Journal article

The image or perception one has of a product or service drives the individual’s buying decision and, thus, drives that product’s or service’s market share. Image affects all kinds of decisions such as which university to attend, which automobile to purchase, which political candidate to support, and even which motel to choose. That image is important to a product’s or service’s success is well understood. What is not well understood is how image data might be used to optimize a product’s or service’s image to maximize market share. While many sophisticated models of the buying decision process have been developed to do just that (e.g., Weerahandi & Moitra, 1995; Andrews & Srinivasan, 1995), no evidence is apparent that such models are being used by marketing practitioners to increase market share. Given the complexity of most models, it may not be apparent how a marketing practitioner might use them to identify an image that would optimize a brand’s share.

This paper was designed to identify very straightforwardly (1) how meaningfully to assess the image a product or service conveys to the marketplace and (2) how to use a model of the brand choice process to identify those changes to that product’s or service’s image that could be expected to maximize market share.

Upper-divisional college students in a large marketing class (219 students) completed a paper-and-pencil questionnaire about the fast-food industry. Specifically, they were asked to indicate the relative importance of 15 attributes of fast-food restaurants by distributing a total of 15 points among them, an approach previously used by Schori (1995). For example, if a student thought all 15 attributes were of equal importance, then he could assign one point to each attribute or if he thought that one attribute far out-weighed the remaining 14 in importance, then he could assign all 15 points to that single attribute. The students were also asked to rate Burger King , McDonald’s , Wendy’s , and their preferred fast-food restaurant, if it was other than one of these three, on the 15 attributes by assigning a score anchored by 1 ("strongly agree") and 5 ("strongly disagree") on each scale. In addition, they were asked to indicate how high they would rate their ideal fast-food restaurant on the 15 attributes using the same 5-point scale. The 15 attributes used were low prices, ample parking, fried food, burgers-only, 24-hour service, kid-oriented, clean/safe facility, trained staff, courteous staff, can be found everywhere, broad menu, low-calorie food, drive-through discount, near major highway, and food low in cholesterol.

Had this study actually be done for one of the fast-food companies being evaluated, we would have selected participants who were representative of that company’s entire target market, not just college students at a single institution. Likewise, we would have scoured that company’s previous research for attributes that had repeatedly been identified by customers in the past.

Actual share of fast-food restaurant preference among the 219 respondents was Burger King at 11%, McDonald’s 51%, Wendy’s 21%, and other fast-food restaurants 17%.

To evaluate importance we simply looked at the proportion of respondents who assigned more than 1 point to a given attribute, i.e., gave the attribute more than its "fair share" of points. The relative importance the 219 respondents ascribed to the various attributes is shown in Table 1. Importance ratings are really spread out among these 15 attributes, viz., by a factor of 5. Had we access to attributes that had probably come up in the fast-food industry’s own research, the spread in relative importance might have been greater.

Attribute	%	Attribute	%
24-hour Service	14	Food Low in Cholesterol	34
Drive-through Discount	17	Courteous Staff	37
Fried Food	17	Clean/Safe Facility	41
Ample Parking	22	Can Be Found Everywhere	49
Near Major Highway	26	Burgers-only	50
Low Prices	26	Trained Staff	56
Broad Menu	27	Kid-oriented	70
Low Calorie Food	29

Some years ago, we described an approach that permits one to identify a position (image) that would maximize a brand’s market share (Schori & Meadow, 1985). We developed a model of the process of brand choice that permitted prediction of an individual’s choice of brands, given the importance the person ascribed to a set of attributes that depicted the product/service category, how well the rater believed alternative brands scored on those attributes, and how the person’s ideal brand scored on those attributes, i.e., exactly the questions we asked of the respondents in this study. In simple terms, the model looks like this:

Share of Preference_k= S Minimum [Over-all Assessment_k/ (Total Individuals) * 100],

Since the model permits prediction of brand choice (actually, preferred brand), we were able to simulate changed beliefs about a brand, i.e., fast-food restaurant, and observe what those changed beliefs would do to each respondent’s predicted brand choice. This was done from the point-of-view of each of these brands by examining the potential effect on market share of changing beliefs on just one attribute at a time. These simulated belief changes were modest in that they moved each respondent’s rating only one point up or one point down (if not already at the maximum or minimum of the 5-point scale).

As Table 2 illustrates, if Burger King were able to change the market’s belief about them on just one attribute¾ located everywhere¾ their share could double! This gain, incidentally, would be mostly at McDonald’s expense. Image changes would not hurt them much, since in the worse case (with kid-oriented), they’d stand to lose only one share point. From the table, it is apparent that there is no upside to McDonald’s in changing beliefs about their brand¾ only a downside. Changing the market’s beliefs about the extent that "burgers-only" characterizes them would not damage their share but it would not help it either. On the other hand, that change by McDonald’s would benefit Burger King at the expense of Wendy’s . Clearly, image is dynamic. The damage McDonald’s could do to themselves by changing their own image, i.e., making themselves appear less kid-oriented, could lead to the loss of 18 share points¾ within the group of students surveyed. Given these findings, they obviously would be well-advised not to change their image. One of their competitors, however, might think of how they might change McDonald’s image in the marketplace to take some of McDonald’s share.

TABLE 2
PREDICTED MAXIMUM AND MINIMUM MARKET SHARE BASED UPON SIMULATED CHANGES IN BELIEFS ABOUT A BRAND VERSUS ACTUAL SHARE OF PREFERENCE (N=219)

	Burger King	McDonald’s	Wendy’s
Actual share of preference	11	51	21
Predicted share of preference
for Burger King
If Everywhere [more]	23	41	19
If Kid-oriented [less]	10	52	21
for McDonald’s
If Burger-only [more]	13	51	19
If Kid-oriented [less]	24	33	26
for Wendy’s
If Trained staff [more]	11	46	26
If Burgers-only [less]	17	49	18

Table 2 also shows Wendy’s has the potential for a five share-point gain were they rated more favorably by the market on "trained staff"¾ and all at the expense of McDonald’s . Were they rated less strongly as "burgers only," they would stand to lose four share points.

In exercising the model of the process of brand choice, we looked at the effect on share of changing beliefs on each and every attribute, one attribute at a time for every single respondent and for each restaurant. These results go beyond simply directing attention toward what constitutes the most important attributes to these consumers. They also provide an indication of the extent to which changes in beliefs about a brand might increase or decrease market share. The ability to view these data from the perspective of any given brand is especially useful.

Assuming that we’ve considered the "right" attribute set and that we’ve addressed the "right" target market, there are some conclusions a marketing practitioner might draw: Burger King should open more stores. Doing so has the potential of helping them double their share¾ most of which would come at the expense of McDonald’s . Knowing this, McDonald’s should do what they can to discourage Burger King from opening more stores. McDonald’s should do everything possible to avoid changing their image. Almost any changes in beliefs about them would have a negative effect on their share¾ especially anything that would result in their being viewed as less kid-oriented. Knowing this, Burger King and Wendy’s may want to consider how they might change McDonald’s position on the kid-oriented attribute. Wendy’s should address how they might improve the market’s perception of how well their staff is trained. If successful, McDonald’s would be the prime contributor to Wendy’s share increase.

This effort was designed not to explore the fast-food industry but to serve as an example of how changing a brand’s image may be done in a way that would be predicted to maximize the brand’s share. Indeed, any industry could have been selected to illustrate the approach. The respondents were, in all likelihood, not representative of what the various fast-food restaurants would consider target markets. Likewise, the attributes considered probably fall short of what might be considered by these restaurants as appropriate; however, this exercise showed how an image can be used to describe which change in a brand’s image might maximize share¾ something not commonly seen in studies of brand image.

Andrews, R.L., & Srinivsan, T.C. (1995) Studying consideration effects in empirical choice models using scanner panel data. Journal of Marketing Research, 32, 30-41.

Schori, T.R. (1995) Note on attribute importance. Perceptual and Motor Skills, 80, 1129-1130.

Schori, T.R. , & Meadow, H.L. (1985) Brand choice modeling; identifying a brand’s optimal positioning, Psychological Report, 57, 1260-1262.

Weerahandi, S., & Moitra, S. (1995) Using survey data to predict adoption and switching for services, Journal of Marketing Research, 32, 85-96.

GETTING THE MOST OUT OF IMAGE: AN EXAMPLE FROM THE FAST-FOOD INDUSTRY