University of Wollongong

Research Online
Faculty of Commerce - Papers (Archive) Faculty of Business

2012

Emotional branding pays off: how brands meet

share of requirements through bonding,
companionship, and love
John Rossiter
University of Wollongong, [email protected]

Steven Bellman
Murdoch University

Publication Details
Rossiter, J. & Bellman, S. (2012). Emotional branding pays off: How brands meet share of requirements through bonding,
companionship, and love. Journal of Advertising Research, 52 (3), 291-296.

Research Online is the open access institutional repository for the University of Wollongong. For further information contact the UOW Library:
[email protected]
Emotional branding pays off: how brands meet share of requirements
through bonding, companionship, and love
Abstract
Emotional branding is defined here as the consumers attachment of a strong, specific, usage-relevant
emotionsuch as Bonding, Companionship, or Loveto the brand. The present large-scale survey of buyers
of frequently purchased consumer products finds that, for such products, full-strength emotional branding is
attained among, at most, only about 25 per cent of the brands buyers but that, if attained, it pays off massively
in terms of personal share of purchases. Emotional branding may well be more widely effective for high
involvement, positively motivated products (not surveyed here). It seems that advertising can generate the
expectancy of strong, specific, emotional attachment, but very favorable brand usage experience must follow if
this approach is to be successful. In general, the traditional benefit-based USP advertising strategy seems less
risky with lesser though more widespread effectiveness.

Keywords
branding, pays, off, brands, meet, emotional, share, love, requirements, bonding, companionship

Disciplines
Business | Social and Behavioral Sciences

Publication Details
Rossiter, J. & Bellman, S. (2012). Emotional branding pays off: How brands meet share of requirements
through bonding, companionship, and love. Journal of Advertising Research, 52 (3), 291-296.

This journal article is available at Research Online: http://ro.uow.edu.au/commpapers/2857

 Emotional Branding Pays Off

JOHN R. ROSSITER
University of Wollongong
[email protected]

STEVEN BELLMAN
Murdoch University
[email protected]

Journal of Advertising Research, 2010 or 2011, forthcoming.

 2

Emotional Branding Pays Off

Abstract: This large-scale survey demonstrates that consumers acquisition of strong emotional

attachment to the brand pays off in terms of a significantly greater volume of purchase of the

brand (SOR%). The specific emotions are measured by the practitioners pick any item-

battery, in which the consumer checks only those emotions felt at full strength. The specific

emotions of Bonding (in which the consumer checks yes to the item Its my brand) and

Love (yes to I would say that I feel deep affection, like love, for this brand and would be

really upset if I couldnt have it) are shown to be effective for a wide range of consumer

products. The specific emotion of Companionship (yes to This brand is like a companion to

me) seems to be effective for functional products such as gasoline that provide an ongoing

complementary usage experience. The share-of-requirements (SOR%) return to these

attachment-like emotions is at least as great as that of a favorable Brand Attitude for hedonic

products and is actually greater for utilitarian products. However, emotional attachments to

the brand are much more difficult for brand marketers to attain for the brand than a favorable

attitude, and further research is needed to verify that advertising plays a significant role in

creating these attachments.

 3

INTRODUCTION

Emotional branding defined as the successful attachment of a specific emotion to a

brand is an increasingly popular advertising strategy (see Roberts, 2004; Rossiter and Bellman,

2005). Notable examples include attachment of the specific emotion of nostalgia to the Kodak

brand of film, bonding to the Jim Beam bourbon brand, and love to the McDonalds brand.

Perusal of recent campaigns in Advertising Age and Admap suggests that emotional branding is at

least as widely used nowadays as the more traditional key benefit advertising strategy, which

works by instilling a favorable overall attitude toward the brand (see Mayer, 1958, for the first

account of the USP key-benefit approach and see Rossiter and Bellman, 2005 for a benefit-

based and emotion-based brand attitude model). Prior to the present study, however, no study

has documented the behavioral effectiveness of emotional branding.

The present study, conducted with a large and nationally representative sample of

consumers, examines the effectiveness of five quite widely targeted attachment-like emotions,

labeled as Trust, Bonding, Resonance, Companionship, and Love. These emotions are measured

in an extreme full-strength manner by using the practitioners free choice, or pick any

procedure (see Barnard and Ehrenberg, 1990). The effectiveness measure is the consumers

purchase share-of-requirements, or SOR%, a volume-based sales metric favored by marketing

managers (see Baldinger and Rubinson, 1996, and Hu, Lodish, Krieger, and Hayati, 2009). The

behavioral effectiveness of the emotional approach overall is assessed by forming a summary

variable called Brand Emotional Attachment. The SOR% predicted by Brand Emotional

Attachment is then compared with that of the traditional predictor of purchase, Brand Attitude

(see Axelrod, 1969, and Achenbaum, 1972).

 4

METHOD

Overview

For the present study, an online survey of brand perceptions, attitude ratings, and usage

was conducted by the Interactive Television Research Institute in Perth, Australia, among

members of the public who belong to the institutes consumer panel. In the Perth market, U.S.

and U.K. brands dominate in most consumer product categories aside from beer, where

Australian brands have the largest market shares. The product categories included in the survey

were gasoline and laundry detergent (utilitarian products for which the emotional approach

would not be expected to work well) and beer and instant coffee (hedonic products for which

the emotional approach should work well). Men were surveyed for gasoline and beer, and

women were surveyed for laundry detergent and instant coffee, representing historically and

respectively the main buyers of these types of products.

Consumer sample

Just over one thousand (1,025) consumers from the panel volunteered to participate in the

survey. The screening questions required participants to be buyers of the product category, and

so the more widely bought products resulted in women making up approximately two-thirds of

the overall sample. Sample sizes by product category were 338 men for gasoline, 244 men for

beer, 606 women for laundry detergent, and 468 women for instant coffee.

Almost half of the participants, 46%, were in the 18-34 age group, which is generally

regarded as the prime age target for new consumer product brands. The participants were also
 5

more upmarket, with 32% in professional occupations, than either the Australian population

(19% professionals) or the U.S. population (13% professionals).

Consumer product brands

All brands available in the Perth market in each product category were included in the

survey. There were 25 brands of laundry detergent, 18 brands of beer, 13 brands of instant

coffee, and eight brands of gasoline. The brands were represented in the online questionnaire by

their full-color packs or logos so as to facilitate correct recognition (see Rossiter and Bellman,

2005) and to more validly measure brand emotional associations and brand attitude by presenting

the brand as it is encountered at the point of purchase.

Measures

New single-item measures of the specific brand emotions the main independent

variables were developed in a pretest for the present study. In keeping with best measurement

practice in brand and advertising tracking studies (the practice of leading research companies

such as Ipsos/ASI and Millward Brown), the new measures were administered in pick any, or

free choice, format; participants were asked to check that is, pick only those descriptions

that, in their mind, apply to the brand. Consumers had to pick them absolutely, not rate them

relatively or by degree. Participants answers were scored as picked = 1, and not picked = 0.

The wording of the specific-emotion measures thoroughly checked for clear consumer

understanding in the open-ended pretest was as follows:

Trust: I trust this brand

Bonding: I regard it as my brand

 6

Resonance: I use this brand because it fits my self-image

Companionship: It is like a companion to me

Love: I would say that I feel deep affection, like love, for this brand and would be

really upset if I couldnt have it

The alternative independent variable, Brand Attitude, or Ab, was measured with a single

semantic-differential item: Bad...Good (see Bergkvist and Rossiter, 2007, for evidence of the

validity of this single-item measure of attitude). This was a 7-step bipolar measure scored 3 to

+3. Participants rated all brands and were instructed to give an expected Ab rating, based on the

brands pack or logo, if they didnt know the brand and to give a zero rating only if they truly felt

neutral about a brand.

The dependent variable, Brand Purchase Volume, was measured as share-of-requirements

(SOR%) using the constant sum method (e.g., Baldinger and Rubinson, 1996). Each participant

was asked to allocate 100 percentage points across the brands that he or she had bought over the

past year. The online software in the questionnaire prompted the participant to check that the

shares added to 100%.

FINDINGS

Three major findings are reported and discussed. First, the incidence of the specific

attachment-like emotions for the average brand the consumer buys in each product category is

examined. Second, the relationships of the specific emotions to the dependent variable, SOR%,

are graphed for brands that the consumer buys in each product category. Finally, an overall

Brand Emotional Attachment variable is computed and its predictive correlation is compared

with the predictive correlation of Brand Attitude for predicting the brands percentage SOR.
 7

Incidence findings

The incidences (in percent) of consumers who have attained the specific associations of

Trust, Bonding, Resonance, Companionship, or Love for the average brand that they buy in each

of the four product categories are shown in Figure 1. The results in the figure show that

consumers fully Trust most but not all of the brands that they buy. Gasoline brands had the

lowest incidence of Trust, with the average brand of gasoline bought trusted by 59% of male

buyers. The incidence of Trust was 69% to 80% for the average brand in the three other product

categories.

Figure 1 about here

The results in the figure reveal the difficulty of attaining stronger attachment-like

emotions for brands of consumer products. Bonding (its my brand) was achieved among only

about 20% of consumers for the average brand that they buy lower in the gasoline category

among men and higher among women consumers for the average brand of instant coffee.

Resonance (fits my self-image) was achieved for just 20% of gasoline brands and reached only

9% for the average brand of beer that men buy. Companionship (its like a companion) and

Love (feeling deep affection) were emotions reported by just 1% to 8% of consumers for the

average brand bought, or about 4% overall for brands of consumer products.

Specific emotions and SOR%

 8

Cross-tabulations of Trust, Bonding, Resonance, Companionship, and Love with the

personal share of requirements (SOR%) that each specific attachment-like emotion delivers for

the brand are shown in Figure 2. The horizontal line in each of the graphs indicates the average

share of requirements, in percent, given to those brands that the consumer buys.

Figure 2 about here

For gasoline, the average brand bought (by men) accounted for an SOR of 30% (see first

graph in the figure). Being regarded as a Trusted brand of gasoline did not guarantee a high

volume of purchase of that brand. On the other hand, two of the attachment emotions, Bonding

(its my brand) and Companionship (its like a close friend), deliver more than double the

average purchase volume (SOR%), such that the bonded or companion brand accounts for

two-thirds of the consumers purchases in the category.

For beer, the average SOR of 21% indicates that the average male beer drinker shares his

purchases over about five brands and is thus not very loyal (see second graph in the figure).

Trust in the brand does not raise the brands share of consumption above this average level.

However, two of the emotional attachments Bonding and, somewhat surprising for men to

admit, Love evidently increase the brands SOR to an above-average 32% and 38%,

respectively.

For laundry detergent, women buyers, on average, report using an average of just over

three of the 25 available brands (shown in the third graph in the figure by the average SOR of
 9

31%). Trust, surprisingly, delivers a below-average SOR of only 24%. On the other hand,

Bonding delivers almost double the average SOR, 60%, for my brand of laundry detergent, and

Love, somewhat surprisingly for this utilitarian product, delivers an above-average SOR of 54%

for the loved brand.

For instant coffee, the average female buyer reported buying slightly fewer than three

different brands over the past year. Instant coffee brands have the highest loyalty highest of the

four products surveyed an SOR of 37% for the average brand bought (see fourth graph in the

figure). Trust is not sufficient to ensure loyal purchasing, delivering a below-average SOR of

29%. However, those women who say its my brand (Bonding) or that they feel something

like love for the brand and would be upset without it (Love) buy and use that brand over 50%

of the time, compared with the 37% share of requirements for the average instant coffee brand.

Brand emotional attachment versus brand attitude

The traditional best predictor of Brand Purchase Volume (SOR%) is Brand Attitude

(called Overall Brand Rating in the study by Achenbaum, 1972, and see also Rossiter and

Eaglesons, 1994, reanalysis of the results of the ARFs Copy Research Validity Project, which

reveals Overall Brand Rating to be a strong and reliable predictor of brand purchase volume). To

test this traditional predictor against emotional attachment to the brand, a new variable called

Brand Emotional Attachment was constructed. This new variable was a weighted composite of

the four strongest attachment-like emotions, omitting Trust because it was such a weak predictor

(see Figure 2), and ordering the other four emotions in terms of their apparent strength of

attachment, thus Resonance = 1, Bonding = 2, Companionship = 3, and Love = 4. For a given

 10

brand, only the consumers strongest checked emotion was counted, so that scores on Brand

Emotional Attachment could range from 0 to 4.

Brand Emotional Attachment (EAb) and Brand Attitude (Ab) were then compared as

separate predictors of the brands Share-of- Requirements (SOR%). The predictive validity

coefficients were computed using polychoric correlations, a strength-of-relationship statistic that

is fair to the different-length ordinal rating scales (0 to 4 for EAb and 3 to +3 for Ab). The

results are shown in Table 1.

Table 1 about here

Brand Emotional Attachment proved to be a significantly better predictor overall than

Brand Attitude: rp = .67 for EAb versus rp = .55 for Ab (p < .05). By product category, Brand

Emotional Attachment predicts utilitarian gasoline brand purchase share much better than does

Brand Attitude (surprisingly, among male buyers) and also significantly better for the other

utilitarian product, laundry detergent (among women buyers). For the two more hedonic

products, beer and instant coffee, Brand Emotional Attachment predicts SOR% equally as well as

does the traditional predictor, Brand Attitude.

DISCUSSION

This is the first study to investigate the behavioral effects of specific brand emotions of

the attachment variety, starting with Trust and progressing to the very strong emotion of

(quasi-romantic) Love for the brand. The attachment-like emotions, Trust, Bonding, Resonance,
 11

Companionship, and Love were measured in the extreme zero-one, pick any manner favored by

practitioners in brand-tracking studies. This extreme measure records absolute, not relative, full-

strength levels of specific emotions, which partly accounts for their incidence (beyond Trust)

being so low. For the average brand bought by the consumer, Bonding is achieved by about 20%

of consumers and Resonance, Companionship, and Love by only about 4% of consumers.

Those consumers who become Bonded to the brand (its my brand), those who regard

it as a Companion, or those who admit that they feel something akin to Love for the brand,

reward the marketer with substantially greater purchase and usage of the brand (higher share-of-

requirements, SOR%) than if they merely develop a favorable, above-neutral Brand Attitude.

These emotionally attached consumers are therefore the brands most profitable customers,

especially considering that they would have no need of price promotions to keep buying the

brand.

The one specific attachment emotion, in addition to Trust, that doesnt work well

perhaps because it may apply only to major lifestyle products is Resonance (the feeling that

the brand fits my self-image).

Most interestingly, Brand Emotional Attachment predicts brand purchase volume better

than Brand Attitude for utilitarian products, such as gasoline and laundry detergent, perhaps

because these products are usually unemotional so that the rare emotionally regarded brand

stands out. The two approaches work equally well for hedonic products like beer and instant

coffee an unexpected finding because the EAb approach (emotional branding) might be

expected to be superior for these presumably more emotional products. Emotional branding

also seems to be just as effective among men as among women, contrary to the stereotype that

women are more receptive than men to emotional appeals, and especially to the subclass of
 12

appeal based on personal attachment. Fournier and Yaos (1997) widely cited brand attachment

results, for instance, were obtained with women consumers.

Future research is needed to determine whether advertising can be used to instill or at

least facilitate consumers emotional attachment to brands. It is not at all clear that consumers

can be told or urged via ads to regard the brand as their brand (Bonding), or to see it as a

Companion or a Love-object. Indirect modeling seems more likely to be effective, such as in

McDonalds Im lovin it campaign (see Rossiter and Bellman, 2005, for a favorable

assessment of this successful but initially professionally criticized campaign). Nevertheless, the

possibility remains that most consumers who become attached to brands do so solely as a result

of very rewarding usage experience and very little or not at all from advertising. Advertisers

therefore need a definitive answer from well-controlled field tests of their campaigns to know

whether emotional branding is really worth pursuing in place of the more traditional key-

benefit or USP approach which is aimed at the apparently easier goal of increasing the

consumers brand attitude.

 13

REFERENCES

ACHENBAUM, ALVIN A. Advertising Doesnt Manipulate Consumers. Journal of

Advertising Research, 12, 2 (1972): 3-13.

AXELROD, JOEL N. Attitude Measures That Predict Purchase. Journal of Advertising

Research, 8, 1 (1968): 3-17.

BALDINGER, ALLAN L., and JOEL RUBINSON. Brand Loyalty: The Link Between Attitude
and Behavior. Journal of Advertising Research, 36, 6 (1996): 22-34.

BARNARD, NEIL R., and ANDREW S.C. EHRENBERG. Robust Measures of Consumer
Brand Beliefs. Journal of Marketing Research, 27, 4 (1990): 477-484.

BERGKVIST, LARS, and JOHN R. ROSSITER. The Predictive Validity of Multiple-Item

Versus Single-Item Measures of the Same Constructs. Journal of Marketing Research, 44, 2
(2007): 175-84.

FOURNIER, SUSAN, and JULIE L. YAO. Reviving Brand Loyalty: A Reconceptulization of

Consumer-Brand Relationships. International Journal of Research in Marketing, 14, 4 (1997):
451-472.

HU, YE, LEONARD M. LODISH, ABBA M. KRIEGER, and BABAK HAYATI. An Update
of Real-World TV Advertising Effects. Journal of Advertising Research, 49, 2 (2009): 201-206.

MAYER, MARTIN. Madison Avenue U.S.A. London: The Bodley Head, 1958.

ROBERTS, KEVIN. Lovemarks: The Future Beyond Brands. New York: Powerhouse Books,
2004.
 14

ROSSITER, JOHN R., and STEVEN BELLMAN. Marketing Communications: Theory and
Applications. Sydney, Australia: Pearson Prentice Hall, 2005.

ROSSITER, JOHN R., and GEOFF EAGLESON. Conclusions from the ARFs Copy Research
Validity Project. Journal of Advertising Research, 34, 3 (1994): 19-32.
 15

Incidence (%)

80
80 77

Gasoline
69
70 Beer
Laundry detergent
59
60 Instant coffee

50

40

30
24
19 19
20

10 9 8
10 5
3 4 3 4 4 4
2 1 .5
0
Trust Bonding Resonance Companion Love

Figure 1 Percent of Consumers Who Report Trust, Bonding, Resonance,

Companionship, or Love for the Average Brand That They Buy in Each Product
Category
 Gasoline (Men) Beer (Men)
SOR (% ) SOR (% )
80 80

70 67
65 70

60 54 60

50 50
40
40 40 38
Average = 30 32
30 30
28 24
Average = 21 21
20 20
19
10 10

0 0
Trust Bonding Resonance Companion Love Trust Bonding Resonance Companion Love
 17

Laundry Detergent (Women) Instant Coffee (Women)

SOR (%) SOR (% )

80 80

70 70
60
58
60 54 60 55

50 50 46
41
40 40 Average = 37
Average = 31
30 30 33
27 29
20 24 20

10 10

0 0
Trust Bonding Resonance Companion Love Trust Bonding Resonance Companion Love

Figure 2 Share-of-Requirements (SOR%) Delivered by Trust, Bonding, Resonance, Companionship, and Love for
the Average Brand that the Consumer Buys in Each Product Category
TABLE 1
Predictive Validity Coefficients (Polychoric Correlations) for Brand Emotional
Attachment and Brand Attitude Predicting Share-of-Requirements (SOR%) for
Brands in Each Product Category and Overall

Laundry Instant
Predictor Gasoline Detergent Beer Coffee Overall
Variable (338 men) (606 (244 men) (468 (N = 1,656)
women) women)

Brand .63a .68a .64a .71a .67a

Emotional
Attachment

Brand .36b .58b .67a .64a .55a

Attitude

Notes: Different superscripts (a and b) mean that the correlations differ on a z-test at p < .05.
Overall results (italicized) are equally-weighted averages across the four product categories.