example of the model and show how Lancaster's model can be applied to brand positioning or repositioning questions.

OVERVIEW OF THE PRODUCT ATTRIBUTES MODEL

The model in Lancaster ( 1 966, 1 979), which we refer to as the product attributes model, assumes that consumer choice is based on the c haracteristics ( or attributes) of a brand. The consumer derives utility based on the level of these attributes existing in the brand that she/he chooses within a budget constraint.

In the model, the consumer maximizes utility (U) in her/his choice of a brand. This choice is subject to the consumer's budget constraint (k), which must be greater than or equal to price (p) in combination with the quantity of the brand purchased (x). The choice is based on the vector of product attributes (z) that maximizes the consumer's utility based on the perceived levels of the attributes in the brand and the quantity t hat c an b e p urchased. B r epresents t he m atrix o f products-to-attributes transformation coefficients b¡j, showing the level of z'th attribute a¡ attained from the consumption of

one unit o f/th product Xj. A simple form of this model is shown below (Lancaster 1966):

Maximize U(z) (1)

Subject to px < k (2)

With z = Bx (3)

z, x > 0. (4)

There are three key components to the analysis for the product attributes model: the perceived levels of the product attributes for a brand, the budget constraint (referred to as the efficiency frontier in this application), and the indifference curve of a consumer or segment of consumers. The following sections discuss each of these elements briefly in terms of explicating the concept and its importance to the model.

Attributes of Products

Lancaster (1966, 1971, 1979) shows that consumers have preferences for characteristics (or attributes) of products. Each product is a bundle of attributes. For example, automobiles and motorcycles differ in gas mileage, horsepower, styling, safety, and so on. Rather than comparing the products themselves (as in indifference curve and budget constraint analysis), the theory assumes individuals choose among the more basic attributes of the products. Understanding why a consumer chooses a product based upon its attributes helps us to understand why some consumers have preferences for specific brands. This allows an analysis of brand competition.

The product attributes model explains individual choice as a process of choosing bundles of product attributes inherent in goods and services. The model assumes that consumer choice is b ased on m aximizing u tility ( or t he 1 evel o f s atisfaction received) from the product attributes subject to a budget constraint. The model is particularly useful in analyzing differentiated product markets. These markets have brands that are substitutes for each other and are distinguished by their makeup of a specific set of characteristics. Understanding consumer preferences for attributes that distinguish among brands can help in defining the best positioning and marketing mix for a particular brand.

To illustrate the model, we will use a simple example with only two product attributes. In this case, a two-dimensional graph reveals the model's main features and links it to the traditional indifference curve and budget constraint analysis of individual choice.

Figure 1 shows three products (A, B, and C), each offering specific amounts of attribute X and attribute Y. Each product is depicted in attribute space as a ray from the origin. The slope of each ray is determined by the ratio of attribute Y to attribute X. The rays are linear if we assume each product




TABLE 1 Summary of Methods for Determining a Brand's Positioning1

Method Input Analysis Advantages Limitations Best Use Multidimensional scaling Similarity Based on 1. The resulting 1. Dimensions are often Established categories (cf. Green, Wind, and or proximity - more dimensions are difficult to interpret. with numerous brands Jain 1972; Green, Wind, preference similar brands are based on 2. Repositionings and new in the evoked set. and Claycamp 1975; ratings placed closer consumer products are hard to evaluate • Existing brand Moinpour, McCullough, together. judgments or because the analysis only and/or and MacLachlan 1976) preferences of the considers existing dimen- attributes.

brands. The sions of existing brands, results are not 3. Direct similarity judgments dependent on have been shown to have attributes asked. issues with reliability

2. Maps are (Summers and MacKay constructed to 1976). show consumer 4. Data collection is difficult ideal points and for the researcher and brands on the respondent. Respondents same map. have to rate or rank order

numerous pairs of objects in terms of similarity or dissimilarity. (Katahira 1990)

5. The analysis requires at least 7 or 8 brands to get a map with 2 or 3 dimensions. This is a problem if the evoked set is small. (Hauser and

Koppelman 1979) Factor Analysis Ratings of Determines key 1. The product The factors are a function of the Promotional and (cf. Hauser and Urban brands on dimensions based dimensions are data collected (i.e., the attributes communications 1977; Huber and attributes on the explana- relatively easily asked) rather than the strategies because it Holbrook 1979) tion of variance determined from importance of the attributes to yields more affective

in the total the factor consumers. dimensions (Huber and attribute set using loadings. Holbrook 1979). correlation- 2. The analysis • Potential gaps among the reduces many or opportunities attributes. The attributes to key for new factor scores are dimensions. products, the input for the

maps. Discriminant Analysis Ratings of Determines the The dimensions are 1 . The factors are a function of Engineering or new (cf. Pessemier 1977; brands on linear based on attributes the data collected (i.e., the product design Katahira 1 990) attributes combinations of that differentiate the attributes asked) rather than because it yields more

attributes that brands. the importance of the objective dimensions best discriminate attributes to consumers. (Huber and Holbrook among brands. 2. Dimensions are not easily or 1979).

intuitively interpretable into managerial actions.

3. Dimensions are based on attributes that are different across brands regardless of relevance to preference or

choice. Multiattribute Tradeoffs Determines the 1 . The analysis pro- 1 . It requires other analysis to New product design Compositional Models for utilities of vides insight about show brand positioning in • Product re-design (The most commonly consumers combinations of consumer prefer- relation to competitors and or repositioning used is conjoint analysis.) between attributes. enee among attri- segments, (cf. Wilkie and Pessemier attributes butes and brands. 2. The data collection for a 1973; Green and DeSarbo and product 2. The analysis can large number of brands could 1978; Green, Carroll, and configur- provide be onerous for the Goldberg 1981) ations simulations for respondent and costly to the

determining the researcher, optimal combination of attributes for segments of

I I consumers.

1 Keon (1983) and Hauser and Koppelman (1979) formed the primary basis of this table, as supplemented by other papers shown.

Spring 2003 33



FIGURE 1

s

/ ' ,/' y- Efficiency / i / 'm Frontier

v Attribute X AA

offers a constant ratio of attribute Y to attribute X as quantity consumed increases. The highest ratio is offered by product A, the lowest by product C.

The Budget Constraint and Efficiency Frontier

The quantity of a product that a consumer can afford to buy will depend on the price of the product and the budget of the consumer. For a given budget constraint and set of prices for the products, the end points A, B, and C represent the limits of consumption along each product ray. The end points (or total attainable attributes) are calculated by multiplying each attribute level for a single unit of the product by the quantity of the product that the consumer can afford to buy. The line segment ABC defines the budget (or efficiency) frontier for the consumer. The frontier is called efficient because a consumer maximizes utility by choosing combinations of attributes on this frontier.

The efficiency frontier shows the total amounts of attribute X and attribute Y that the consumer can afford to buy. Product A offers XA of attribute X and YA of attribute Y. Similarly, products B and C offer the attribute bundles (XB, YB) and (Xc, Yc) respectively.

Indifference Curves

The consumer's choice is made by maximizing her/his utility, as defined by the consumer's set of indifference curves, subject to the budget constraint. In the product attributes model, we interpret the slope of the indifference curve at a particular point (also called the marginal rate of substitution) as the rate at which the consumer is willing to trade off units of attribute Y for an additional unit of attribute X to maintain the same level of utility. Thus, the consumer's choice is influenced by her/his preference pattern in attribute space.

Consumer Equilibrium

As shown in Figure 2, the consumer shows a strong preference for attribute Y and therefore chooses product A. The consumer's preference is represented by the relatively flat indifference curves (IC, IC*, IC). The highest indifference curve that intersects the efficiency frontier is IC*. (We will always let IC* represent the optimal indifference curve throughout our discussion of the product attributes model.) This intersection is at the end point of the ray for product A. The consumer maximizes utility at this intersection and will choose Product A.

On the other hand, a strong preference for attribute X would lead her/him to choose product C. In this case, the indifference curves are relatively steep. The proliferation of differentiated products in a particular market can be explained as the result of the dispersion of tastes and preferences for various attributes among different segments of consumers.

FIGURE 2

X Attribute X

Mixing Products

Note that the consumer will spend her/his entire budget on a single product A, B, or C if the highest indifference curve just touches the respective end point. If the highest indifference curve touches a point on the line segment between two product ray end points, then the consumer would choose to split consumption between the two adjacent products. In Figure 3, the highest indifference curve that intersects the efficiency frontier is IC*. The intersection is represented on the graph as point M. However, there is no product that satisfies the attributes in the ratio represented by the point M. The consumer may be able to maximize utility by mixing products B and C. The consumer buys product B some times and product C other times. This can lead to brand switching by consumers for specific uses or occasions or just for variety. An example of mixing products would be dining out at various restaurants.

Product Indivisibility

If the product's consumption is indivisible (as in the case of an automobile or house), then consumption (or the consumer's




FIGURE 3

I ¡

XA Attribute X

choice) is determined by the highest indifference curve that touches an end point. This may occur if the price of the product is large relative to the consumer's income, and the consumer can only afford to purchase one of the products.

As shown in Figure 4, the consumer cannot attain the highest indifference curve IC* if she/he can only afford to buy one of the products. The consumer will choose to buy product B (where the lower indifference curve IC intersects the efficiency frontier), and then utility will be less than optimal.

FIGURE 4

X Attribute X

Lancaster's Model in the Marketing Literature

In the 1970s and early 80s, several papers looked at Lancaster's model, its advantages and limitations, and the potential application to marketing. Ratchford (1975) provides

an overview of the model and describes several contributions of Lancaster's model to consumer behavior (p. 67):

• An explanation for the role of price in determining demand for differentiated products.

• A framework for estimating the sensitivity of demand to changes in the relative price of a brand.

• A theoretical perspective for brand share models. • An economic explanation for brand loyalty. • A method to understand how brand preferences are formed

given preferences for characteristics and prices.

Ratchford (1975) cites two limitations of the model. First, the model is static and deterministic. As noted by Ratchford (1975), most of the models used for evaluating brand positioning possess these same characteristics so this is a limitation across many of the methods used, not just Lancaster's model. Ratchford (1979) in fact provides an extension of the Lancaster model for stochastic choice. Second, the model does not address how preferences for the characteristics are formed. Again, most of the methods for evaluating brand positioning do not describe this relationship. To the extent that understanding the formation of preferences is important for a specific application, Hauser and Sirnmie (1981) extend Lancaster's model to include a full information processing model (physical feature -* perceptions -► preferences).

Ladd and Zober (1977) claim the model makes three assumptions that limit its usefulness: (1) every characteristic has nonnegative marginal utility, (2) utility is independent of the distribution of characteristics among products, and (3) the model uses linear consumption technology. Ratchford (1979) refutes that these are necessary assumptions of the Lancaster model except in unusual circumstances (e.g., when a consumer buys all brands of the product).

For our proposed application to brand positioning as opposed to product choice, Lancaster's model ( 1 966, 1 979) provides an effective method for incorporating price into positioning analysis without requiring more elaborate extensions to the model. The product attributes model is particularly effective for brand positioning decisions involving product design and repositioning, especially when price plays a critical role in the perception of value for the product. In this context, the product attributes model provides a superior way to look at the combination of attributes with price and consumer budget constraints to guide managers in decision making.

Ratchford ( 1 975) concludes: "In defense of the work presented here [Lancaster's model], it can only be said that no worthwhile model of consumer behavior could take account of all factors.... The judgment about whether the models discussed here are 'good' can only be made after then- predictions have been tested empirically (p. 74)." To encourage this empirical testing, we propose several ways that the product attributes model can provide strategic guidance in marketing. We conclude with an example of an application in

Spring 2003 35



the automobile industry. Our goal is to show the usefulness of the model in a specific context - brand positioning - and make the model tractable for academician and practitioner alike.

STRATEGIC IMPLICATIONS FOR BRAND POSITIONING

The product attributes model can be used to analyze a brand's positioning and its effect on consumer choice. The product attributes model has important implications for the brand's positioning and marketing mix.

Market Segmentation/Target Market

The product attributes model can help the firm understand the market segment(s) for a product. It can then enable the firm to identify the appropriate target market(s) for its specific brand.

For example, in Figure 5, the consumer with indifference curve ICY has a relatively low marginal rate of substitution (MRS) between attribute Y and attribute X. This consumer has a relatively flat indifference curve and will tend to purchase product A. Consumers with a low MRS can be called the "Attribute Y segment."

The consumer with indifference curve ICX has a relatively high MRS between attribute Y and attribute X. This consumer has a much steeper indifference curve and will tend to purchase product C. She/he belongs to the "Attribute X segment."

FIGURE 5

/i^vT / ICy

X Attribute X

A firm that knows the tastes and preferences of consumers for the various attributes can incorporate the attributes into a product designed for a particular segment of the consumer base. The firm may also look for market segments) that are not satisfied with the current market offerings and develop a product to appeal to this segment.

Positioning

The product attributes model allows the firm to see how its brand is perceived relative to the competition on critical attributes. The firm can assess whether it operates from a position of strength or if it is vulnerable to the competition. If it is vulnerable, it can determine the appropriate action to take: a new product, changes in the existing product, changes in price, or a new promotional strategy.

New Product

A firm may introduce a new product positioned to take advantage o f an opportunity represented by a "gap" in the attribute space between rays. This "gap" indicates that existing products do not satisfy a specific ratio of attributes. If a sufficient number of consumers want this ratio, a new product may be developed with the appropriate ratio of attributes that maximizes the utility of this market segment.

Changes in Existing Product

The product attributes model indicates on what attribute(s) the existing product may be deficient. The firm can then make the product improvements needed to shift the product to a more favorable position in the market. The product attributes model can also help predict increases in market share that will result from these product improvements.

Changes in Price

If the product's price is reduced, the budget constraint will shift outward along the product ray (the consumer can afford more). If the product's price is increased, the budget constraint will shift inward along the product ray (the consumer can afford less).

Look again at Figure 4. What can the firm that sells product C do to capture the consumer and make a sale? One option would be to cut price. Cutting the price will shift the budget constraint outward. A cut in price moves the end point for Product C from C to C as shown in Figure 6. The consumer equilibrium w ill b e a 1 1 he i ntersection of I C* a nd t he n ew efficiency frontier. The consumer then will choose to buy product C.

Pricing a Product Out of the Market

Based upon a consumer's perception of the attributes offered by a product, there is a maximum price he or she will pay for the product. In Figure 7, the consumer maximizes utility by mixing between products A and C. The price of product B at the point B on the product ray is too high for the consumer. The price of product B must be decreased to the point represented by B' in order to capture the consumer. An alternative approach is to add value through product re-design to move the product to the same position.




FIGURE 6

>

! •e

/ / ^ - IC

Attribute X

FIGURE 7

2

v Attribute X AA

Promotion / Changes in Consumer Perceptions

Advertising can change the perception of the product in terms of attribute content and proportion (distance from the origin to the e nd p oint a nd/or s lope o f t he p roduct r ay). 1 1 c an a lso influence a consumer's tastes for attributes (shape of individual consumer's indifference curve).

Advertising may make an attribute important to consumers that might not have been considered previously. In this case, a new attribute dimension must be considered in the product attributes model.

Price deals and coupons can be used to effectively lower the price for specific purchase occasions. These promotions adjust the efficiency frontier on the promoted product for those particular purchases.

An Example

We will start with a simple example based on two automobile attributes (gas mileage and horsepower) to analyze the choices faced by two market segments: environmentalists and racing enthusiasts. Table 2 summarizes gas mileage, horsepower, and price data for three automobiles (Focus SE, Taurus, and Mustang Mach 1), as reported by Edmund's New Car Guide. Note that these models do not compete in the same subcategory. However, the analysis shows the positioning strategies used by one company (in this case, Ford) to appeal to multiple segments based on different combinations of attributes.

Figure 8 shows the product rays associated with each brand of automobile. The slope of each ray is equal to the ratio of gas mileage to horsepower.

FIGURE 8

& 25°- J . Focus

1 200- / 3 /

150 - I Taurus

100« /

/ / Mustang 50 ' // y^^

0 F I ■ I I I 0 50 100 150 200 250

Horsepower

Table 3 shows the total attainable attributes for each brand of automobile given the consumer's budget. Total attainable attributes are calculated as the attribute rating of the automobile multiplied by the quantity of that automobile which the consumer can afford to buy (in this case, automobiles per $30,000). Clearly, the Focus is positioned as an e conomy c ar b ased o n i ts superior p erformance o n g as mileage. Similarly, Mustang has a well-defined position as a muscle car based on its strength in terms of horsepower. The Taurus performs at a mid-level on each of these attributes.

Spring 2003 37



TABLE 2 ATTRIBUTES AND PRICES OF THREE AUTOMOBILES

- GAS MILEAGE AND HORSEPOWER -

Attribute Rating* Ratio of Gas Automobiles per Brand of Price 1 Mileage to $30,000** Automobile (MSRP) Gas Mileage Horsepower Horsepower

(Highway)

Focus SE $15,595 100 36 2.78 1.92

Taurus $19,990 74 51 1.45 1.50

Mustang Machi $28,995 63 100 0.63 1.03

TABLE 3

TOTAL ATTAINABLE ATTRIBUTES - GAS MILEAGE AND HORSEPOWER -

Brand of Automobile Total Attainable Attributes

Gas Mileage Horsepower

Focus SE 192 69

Taurus 111 77

Mustang Machi 65 103

Figure 9 builds on Figure 8. The total attainable attributes define the end point on each product ray. We then connect the outermost end points to define the efficiency frontier. Any end points inside the efficiency frontier are effectively priced out of the market. In this case, the end points are defined by the Focus and the Mustang. The Taurus falls inside of this efficiency frontier, and so it appears to be priced out of the market. We will discuss this more later in the example.

FIGURE 9

& 250-

I /Focus j- 150- x / ' 150- / ' Taurus -

/ M / / / W Mustang

0 50 100 150 200 250

Horsepower

People who are concerned about the environment or economy would be expected to give up only a little gas mileage for increases in horsepower since this segment values fuel

efficiency. As a result, this segment has a very low marginal rate of substitution, and its indifference curves are relatively flat. We call this segment the environmentalist segment with the highest attainable indifference curve ICE. Figure 10 shows a representation of this segment's highest attainable indifference curve superimposed on the graph of Figure 9.

The environmentalist segment will choose to purchase the Focus SE since its highest indifference curve (ICg) intersects the efficiency frontier at the end point for the Focus SE. This segment's willingness to trade off only a little bit of gas mileage for m ore h orsepower i s b est s atisfied b y t he F ord Focus SE.

On the other hand, the other segment (referred to as the racing enthusiasts) cares little about gas mileage and wants a lot of horsepower. As a result, this segment's indifference curve (ICr) is relatively steep in the area of the efficiency frontier, and a member of this segment will choose the Mustang Mach 1.

Note that the Ford Taurus is "priced out of the market." This result is driven by our assumption that consumers only care about two automobile attributes, gas mileage and horsepower. We know that there are consumers who buy a Ford Taurus (in fact, the Taurus is one of the best selling automobiles in the world). This must mean that there are attributes other than gas mileage and horsepower that are important to automobile consumers. This result points to a need to consider additional attributes in the product attributes model. The analysis can be extended to multiple attributes through additional maps or




FIGURE 10

gj 250-1 '

± ' . Focus

^ 200- 'y ü /'^

ICe 150 "

/ '/Taurus

íoo« / 7 V / / 's Mustang

^

//^ ^ ICr

0 F I I I I I

0 50 100 150 200 250

Horsepower

dimensions. Alternatively, the analysis can be extended through multiattribute modeling of utility maximization. These analyses are demonstrated in a continuation of our example below.

Why is the Taurus so popular? The Taurus is marketed by Ford as "America's smart family car" on its website. The specifications show that the car is longer and wider than either the Focus or Mustang, plus the Taurus can seat 6, while the Focus can seat 5, and the Mustang only 4. However, the greatest difference in the cars comes in the luggage capacity where the Taurus has significantly more room in the trunk to haul around sporting gear for the children, groceries for the household, and of course luggage for a family vacation. As a result, we have used luggage capacity to extend our analysis to a third dimension that helps us understand the popularity of the Taurus.

Table 4 builds on Table 2, presenting the ratings and ratios for luggage capacity. The total attainable attributes for luggage capacity are summarized in Table 5.

Table 4 Attributes and Ratios of Three Automobiles

- Luggage Capacity - Attribute Ratings*

Brand of Ratio of Ratio of Automobile Luggage Gas

Capacity to Mileage to Horsepower Luggage

Capacity

Luggage Capacity Focus SE 76 2JJ 1.32 Taurus 100 L96 0.74 Mustang

Mach 1 1 64

1 0.64

1 0.98

♦Attribute ratings have been scaled to 100 points to allow an equivalent comparison of the attributes. Actual luggage capacity: Focus SE 12.9 cubic feet, Taurus 17 cubic feet, and Mustang Mach 1 10.9 cubic feet.

As can be seen in Table 5, the Taurus has the greatest total attainable attributes for luggage capacity; however, the Focus SE is very close. This proximity is primarily driven by the difference in price since the luggage capacity for the Taurus is approximately 40% larger than that for the Focus SE. This finding demonstrates the importance of including price in the analysis.

TABLE 5 Total Attainable Attributes

- Luggage Capacity - Total Attainable Attributes

Brand of Automobile Luggage Capacity Focus SE 146 Taurus 150 Mustang Mach 1 66

Figure 1 1 shows the relative positioning of the Focus, Taurus, and Mustang with respect to luggage capacity and horsepower. For segments emphasizing luggage capacity or horsepower, the Focus does not provide the value offered by Taurus and Mustang. As a result, the Focus, despite its relatively modest price, is priced out of the market because it does not possess sufficient quantities of the attributes desired for the price. Of the two cars remaining, the Taurus clearly appeals to the segment that desires luggage capacity which we will call the family segment. The indifference curve for this segment is designated by ICF. The racing enthusiasts who greatly value horsepower over luggage capacity will select the Mustang (as shown by ICr).

For completeness, we also show the comparison of luggage capacity and gas mileage. In this case, the Mustang is priced out of the market since it does not offer greater value for either attribute. The family segment will select the Taurus for its greater luggage capacity (as shown by ICF), and the environmentalists will choose the Focus (as shown by ICE) for its greater fuel efficiency.

FIGURE 11

£> 20°l ' /Taurus

õ 150 • ■ "VtK^

& y/'~"~

ICf

J 100- // '

// '/ Mustang

Orí i i i 0 50 100 150 200

Horsepower

Spring 2003 39



FIGURE 12

& 250 i

l 200- VfocuS

J yt^ IC*

150- / ' y Taurus

Mustang v/ 's

50- //C'

Orí i i i i

0 50 100 150 200 250

Luggage Capacity

Multiattribute modeling of utility maximization requires an understanding of the willingness of consumers to make tradeoffs among the three attributes. To estimate consumer utility, m arketers c an u se s urveys ore onjoint a nalysis. F or purposes of our example, we assume a linear utility function which means that consumers have a constant marginal rate of substitution. In other words, we assume that the consumers' willingness to tradeoff among the attributes does not vary across the levels of the attributes. This assumption seems valid at the brand level since in any given purchasing situation, consumers will most likely have one preference based on the attributes, price and their budget. For our three attribute example, the utility function takes the form of: Uj = ax + by +cz (5)

In this example, U¡ is the utility of the individual (or a segment of individuals); a, b, c are the weights for each product attribute; and x,y, z are the total attainable attributes for each product attribute.

As can be seen in this generalized example, the utility function is a weighted combination of the three attributes. The weights are assumed to vary across segments (and in fact, define the segments). For example, the environmentalists value gas mileage above all other variables. That does not mean that they do not want horsepower or luggage capacity, just that the weights for these attributes will be much lower. We will model this utility as:

UE=10GM + HP + LC (6)

In this equation, UE stands for the utility function for the environmentalists, GM for gas mileage, HP for horsepower, and LC for luggage capacity. Ten is the weight (or importance) assigned to gas mileage while the relative weights for horsepower and luggage capacity are one for this segment. In other words, this segment values gas mileage at the rate of 10 to 1 over any of the other attributes.

Similarly, for expositional purposes, we develop utilities for the family and racing enthusiast segments. The family segment (UF) values luggage capacity over horsepower, and horsepower over gas mileage. The racing enthusiasts (Unjust care about horsepower and are willing to pay for it.

UF = GM+10HP + 20LC (7)

UR = GM + 20HP + LC (8)

Based on these utility functions, Table 6 shows the total utility assigned to each car by each segment. These values are assigned by weighting the total attainable attributes for each car. The bold values show the car preferred by each segment.

TABLE 6 Total Utility for Three Consumer Segments Across Three Brands of Automobiles Up Up Up Focus SE 2135 3802 1718 Taurus 1337 3882 1801

"Mustang

Mach 1 1819 12415 12191

As can be seen in these examples, the environmentalist segment has a strong preference for the Focus SE due to its superior gas mileage. The family segment prefers the Taurus for its greater luggage capacity, and the racing enthusiasts prefer the Mustang for its horsepower. Through this analysis, marketers can identify the key segments and their preferences in terms of attributes and brands to determine the best segment to target, opportunities for new products or the need for modifications to existing products, appropriate positioning and promotional strategies, and potential pricing adjustments.

Conclusion and Directions for Future Research

The product attributes model provides a powerful tool for analyzing marketing strategy based on the positioning of a brand. It provides a framework for incorporating brand price, consumer preferences, and budget constraints into an analysis of the optimal combination of attributes that appeal to specific market segments. The model provides direction for marketing strategy with implications for pricing, product development, promotion, positioning, and segmentation.

The product attributes model offers many opportunities for a future research agenda, including:

• Empirically testing the application of the model to marketing strategy.

• Using empirical research to compare the results of the product attributes model to other positioning analysis tools to better understand the relative strengths and weaknesses of these methods.

• Developing measures to define indifference curves for consumers or consumer segments.

• Empirically evaluating the application of the multiattribute model and its implications for marketing strategy beyond two dimensions.




REFERENCES

Aaker, David A. (1998), Strategic Market Management, Fifth edition, New York: John Wiley & Sons, Inc.

Aaker, David A. (1996), Building Strong Brands, New York: The Free Press.

Bijmolt, Tammo H.A. and Michel Wedel (1999), "A Comparison of Multidimensional Scaling Methods for Perceptual Mapping," Journal of Marketing Research, 36 (May), 277-285.

Carroll, J. Douglas and Paul E. Green (1995), "Psychometric Methods in Marketing Research: Part I, Conjoint Analysis," Journal of Marketing Research, 32 (November), 385-391.

and (1997), "Psychometric Methods in Marketing Research: Part II, Multidimensional Scaling," Journal of Marketing Research, 34 (May), 193-204.

DeSarbo, Wayne and Vithala R. Rao (1986), "A Constrained Unfolding Methodology for Product Positioning," Marketing Science, 5, 1 (Winter), 1-19.

DeSarbo, Wayne S., Martin R. Young, and Arvind Rangaswamy (1997), "A Parametric Multidimensional Unfolding Procedure for Incomplete Nonmetric Preference/Choice Set Data in Marketing Research," Journal of Marketing Research, 34 (November), 499-516.

Dillon, William R., Teresa Domzal, and Thomas J. Madden (1986), "Evaluating Alternative Product Positioning Strategies," Journal oj Advertising Research, 26 (4), 29-36.

Douglas, Evan J. (1992), Managerial Economics: Analysis and Strategy, 4th Edition, Englewood Cliffs, New Jersey: Prentice Hall.

Green, Paul E., J.D. Carroll, and S.M. Goldberg (1981), "A General Approach to Product Design Optimization via Conjoint Analysis," Journal of Marketing, 45 (Summer), 17-37.

and W. S. DeSarbo (1978), "Additive Decomposition of Perceptions' Data Via Conjoint Analysis," Journal of Consumer Research, 5 (June), 58-65.

and Abba M. Krieger (1992), "An Application of a Product Positioning Model to Pharmaceutical Products," Marketing Science, 11, 2 (Spring), 117-132.

, Y. Wind, and H.J. Claycamp (1975), "Brand- Features Congruence Mapping," Journal of Marketing Research, 12 (August), 306-313.

, , and A. K. Jain (1972), "A Note on Measurement of Social-Psychological Belief Systems," Journal of Marketing Research, 9 (May), 204-208.

Hauser, John R. and Steven P. Gaskin (1984), "Application of the 'Defender' Consumer Model," Marketing Science, 3, 4 (Fall), 327-351.

and Koppelman (1979), "Alternative Perceptual Mapping Techniques: Relative Accuracy and Usefulness," Journal of Marketing Research, 1 (November), 495-506.

and P. Simmie (1981), "Profit Maximizing Perceptual Positions: An Integrated Theory for the Selection of Product Features and Price," Management Science, 27 (January), 33-56.

and G. Urban ( 1 977), "A Normative Methodology for Modeling Consumer Response to Innovations," Operations Research, 25 (July-August), 579-619.

and (1986), "The Value Priority Hypotheses for Consumer Budget Plans," Journal of Consumer Research, 12 (March), 446-462.

Hendler, Reuven (1975), "Lancaster's New Approach to Consumer Demand and Its Limitations," The A merican Economic Review, 65, 1 (March), 194-199.

Huber, J. andM.B. Holbrook(1979), "Using Attribute Ratings for Product Positioning: Some Distinctions Among Compositional Approaches," Journal of Marketing Research, 16 (November), 507-515.

Katahira, Hotaka ( 1 990), "Perceptual Mapping Using Ordered Logit Analysis," Marketing Science, 9, 1 (Winter), 1-17.

Keon, John W. (1983), "Product Positioning: TRINODAL Mapping of Brand Images, Ad Images, and Consumer Preference," Journal of Marketing Research, 20 (November), 380-392.

Ladd, George W. and Martin Zober (1977), "Model of Consumer Reaction to Product Characteristics," Journal of Consumer Research, 4 (September), 89-101.

Lancaster, Kelvin J. (1966), "A New Approach to Consumer Theory," Journal of Political Economy, 74, 132-157.

(1971), Variety, Equity and Efficiency, New York: Columbia University Press.

(1975), "Socially Optimal Product Differentiation," American Economic Review (September), 567-585.

Spring 2003 41



(1979), Consumer Demand: A New Approach, New York: Columbia University Press.

Monpour, R., J.M. McCullough, and D.L. MacLachlan ( 1 976), "Time Changes in Perception: A Longitudinal Application of Multidimensional Scaling," Journal of Marketing Research, 13 (August), 245-253.

Neal, William D. (1980), "Strategic Product Positioning: A Step-by-Step Guide, Business, 30 (3).

Perreault, William D., Jr. and E. Jerome McCarthy (1999), Basic Marketing: A Global-Managerial Approach, Boston, MA: Irwin McGraw-Hill.

Pessemier, E.A. (1977), Product Management Strategy and Organization, New York: Wiley/Hamilton, 234-41.

Ratchford, Brian T. (1975), "The New Economic Theory of Consumer Behavior: An Interpretive Essay," The Journal of Consumer Research, 2, 2 (September), 65-75.

(1979), "Operationalizing Economic Models of Demand for Product Characteristics," Journal of Consumer Research, 6 (June), 76-85.

Schnedler, David E. (1996), "Use strategic market models to predict customer behavior," Sloan Management Review, 37 (3), 85-94.

Seggev, Eli ( 1 992), "Point of View: Advertising Effectiveness Measurement for Contribution-Based Compensation," Journal of Advertising Research, 32 (4), 73-80.

Taylor, Lester D. and George H. Haines, Jr. (1975), "Commentaries on Ratchford, 'The New Economic Theory of Consumer Behavior: An Interpretive Essay,'" The Journal of Consumer Research, 2, 2 (September), 76-78.

Urban, Glen L. and John R. Hauser (1993), Design and Marketing of New Products, Second Edition, Englewood Cliffs, NJ: Prentice Hall, 213-216.

Wilkie, W.L. and E.A. Pessemier (1973), "Issues in Marketing's Use of Multi-Attribute Models," Journal of Marketing Research, 10 (November), 428-441.

AUTHOR BIOGRAPHY Carol F. Gwin (Ph.D., University of South Carolina) is a lecturer of marketing in the Hankamer School of Business at Baylor University. Dr. Gwin's areas of research are marketing strategy and consumer policy. In the area of marketing strategy, the research stream encompasses strategic implications for decisions in sales management, promotion, and e- commerce. For consumer policy, her research includes privacy issues and compulsive buying. Dr. Gwin has presented at the Annual Conferences for AMA and ACR and has a publication in the Journal of Real Estate Research.

AUTHOR BIOGRAPHY Carl R. Gwin (Ph.D., Indiana University) is an assistant professor of economics in the Hankamer School of Business at Baylor University. Dr. Gwin's areas of research interest are Industrial Organization and Applied Microeconomics. His stream of research emphasizes the application and testing of game theoretic principles in understanding how business responds to asymmetric information problems. Dr. Gwin's publications include Economic Inquiry, the Journal of Property Investment and Finance, the Journal of Real Estate Research, and the Antitrust Bulletin. He has an upcoming publication in the Journal of Money, Credit and Banking.




product attributes model: a tool for evaluating brand positioning

Documents