Prioritizing Marketing Image Goals under Resource Constraints
Image assessment and enhancement are essential strategic management tools. Managers must be able to assess their company’s current image or reputation in the marketplace and improve it. Although it would be ideal to achieve top ratings on every attribute, such as product quality, after-sales service, and reliability, it is not realistic. Given limited resources, managers must decide which of the attributes are more important to the company’s target market and consider how difficult it would be to enhance the image on those attributes. In this paper, we propose an approach for prioritizing marketing image goals by taking into account market preferences and resource requirements. The approach is potentially applicable to small, medium, and large companies, including not-for-profit organizations. In the first half of the paper, we explain the method; in the second half, we describe how a company applied the method.
Marketing Image
The impressions, beliefs, and feelings that people have about a company constitute the firm’s image.1 These impressions may be true or false, real or imagined. If people hold incorrect negative perceptions about the company or its products, then management needs to communicate with them in order to change the incorrect perceptions. On the other hand, if the negative perceptions are accurate, then the company needs to address the actual problems.
A company has different types of images: corporate images, product images, and brand images. We focus here on marketing image, the way customers view the company’s overall marketing offer and marketing mix.2 The marketing image is broader than brand and product line images but narrower than corporate image, which also considers the company as an employer and as a citizen in terms of community involvement, concern for the environment, patronage of the arts, and the like. In the case of a large diversified corporation, marketing images will typically be different for different business units. Consequently, when we use the word “company” in this paper, we mean a corporation’s business unit.
Image is an integral part of any company. The image of a company often determines its success with various constituencies.3 Marketing image plays an important role in the attractiveness of the firm and its products and services to customers and thereby has an impact on the bottom line.
An Approach to Prioritizing Marketing Image Goals
Company image is typically measured with rating scales, such as determining average customer ratings on attributes such as after-sales service, reliability, and quality. We propose an approach that prioritizes marketing image goals by considering both the relative importance the market places on the different attributes and the resources it would take to make improvements. The idea is to improve those attributes that provide the maximum benefit per unit cost — to maximize the “bang for the buck.” We measure the benefit of an improvement from a customer point-of-view by using conjoint analysis methodology.4 We rely on managerial judgment to determine the resource requirements for making improvements. The steps for the method are as follows:
1. Determine the Sample of Customers
The company’s overall marketing strategy dictates its target market. For instance, a department store positioned to serve an upscale market would not want to include in its sample those customers who shop mostly in discount outlets. The market research study should draw a random sample from the target market, including both current and potential customers, for use in the following research.
2. Select the Attributes
To determine which attributes are salient to the study, the company can conduct focus groups and one-on-one interviews. During this qualitative research phase, managers should attempt to understand the language customers use in describing perceptions of a company and its competitors. As the purpose is to understand the marketing image of the company rather than the image of a specific brand, the attributes will tend to be general and subjective. For instance, in the case of an automobile company, a marketing image attribute would be engineering excellence rather than a particular car’s speed of acceleration.
3. Collect Quantitative Data
For the quantitative phase, the company can conduct a survey (by mail, phone, in-person interview, or a combination). The survey asks customers about their familiarity and image ratings of the company and its key competitors. It also collects information on demographics and purchase behavior.
4. Apply Conjoint Analysis
Next, conjoint analysis is used to assess the importance customers place on the attributes.5 Conjoint analysis is a procedure for inferring a customer’s priorities or importances for the attributes. In the full-profile approach, researchers describe hypothetical companies in terms of how they perform on the multiple attributes. Given the subjective nature of attributes, the levels of attributes are described by labels such as poor, fair, good, and excellent. Customers rank or rate the hypothetical companies in terms of their overall appeal.
In the empirical application presented later, preliminary research indicated that most customers perceived the company as at least fair on all the chosen attributes. Consequently, we consider only the fair, good, and excellent attribute levels in our discussion.
Instead of the full-profile approach, customers may choose to compare two attributes at a time or use tradeoff tables. The full-profile approach is particularly appropriate in the context of a marketing image study because it is possible to describe the company simultaneously in terms of all the chosen attributes. Because conjoint analysis methodology is well known, we do not go into its details here.6 Although it has been used extensively to estimate the impact of attributes on customer preferences for products and services, its application in the context of prioritizing marketing image goals is new, to the best of our knowledge.7
An important output of conjoint analysis is the average (across the customer sample) part-worth function for each attribute. The part-worth function specifies the utility or value the customer places on specific levels of that attribute. The overall utility is given by summing the part-worths across all attributes.
Figure 1 provides a method for interpolating the part-worth utility for attribute values throughout the entire range. For convenience in notation, we set the minimum and maximum values of the attribute at zero and one, respectively, and set the part-worth for the least preferred level (“fair” in Figure 1) at zero. The curvilinear function in Figure 1 captures diminishing returns; that is, the marginal value of an attribute improvement becomes smaller at higher values of the attribute. (The application provides empirical support for diminishing returns.) Consequently, a straight-line relationship (referred to as the “vector model” in the conjoint analysis literature) may be somewhat limiting.8 The linear model also has the limitation that, unlike a curvilinear model, it cannot compensate for any errors introduced by the assumption that the attribute value for the middle level (“good” in Figure 1) is exactly halfway between the attribute values for the two extreme levels (“fair” and “excellent”).9 The height (or range) of the part-worth function, equal to I in Figure 1, provides a measure of the relative importance of the attribute. The curvature of the part-worth function is characterized by the shape parameter S. (S equal to one would denote a linear relationship; S less than one would indicate a diminishing returns curve of the type shown in Figure 1; and S greater than one would indicate a curvilinear relationship with increasing returns.)10
5. Determine Resource Requirements for Enhancing Image
Now we have the following measurements for each of the attributes (where j refers to each attribute, 1, 2, . . .n): the company’s current image, Cj on a scale of zero to one; the attribute’s importance to customers, Ij; and the shape parameter, Sj Next we ask managers who are responsible for action plans how difficult it would be to change the company’s marketing image on each of the attributes. Management is first apprised of the current marketing image of the company and its competitors. At this point, they are not told about the relative importances of the attributes so as not to potentially bias their inputs. For each attribute, managers consider how the perceived image fits with reality and subjectively estimate the least amount of resources required to make a unit improvement on that attribute. A common unit is used across all attributes. For instance, in the context of Figure 1, the common unit could be an improvement of 0.25 on the attribute. As a first approximation, we assume that the resources required are proportionate to the amount of change on an attribute, that is, two units of improvement will require twice as many resources. Later we will discuss how this assumption can be relaxed.
Improvement efforts may use resources to conduct communication campaigns, add personnel, broaden the product line, and so on. The resource estimates should reflect the most effective way to accomplish a unit change on one attribute relative to other attributes, given the competitive environment. Let Rj (on a scale of zero to ten) denote the resource requirement (subjective ratings averaged over managers) for enhancing the company’s image by one unit on attribute j. Thus, R5 = 8 and R1 = 4 would mean that it is twice as costly to make a unit enhancement on attribute five as it is on attribute one. As will be seen later, only the relative values of R across the attributes are needed in our method.
Alternatively, conjoint analysis can be used to assess the resource requirements on each attribute. The researchers can develop several scenarios regarding the company’s future, each scenario corresponding to the image the company might attain on the multiple attributes. Managers then rank (or rate) the scenarios in terms of the resources required to achieve the image described in the scenario. An analysis of the data would then reveal the resource requirement on each attribute.11
6. Prioritize Marketing Image Goals
Consider now an improvement of, say, 0.05 in the company’s image on attribute j. From Figure 1, the value of the improvement from the customer perspective is given by
where V is the value of the improvement; I is the importance; and C is the current image.
The expected cost of accomplishing the enhancement is Rj. Thus the highest priority improvement — that is, among improvements of 0.05 along each of the n attributes, the one that provides the maximum benefit per unit cost — can be obtained by comparing the ratios Vj/Rj across attributes and choosing the attribute, say, j* for which Vj/Rj is the maximum.12
To obtain the next most worthwhile improvement, we first reset customer image to reflect the change chosen in the earlier step; that is, we increase the (current image) C of attribute j* by 0.05, leaving all other Cj at current levels. We repeat the computation of Vj and recalculate the ratios Vj/Rj. [For attribute j*, the new value of V would be somewhat smaller than that obtained in the previous step because of the diminishing returns property (concavity) exhibited in Figure 1. For other attributes the ratios Vj/Rj would remain unaltered because no changes were made on those attributes.] The next improvement in image to make would be on the attribute for which the benefit-cost ratio Vj/Rj is maximum. (The chosen attribute in this step may or may not coincide with the attribute in the previous step.)
By repeating the above step many times, we trace a trajectory of changes in marketing image corresponding to increasing levels of resources that could be committed to enhancing the company’s marketing image.13 Of course, the greater the total resources committed, the greater would be the value of the enhanced market image. Thus the choice of image goals is based on available resources.
The Application
We applied this approach to a business unit of a Fortune “500” corporation. Although the results presented here are real, the industry and the attributes have been disguised to protect confidentiality. Suppose the management of a high-end department store chain in a metropolitan area wanted to assess its current marketing image and to set feasible goals for the next three years. Based on prior customer research using focus groups, management decided to research the following attributes: product variety, product quality, store attractiveness, reasonable prices, convenience (hours and location), and customer service.
The researchers interviewed a random sample of about one hundred customers (current and potential) from the relevant geographic area. Because the chain’s marketing strategy was aimed at the high end of the department store market, customers outside this target market were screened out with preliminary questions.
The marketplace interviews consisted of questions related to current shopping behavior, demographics, preferences, and image ratings of the department store chain and its key competitors. We used the full-profile method of conjoint analysis. Researchers described each attribute in detail to customers. Based on qualitative market research, three levels, fair, good, and excellent, had been chosen to describe each attribute. Six attributes at three levels makes 36 = 729 possible combinations. Researchers used a fractional factorial design using eighteen hypothetical combinations (cards) and asked customers to rank order the hypothetical department stores from the most to the least in terms of their likely patronage.14 The resulting rank order data were analyzed using the LINMAP computer program.15 The average attribute importances and shape parameters are displayed in Figure 2.
An attribute’s importance in conjoint analysis is defined as the range (height) of the part-worth function. As shown in Figure 2, it appears that there are two levels of importance over the six attributes. Four of the six attributes, product variety, product quality, convenience, and customer service, appear to be the most important, and the other two attributes are somewhat less important. All six part-worth functions had the diminishing returns property. That is, the increase in utility from fair to good is larger than the increase from good to excellent; the shape parameter S for each of the six attributes is less than one.
Each respondent also rated the department store on each of the six attributes. The customers, on average, rated the department store as good on all attributes except product variety, where they perceived it to be somewhat better than good, and customer service, where they rated it close to excellent.
Nine managers of the department store took the same questionnaire, rating the company’s image as if they were customers. These data permit us to examine how well managers understand customer perceptions and preferences. In Figure 3, we compare management’s perceptions of market image with the customers’ perceptions. Overall, managers were tougher on themselves than was warranted by customer perceptions. In particular, the managers viewed the image on store attractiveness, reasonableness of prices, and convenience to be much poorer than customers’ perceptions.
In Table 1, we compare the managers’ perceptions of the average market importances with the customers’ importances. Managers severely underestimated the importance of product variety (assortment, fashion, and styling) and overestimated, in particular, the importance of reasonableness of prices and customer service.
Next, we shared the customer image results with the managers and discussed them in order to ensure that the managers understood and adjusted to the customer perceptions. We then asked the managers to consider how the market image could be enhanced on each of the attributes (e.g., with communication campaigns, additional personnel, new information systems, improvements in depth and breadth of the product line, additional store hours, or the opening of new stores). They were asked to take into account the competitive environment and judge the level of resources (financial and human) required to make a unit improvement on each of the rating scales. To the extent that there were multiple ways of accomplishing an improvement with corresponding resource implications, they were to focus on the most effective (least costly) way. We also obtained ratings on a zero-to-ten scale, where ten denotes that a unit improvement on that attribute is the most difficult to accomplish. For each of the six attributes, we then averaged the resource requirement ratings across the nine managers.
Table 2 summarizes the key data required for prioritizing marketing image goals, including the data on resource requirements. For instance, the managerial data indicate that a unit move on convenience is the most difficult (i.e., requires the greatest amount of resources), whereas the comparable move on product variety is the least costly.
With this information, we can calculate the relative advantages of improving the different attributes. Consider, for instance, product variety. The utility U for a rating of C on this attribute is given by:
The current image on this attribute is 0.65. Suppose we consider enhancing its image. The value of an improvement from a rating of 0.65 to 0.70 on this attribute is given by
The cost of making a unit improvement on product variety is R = 4.8. Consequently, the benefit-cost ratio for this improvement is
Likewise, the benefit-cost ratio for improving from 0.70 to 0.75 on the same attribute is
Such computations were carried out for all possible improvements on each of the six attributes.
In Table 3, we rank order the benefit-cost ratios. The highest priority improvement, considering the value to the market and the cost of making the improvement, is on the product variety attribute. This is because product variety has one of the highest attribute importances (I = 19.2) and the smallest resource requirement (R = 4.8). By comparison, improvement in store attractiveness received the lowest priority because of the low importance attached to this attribute (I = 10.9) and the moderately high resource requirement (R = 6.7). Improvements in convenience received an intermediate level of priority because of the high resource requirements (R = 8.0), despite a high importance (I = 18.7).
The priorities in Table 3 provide a trajectory of desirable image enhancements. If the company has few available resources, it should enhance product variety. If resources are plentiful, it could enhance all the attributes.
Figure 4 provides two of the many possible sets of image goals. Goal A includes enhancements made with a limited level of resources (up to priority 2 in Table 3). Goal B includes enhancements using a larger level of resources (up to halfway on priority 4 in Table 3). Goal B would produce a more valuable market image but would also require considerably more resources than A.
Overall, the highest priority for marketing image enhancement in this case is product variety, followed by product quality and customer service. Management considered several sets of image goals. After extensive discussion regarding the availability of resources, management decided on a particular set of marketing image goals to implement.
Conclusion
We have proposed and applied an approach for prioritizing marketing image goals for a company. The approach is valuable because it takes into account customer inputs as well as the resource requirements needed to satisfy customers.
We assumed that the resources required are proportionate to the amount of change on an attribute. It is possible, however, that the marginal resource requirement may increase at higher levels of change. Given the difficulty and subjectivity of the resource requirement data, we did not include this refinement in the reported application. However, in future applications, it may be possible to provide managers with several scenarios of image enhancements on the multiple attributes and ask them to rank or rate the corresponding resource requirements. A conjoint analysis of the data would reveal the resources required for enhancement on each attribute, analogous to part-worth functions.16 A convex function can then be fitted to the resources required on each attribute so that the resource requirement Rj for a change of 0.05 on an attribute would be greater at higher levels of that attribute. The benefit-cost analysis would still be appropriate in this context.17
The proposed methodology is general enough to be applicable to industrial and consumer products and services and is potentially useful for not-for-profit organizations. The reported application illustrates that the approach yields actionable results in prioritizing the marketing image goals for a company.
References
1. J.C. Bevis, “Corporate Image Studies,” in Handbook of Marketing Research, ed. R. Ferber (New York: McGraw-Hill, 1974), pp. 206–218.
2. H. Barich and P. Kotler, “A Framework for Marketing Image Management,” Sloan Management Review, Winter 1991, pp. 94–104.
3. E.R. Gray and L.R. Smeltzer, “Corporate Image — An Integral Part of Strategy,” Sloan Management Review, Summer 1985, pp. 73–78.
4. P.E. Green and V. Srinivasan, “Conjoint Analysis in Consumer Research: Issues and Outlook,” Journal of Consumer Research 5 (1978): 103–123; and
P.E. Green and V. Srinivasan, “Conjoint Analysis in Marketing: New Developments with Implications for Research and Practice,” Journal of Marketing 54 (1990): 3–19.
5. Ibid.
6. Ibid.
7. D.R. Wittink and P. Cattin, “Commercial Use of Conjoint Analysis: An Update,” Journal of Marketing 53 (1989): 91–96.
8. Green and Srinivasan (1978).
9. The ideal-point model suggested in the conjoint analysis literature, which uses linear and quadratic terms, sometimes runs into the technical difficulty that, even though the observed part-worth values may be increasing over the levels, the fitted ideal-point model reaches a maximum value within the relevant range so that there is a decrease in utility for values beyond the maximum. This is illogical because, in our context, the attributes are such that higher values of the attribute imply greater preference. See:
Green and Srinivasan (1978) and
D. Pekelman and S.K. Sen, “Improving Prediction in Conjoint Measurement,” Journal of Marketing Research 16 (1979): 211–220.
10. In Figure 1, the shape parameter S is given by I (0.5)S = Q, so that S = [log(Q/I)]/[log 0.5].
11. For discussion of the relative merits of the conjoint and self-explicated methods, see:
Green and Srinivasan (1990).
12. Suppose the unit used in Step 5 above to obtain the resource requirement parameter R is 0.25. Then by the proportionality assumption made in Step 5, the resource required for a 0.05 improvement on attribute j is (0.05/0.25) Rj = 0.2 Rj. The constant term 0.2 does not affect the comparison across attributes and hence is not considered.
13. If the part-worth functions are concave (i.e., exhibit diminishing returns) as displayed in Figure 1, this step-by-step marginal analysis has the desirable mathematical property of producing undominated solutions. That is, at every step of the procedure, the customer value of the image is maximized subject to not exceeding the corresponding resource level. See:
B. Fox, “Discrete Optimization via Marginal Analysis,” Management Science 13 (1966): 210–216.
If the part-worth functions are not concave, the step-by-step marginal analysis needs to be replaced by dynamic programming. See:
R.E. Bellman and S.E. Dreyfus, Applied Dynamic Programming (Princeton, New Jersey: Princeton University Press, 1962).
14. P.E. Green, “On the Design of Choice Experiments Involving Multifactor Alternatives,” Journal of Consumer Research 1 (1974): 61–68.
15. V. Srinivasan and A.D. Shocker, “Estimating the Weights for Multiple Attributes in a Composite Criterion Using Pairwise Judgments,” Psychometrika 38 (1973): 473–493.
16. The resource requirement part-worth function would be scaled such that there are no incremental resources required to maintain the image at the current level on that attribute.
17. The step-by-step marginal analysis would produce undominated solutions as long as the customer attribute utility functions are concave and the resource requirement functions are convex. If the concavity-convexity assumptions are violated, then the step-by-step approach needs to be replaced by dynamic programming so as to maximize benefit for different resource levels. See:
Bellman and Dreyfus (1962).