The two main issues to consider in conducting free association tasks are what types of probes to give to subjects, and how to code and interpret the resulting data. In order not to bias results, it is best to move from general considerations to more specific considerations, as we illustrated earlier. Thus, ask consumers first what they think of the brand as a whole without reference to any particular category, followed by specific questions about particular products and aspects of the brand image.

Consumers’ responses to open-ended probes can be either oral or written. The advantage of oral responses is that subjects may be less deliberate and more spontaneous in their reporting. In terms of coding the data, divide the protocols each consumer provides into phrases and aggregate them across consumers in categories. Because of their more focused nature, responses to specific probes and follow-up questions are naturally easier to code.

Classic projective techniques use incomplete or ambiguous stimuli to elicit consumer thoughts and feelings. One approach is “bubble exercises,” which depict different people buying or using certain products or services. Empty bubbles, as in cartoons, are placed in the scenes to represent the thoughts, words, or actions of one or more of the participants. Marketers then ask consumers to “fill in the bubble” by indicating what they believe is happening or being said in the scene. The stories and conversations told this way can be especially useful for assessing user and usage imagery for a brand.

Another useful technique is comparison tasks, in which we ask consumers to convey their impressions by comparing brands to people, countries, animals, activities, fabrics, occupations, cars, magazines, vegetables, nationalities, or even other brands.13 For example, we might ask consumers, “If Dannon yogurt were a car, which one would it be? If it were an animal, which one might it be? Looking at the people depicted in these pictures, which ones do you think would be most likely to eat Dannon yogurt?” In each case, we would ask a follow-up question about why subjects made the comparison they did. The objects people choose to represent the brand and their reasons can provide glimpses into the psyche of the consumer with respect to a brand, particularly useful in understanding imagery associations.

By examining the answers to probes, researchers may be better able to assemble a rich image for the brand, for example, identifying key brand personality associations. Branding Brief 9-3 outlines how hotel chain Joie de Vivre uses magazine imagery to clarify its brand positions.

Archetype research is one technique for eliciting deeply held consumer attitudes and feelings. According to cultural anthropologist G. C. Rapaille, consumers often make purchase decisions based on factors of which they are only subconsciously aware. Conventional market research typically does not uncover these motivations, so Rapaille employs the archetype research technique to find them.14

Rapaille believes children experience a significant initial exposure to an element of their world called the “imprinting moment.” The pattern that emerges when we generalize these imprinting moments for the entire population is the archetype, a fundamental psychological association, shared by the members of the culture, with a given cultural object. Different cultures have dramatically different archetypes for the same objects. In France, the archetype for cheese is “alive” because age is its most important trait. By contrast, the U.S. archetype for cheese is “dead”; it is wrapped in plastic (“a body-bag”), put in the refrigerator (“a morgue”), and pasteurized (“scientifically dead”).

Rapaille uses relaxation exercises and visualization with consumers to find the imprinting moments appropriate to the product he is researching. For example, at a focus group, he will dim the lights, play soothing music, and coax the subjects into a meditative state. He will then elicit stories about the product from the subjects and analyze these stories to illuminate the archetype.

We can assess brand personality more definitively through adjective checklists or ratings. Jennifer Aaker conducted a research project that provides an interesting glimpse into the personality of a number of well-known brands, as well as a methodology to examine the personality of any one brand. Based on an extensive data collection of ratings of 114 personality traits on 37 brands in various product categories by over 600 individuals representative of the U.S. population, she created a brand personality scale that reflected the following five factors (with underlying facets) of brand personality:22

1. Sincerity (down-to-earth, honest, wholesome, and cheerful)

2. Excitement (daring, spirited, imaginative, and up-to-date)

3. Competence (reliable, intelligent, and successful)

4. Sophistication (upper class and charming)

5. Ruggedness (outdoorsy and tough)

Figure 9-4 depicts the specific trait items that make up the Aaker brand personality scale. Respondents in her study rated how descriptive each personality trait was for each brand according to a seven-point scale (1 = not at all descriptive; 7 = extremely descriptive). Aaker averaged responses to provide summary measures. Some brands tend to be strong on one particular factor; some brands like Nike are high on more than one factor; some brands score poorly on all factors.

A cross-cultural study exploring the generalizability of this scale outside the United States found that three of the five factors applied in Japan and Spain, but that a “peacefulness” dimension replaced “ruggedness” both in Japan and Spain, and a “passion” dimension emerged in Spain instead of “competency.”23 Research on brand personality in Korea revealed that two culture-specific factors emerge (“passive likeableness” and “ascendancy”), reflecting the importance of Confucian values in Korea’s social and economic systems.24

Brand recognition requires consumers to identify the brand under a variety of circumstances and can rest on the identification of any of the brand elements. The most basic recognition test gives consumers a set of individual items visually or orally and asks them whether they think they’ve previously seen or heard of these items. To provide a more sensitive test, it is often useful to include decoys or lures—items consumers could not possibly have seen. In addition to “yes” or “no” responses, consumers can also rate how confident they are in their recognition of an item.

Other, somewhat more subtle, recognition measures test “perceptually degraded” versions of the brand, which are masked or distorted in some way or shown for extremely brief duration. For example, we can test brand name recognition with missing letters. Figure 9-6 tests your ability to recognize brand names with less than full information. These more subtle measures may be particularly important for brands that have a high level of recognition, in order to provide more sensitive assessments.40

Brand recognition is especially important for packaging, and some marketing researchers have used creative means to assess the visibility of package design. As a starting point, they consider the benchmark or “best case” of the visibility of a package when a consumer (1) with 20–20 vision (2) is face-to-face with a package (3) at a distance of less than five feet (4) under ideal lighting conditions.

A key question then is whether the package design is robust enough to be still recognizable if one or more of these four conditions are not present. Because shopping is often not conducted under “ideal” conditions, such insights are important. For example, many consumers who wear eyeglasses do not wear them when shopping in a supermarket. Is the package still able to effectively communicate to consumers under such conditions?

Research methods using tachistoscopes (T-scopes) and eye tracking techniques exist to test the effectiveness of alternative package designs according to a number of specific criteria:

• Degree of shelf impact

• Impact and recall of specific design elements

• Perceived package size

• Copy visibility and legibility

• Distance at which the package can first be identified

• Angle at which the package can first be identified

• Speed with which the package can be identified

These additional measures can provide more sensitive measures of recognition than simple “yes” or “no” tasks. By applying these direct and indirect measures of brand recognition, marketers can determine which brand elements exist in memory and, to some extent, the strength of their association. One advantage that brand recognition measures have over recall measures is the chance to use visual recognition. It may be difficult for consumers to describe a logo or symbol in a recall task; it’s much easier for them to assess the same elements visually in a recognition task.

Nevertheless, brand recognition measures provide only an approximation of potential recallability. To determine whether consumers will actually recall the brand elements under various circumstances, we need measures of brand recall.

To demonstrate brand recall, consumers must retrieve the actual brand element from memory when given some related probe or cue. Thus, brand recall is a more demanding memory task than brand recognition because consumers are not just given a brand element and asked to say whether they’ve seen it before.

Different measures of brand recall are possible depending on the type of cues provided to consumers. Unaided recall on the basis of “all brands” provided as a cue is likely to identify

only the very strongest brands. Aided recall uses various types of cues to help consumer recall. One possible sequence of aided recall might use progressively narrower cues—such as product class, product category, and product type labels—to provide insight into the organization of consumers’ brand knowledge structures.

For example, if recall of the Porsche 911—a high-performance German sports car—in non-German markets were of interest, recall probes could begin with “all cars” and move to more and more narrowly defined categories such as “sports cars,” “foreign sports cars,” or even “high-performance German sports cars.” Marketers could ask consumers: “When you think of foreign sports cars, which brands come to mind?”

Other types of cues can help measure brand recall. For example, marketers can ask about product attributes (“When you think of chocolate, which brands come to mind?) or usage goals (“If you were thinking of having a healthy snack, which brands come to mind?”). Often, to capture the breadth of brand recall and to assess brand salience, we might need to examine the context of the purchase decision or consumption situation, such as different times and places. The stronger the brand associations to these non-product considerations, the more likely it is that consumers will recall them when given those situational cues.

When combined, measures of recall based on product attribute or category cues and situational or usage cues give an indication of breadth and depth of recall. We can further distinguish brand recall according to the order as well as the latency or speed of recall. In many cases, people will recognize a brand when it is shown to them and will recall it if they are given a sufficient number of cues. Thus, potential recallability is high. The more important issue is the salience of the brand: Do consumers think of the brand under the right circumstances, for example, when they could be either buying or using the product? How quickly do they think of the brand? Is it automatically or easily recalled? Is it the first brand they recall?

Any research measure must consider the issue of consumers making up responses or guessing. That problem may be especially evident with certain types of aided awareness or recognition measures for the brand. Spurious awareness occurs when consumers erroneously claim they recall something they really don’t and that may not even exist. For example, one market research firm, Oxtoby-Smith, conducted a benchmark study of awareness of health and beauty products.41 In the study, the firm asked consumers questions like this:

“The following is a list of denture adhesive brand names. Please answer yes if you’ve heard the name before and no if you haven’t. Okay? Orafix? Fasteeth? Dentu-Tight? Fixodent?”

Although 16 percent of the sample reported that they had heard of Dentu-Tight, there was one problem: it didn’t exist! Similarly high levels of reported recall were reported for plausible-sounding but fictitious brands such as Four O’Clock Tea (8 percent), Leone Pasta (16 percent), and Mrs. Smith’s Cake Mix (31 percent). On the basis of this study, Oxtoby-Smith found that spurious awareness was about 8 percent for new health and beauty products and even higher in some other product categories. In one case, a proposed line extension was mistakenly thought to already exist by about 50 percent of the sample (a finding that no doubt sent a message to the company that it should go ahead and introduce the product!).

From a marketing perspective, the problem with spurious awareness is that it may send misleading signals about the proper strategic direction for a brand. For example, Oxtoby-Smith reported that one of its clients was struggling with a 5 percent market share despite the fact that 50 percent of survey respondents reported they were aware of the brand. On the surface, it would seem a good idea to improve the image of the brand and attitudes toward it in some way. Upon further examination, marketers determined that spurious awareness accounted for almost half the survey respondents who reported brand awareness, suggesting that a more appropriate solution to the true problem would be to first build awareness to a greater degree. Marketers should be sensitive to the possibilities of misleading signals because of spurious brand awareness, especially with new brands or ones with plausible-sounding names.

The advantage of aided recall measures is that they yield insight into how brand knowledge is organized in memory and what kind of cues or reminders may be necessary for consumers to be able to retrieve the brand from memory. Understanding recall when

we use different levels of product category specificity as cues is important, because it has implications for how consumers form consideration sets and make product decisions.

For example, again consider the Porsche 911. Assume consumer recall of this particular car model was fairly low when all cars were considered but very high when foreign sports cars were considered. In other words, consumers strongly categorized the Porsche 911 as a prototypical sports car but tended to think of it in only that way. If that were the case, for more consumers to entertain the possibility of buying a Porsche 911, we might need to broaden the meaning of Porsche so that it has a stronger association to cars in general. Of course, such a strategy risks alienating existing customers who had been initially attracted by the “purity” and strong identification of the Porsche 911 as a sports car. The choice of appropriate strategy would depend on the relative costs and benefits of targeting the two different segments.

The point is that the category structure that exists in consumers’ minds—as reflected by brand recall performance—can have profound implications for consumer choice and marketing strategy, as demonstrated by The Science of Branding 9-1. The insights gleaned from measuring brand recall are also valuable for developing brand identity and integrated marketing communication programs, as we showed in Chapters 4 and 6 . For example, we can examine brand recall for each brand element to explore the extent to which any one of these (name, symbol, or logo) suggests any other. Are consumers aware of all the different brand elements and how they relate?

We also need a complete understanding of brand image, as covered in the following section.

A more complicated quantitative technique to assess overall brand uniqueness is multidimensional scaling, or perceptual maps. Multidimensional scaling (MDS) is a procedure for determining the perceived relative images of a set of objects, such as products or brands. MDS transforms consumer judgments of similarity or preference into distances represented in perceptual space. For example, if brands A and B are judged by respondents to be the most similar of a set of brands, the MDS algorithm will position brands A and B so that the distance between them in multidimensional space is smaller than the distance between any other two pairs of brands. Respondents may base their similarity between brands on any basis—tangible or intangible.43

Figure 9-8 displays a hypothetical perceptual map of restaurants in a particular market. Segment 1 is more concerned with health than taste and is well targeted by Brand B; segment 2 is more concerned with taste and is well targeted by Brand C. Brand A is trapped in the middle. It either must improve taste to provide a healthy alternative to Brand C for segment 2, or it must improve healthiness to prove a tastier alternative to Brand B for segment 1.

Another set of measures closely related to brand attitudes and consideration looks at purchase intentions44 and focus on the likelihood of buying the brand or of switching to another brand. Research in psychology suggests that purchase intentions are most likely to

be predictive of actual purchase when there is correspondence between the two in the following dimensions:45

• Action (buying for own use or to give as a gift)

• Target (specific type of product and brand)

• Context (in what type of store based on what prices and other conditions)

• Time (within a week, month, or year)

In other words, when asking consumers to forecast their likely purchase of a product or a brand, we want to specify exactly the circumstances—the purpose of the purchase, the location of the purchase, the time of the purchase, and so forth. For example, we could ask consumers:

“Assume your refrigerator broke down over the next weekend and could not be inexpensively repaired. If you went to your favorite appliance store and found all the different brands competitively priced, how likely would you be to buy a General Electric refrigerator?”

Consumers could indicate their purchase intention on an 11-point probability scale that ranges from 0 (definitely would not buy) to 10 (definitely would buy).

Bain’s Frederick Reichheld suggests there is only one customer question that really matters: “How likely is it that you would recommend this product or service to a friend or colleague?” According to Reichheld, a customer’s willingness to recommend results from all aspects of a customer’s experience.46

Reicheld uses answers to this question to create what he calls a Net Promoter Score (NPS). Specifically, in a survey, customers are asked to rate their likelihood to recommend on a 0–10-point scale. Marketers then subtract detractors (those who gave a 0–6) from promoters (those who gave a 9 or 10) to arrive at the NPS score. Customers who rate the brand with a 7 or 8 are deemed passively satisfied and are not included. A typical set of NPS scores falls in the 10–30 percent range, but world-class companies can score over 50 percent. Some firms with top NPS scores include USAA (89 percent), Apple (77 percent), Amazon.com (74 percent), and Google (71 percent).

Several companies have seen benefits from adopting NetPromoter scores as a means of tracking brand health. When the European unit of GE Healthcare overhauled its call center and put more specialists in the field, GE Healthcare’s Net Promoter scores jumped 10–15 points. BearingPoint found clients who gave it high Net Promoter scores showed the highest revenue growth. When Intuit applied Net Promoter to its TurboTax product, feedback revealed dissatisfaction with the software’s rebate procedure. After Intuit dropped the proof-of-purchase requirement, sales jumped 6 percent.

To capture reported brand usage and behavioral loyalty, we could ask consumers several questions directly. Or we could ask them what percentage of their last purchases in the category went to the brand (past purchase history) and what percentage of their planned next purchases will go to the brand (intended future purchases). For example, the marketers or brand managers of Duracell batteries might ask the following questions:

• Which brand of batteries do you usually buy?

• Which brand of batteries did you buy last time?

• Do you have any batteries on hand? Which brand?

• Which brands of batteries did you consider buying?

• Which brand of batteries will you buy next time?

These types of questions can provide information about brand attitudes and usage for Duracell, including potential gaps with competitors and the names of other brands that might be in the consideration set at the time of purchase.

Marketers can make their measures open ended, force consumers to choose one of two brands, or offer multiple choice or rating scales. They can compare the answers with actual measures of consumer behavior to assess whether consumers are accurate in their predictions. For example, if 30 percent of consumers reported, on average, that they thought they would take their vitamins daily over the next two weeks, but only 15 percent of consumers reported two weeks later that they actually had done so during that period, then Centrum brand managers might need to devise strategies to better convert intentions to actual behavior.

In a business-to-business setting, Narayandas advocates analyzing sales records, talking to sales teams, and conducting surveys to assess where customers stand on a “loyalty ladder.”47

Several different approaches have been suggested to measure the second component of brand resonance—brand attachment.48 Some researchers like to characterize it in terms of brand love.49 One study proposed a brand love scale that consists of 10 items: (1) This is a wonderful brand; (2) This brand makes me feel good; (3) This brand is totally awesome; (4) I have neutral feelings about this brand (reverse-coded item); (5) This brand makes me very happy; (6) I love this brand; (7) I have no particular feelings about this brand (reverse-coded item); (8) This brand is a pure delight; (9) I am passionate about this brand; and (10) I am very attached to this brand.50

Another study found 11 dimensions that characterized brand love:51

1. Passion (for the brand).

2. Duration of the relationship (the relationship with the brand exists for a long time).

3. Self-congruity (congruity between self-image and product image).

4. Dreams (the brand favors consumer dreams).

5. Memories (evoked by the brand).

6. Pleasure (that the brand provides to the consumer).

7. Attraction (feel toward the brand).

8. Uniqueness (of the brand and/or of the relationship).

9. Beauty (of the brand).

10. Trust (the brand has never disappointed).

11. Declaration of affect (feel toward the brand).

One promising approach defines brand attachment in terms of two underlying constructs—brand-self connections and brand prominence—where each of those two dimensions have two subdimensions, suggesting the following sets of measures:52

1. Brand-Self Connection

   1. Connected: “To what extent do you feel that you are personally connected to (Brand)?”

   2. Part of Who You Are: “To what extent is (Brand) part of you and who you are?”

2. Brand Prominence

   1. Automatic: “To what extent are your thoughts and feelings towards (Brand) often automatic, coming to mind seemingly on their own?”

   2. Naturally: “To what extent do your thoughts and feelings towards (Brand) come to you naturally and instantly?”

Although measuring behavioral loyalty and attitudinal attachment may require a fairly structured set of questions, both sense of community and active engagement could call for more varied measures because of their diverse set of issues.

One interesting concept that has been proposed with respect to community is social currency, developed by brand consultants Vivaldi Partners. They define social currency as “the extent to which people share the brand or information about the brand as part of their everyday social lives at work or at home.” Figure 9-10 displays the different dimensions that make up the social currency concept according to Vivaldi Partners.

According to the brand resonance model, active engagement for a brand is defined as the extent to which consumers are willing to invest their own personal resources—time, energy, money, and so on—on the brand beyond those resources expended during purchase or consumption of the brand.

For example, in terms of engagement, in-depth measures could explore word-of-mouth behavior, online behavior, and so forth. For online behavior, measures could explore the extent of customer-initiated versus firm-initiated interactions, the extent of learning and teaching by the customer versus by the firm, the extent of customers teaching other customers, and so on.53

The key to such metrics is the qualitative nature of the consumer-brand interaction and how well it reflects intensity of feelings. One mistake many Internet firms made was to put too much emphasis on “eyeballs” and “stickiness”—the number and duration of page views at a Web site, respectively. The depth of the underlying brand relationships of the customers making those visits, however, and the manner in which those relationships manifest themselves in brand-beneficial actions, will typically be more important.

Accordingly, researchers are attempting to determine the brand value of different online and social media activities.54 For example, how important is a “like” from a user on Facebook? One firm estimated that bringing a user on as a fan could be worth between 44 cents and $3.60 in equivalent media value from increased impressions generated from the Facebook newsfeed. Critics of the study, however, pointed out that not all fans are created equal.55

Several different specific approaches have been suggested to measure brand engagement. The Science of Branding 9-2 provides a detailed breakdown of the concept.

Boston University’s Susan Fournier argues that brands can and do serve as viable relationship partners, and she suggests a reconceptualization of the notion of brand personality within this framework.56 Specifically, the everyday execution of marketing mix decisions constitutes a set of behaviors enacted on the part of the brand. These actions trigger a series of inferences regarding the implicit contract that appears to guide the engagement of the consumer and brand and, hence, the type of relationship formed.

Brand personality as conceptualized within this framework describes the relationship role enacted by the brand in its partnership capacity. For example, if the brand expresses behaviors that signal commitment to the consumer, and further if it sends gifts as symbols of affection, the consumer may infer a courtship or marriage type of engagement with the brand.

Fournier identifies a typology of 15 different relationship types characterizing consumers’ engagement with brands (see Figure 9-11). Fournier argues that this relationship role view of brand personality provides more actionable guidance to managers who wish to create and manage their brand personalities in line with marketing actions than does the trait-based

view, which identifies general personality tendencies that might or might not be connected to marketing strategies and goals.

Fournier has conducted fascinating research that reframes the conceptualization and measurement of brand strength strictly in relationship terms. It defines a brand’s strength in terms of the strength, depth, and durability of the consumer-brand relational bond using the multifaceted concept of brand relationship quality, or BRQ. Extensive validation work supported a multifaceted hierarchical structure for the BRQ construct that includes six main dimensions of relationship strength, many with important subfacets. The main facets are (1) interdependence, (2) self-concept connection, (3) commitment, (4) love/passion, (5) intimacy, and (6) brand partner quality.

Fournier argues that these facets and their subfacets (such as trust within the partner quality facet or consumer-to-firm and firm-to-consumer intimacy) have superior diagnostic value over competing strength measures, and she suggests they have greater managerial utility in their application. In her experience, BRQ measures have been successfully incorporated in brand tracking studies, where they provide profiles of brand strength versus competitors, useful ties to marketplace performance indicators, and specific guidance for the enhancement and dilution of brand equity through managerial actions in the marketplace. Although brand relationship quality shares some characteristics with brand resonance, it provides valuable additional perspectives and insights.

The six main facets of brand relationship quality are as follows:

• Interdependence: The degree to which the brand is ingrained in the consumer’s daily course of living, both behaviorally (in terms of frequency, scope, and strength of interactions) and cognitively (in terms of longing for and preoccupation with anticipated brand interactions). Interdependence is often revealed through the presence of routinized behavioral rituals surrounding brand purchase and use, and through separation anxiety experienced during periods of product deprivation. At its extremes, interdependence becomes dependency and addiction.

• Self-concept connection: The degree to which the brand delivers on important identity concerns, tasks, or themes, thereby expressing a significant part of the self-concept, both past (including nostalgic references and brand memories) and present, and personal as well as social. Grounding of the self provides feelings of comfort, connectedness, control, and security. In its extreme form, self-connection reflects integration of concepts of brand and self.

• Commitment: Dedication to continued brand association and betterment of the relationship, despite circumstances foreseen and unforeseen. Commitment includes professed faithfulness and loyalty to the other, often formalized through stated pledges and publicized intentions. Commitment is not defined solely by sunk costs and irretrievable investments that pose barriers to exit.

• Love/passion: Affinity toward and adoration of the brand, particularly with respect to other available alternatives. The intensity of the emotional bonds joining relationship partners may range from feelings of warmth, caring, and affection to those of true passion. Love includes the belief that the brand is irreplaceable and uniquely qualified as a relationship partner.

• Intimacy: A sense of deep familiarity with and understanding of both the essence of the brand as a partner in the relationship and the nature of the consumer-brand relationship itself. Intimacy is revealed in the presence of a strong consumer-brand relationship culture, the sharing of little-known personal details of the self, and an elaborate brand memory containing significant experiences or associations. Intimacy is a two-dimensional concept: the consumer develops intimate knowledge of the brand, and also feels a sense of intimacy exhibited on the part of the brand toward the individual as a consumer.

• Partner quality: Perceived partner quality involves a summary judgment of the caliber of the role enactments performed by the brand in its partnership role. Partner quality includes three central components: (1) an empathic orientation toward the other (ability of the partner to make the other feel wanted, cared for, respected, noticed, and important; responsiveness to needs); (2) a character of reliability, dependability, and predictability in the brand; and (3) trust or faith in the belief that the brand will adhere to established relationship rules and be held accountable for its actions.