1. advertisingresearch
Emotional persuasion
Robert Heath, University of Bath, explains what emotional persuasion is, why it
matters and how it can be measured
HIS ARTICLE DEFINES two differ- But he also envisaged a defensive role Two definitions of persuasion
T ent types of persuasion: rational
and emotional. Rational persua-
sion, exemplified by performance
for repetitive advertising as ‘reinforcing
already developed repeat buying habits’.
Later, he developed this further to
The above suggests that Ehrenberg does
not see persuasion the same way as the
strong theory. The Oxford Compact Eng-
claims, promotions, offers and the like, address split-loyal purchasers (who reg- lish Dictionary’s general definition of
acts as an incentive for sales. But it is ularly purchase more than one brand), persuade is ‘Cause someone to believe,
emotional persuasion that creates rela- and defined a further role for advertis- convince’ (OCED 1996). This clearly
tionships and builds strong, successful ing as ‘nudging’ split-loyals towards a identifies persuasion as a rational, active
brands. An important new research sys- greater purchase proportion of one thinking activity, which involves the
tem is described – the CEP™ Test – brand or another (2). manipulation of thoughts to create
which is able to quantify accurately in Recent experimental work by beliefs and change attitudes.
advance how well ads will perform on Kathryn Braun (3) has confirmed the But this ‘active thinking’ is not the
each of these two different types of per- power of advertising as reinforcement only definition. The OCED also defines
suasion. in a post-purchase situation. Braun cre- persuasion as ‘to induce, lure, attract,
ated orange juice samples of varying entice’. This does not necessarily imply
Ehrenberg’s Reinforcement quality and gave them to subjects to that a verbal or rational process is need-
Model taste. Half the subjects were then ed for persuasion to take place, as the
In 1974, Andrew Ehrenberg wrote a con- exposed to advertising for the supposed words used (induce, lure, attract, entice)
troversial paper about how advertising new brand. It was found that the adver- all relate to feelings and emotions more
works (1). He rejected the notion that tising confounded the subject’s ability than thinking. Ehrenberg’s view of per-
advertising is capable of changing to judge accurately the quality of the suasion arising from advertising that
attitudes on its own, and proposed that juice, leading to substandard samples uses an emotional tone suggests it is this
it usually worked by reinforcing being highly rated. definition of persuasion he envisages. It
opinions formed from what are often The power of ‘reinforcement’ adver- is this definition of persuasion that I call
high levels of consumer knowledge and tising was contested by Jones (4), who ‘emotional’ persuasion.
experience. characterised it as a ‘weak’ theory of In modern practice, the word persua-
Ehrenberg’s attack was focused on advertising, which contrasted with the sion is used to encompass both
the general assumption that advertising traditional strongly persuasive model definitions, and is often used to describe
was a strong form of persuasion, and ‘universally believed in the United any activity that changes the attitudes
his theory gained much popularity States’. Four key differences emerged or behaviour of the recipient. But Ehren-
among advertising agencies. It was, between Jones’s ‘strong theory’ and berg sees reinforcement advertising as
bear in mind, a time when the sales Ehrenberg’s reinforcement model. influencing behaviour without neces-
effects of advertising were seen by many 1. Strong theory sees advertising as a sarily having to change attitudes. This
as longterm, hard to discern even in dynamic force, driving sales and catego- corresponds closely to the model that
hindsight and virtually impossible to ry growth. Reinforcement identifies an dominates academia in the US, Petty
predict. important additional defensive role, and Cacioppo’s Elaboration Likelihood
Ehrenberg had established that in especially for repetitive advertising. Model (ELM).
most markets there were few 100% loyal 2. Strong theory sees advertising oper-
buyers, and the majority bought more ating on an ‘apathetic and rather stupid Elaboration Likelihood Model
than one brand. He found that brand consumer’ (sic.); reinforcement sees con- The ELM divides consumers into those
users held consistently stronger atti- sumers as knowledgeable and who are ‘involved’ and those who are
tudes than non-users, but could not intelligent. not. Involved consumers tend to process
satisfactorily explain how these atti- 3. Strong theory sees advertising work- advertising using a higher level of
tudes came about. This led him to ing by changing attitudes, which leads thoughtfulness, which they term ‘cen-
question the core assumption within to changing behaviour. Reinforcement tral’ processing. Uninvolved consumers
hierarchy-of-effects models: that atti- rejects the idea that attitude change use a lower level of thoughtfulness –
tude change precedes and drives must always precede purchase. ‘peripheral’ processing. The key differ-
behaviour change. He accepted that 4. Reinforcement sees persuasion as ence between the two is ‘the extent to
advertising can create, re-awaken or arising from advertising that takes ‘an which the attitude change that results …
strengthen brand awareness, and can be emotional instead of an informative is due to active thinking’ (5). Attitude
one factor that facilitates trial purchase. tone’. changes resulting from central process-
46 Admap • July/August 2006 © World Advertising Research Center 2006

2. Robert Heath is a visiting
professor at Copenhagen
Business School, and
lectures at Bath School of
Management, where he is in
the final year of a PhD. He
also runs the Value Creation
Company, a brand
consultancy that specialises
in low attention processing.
ing are enduring, so it is a strongly per- ing has been made by Bornstein tionship’ metacommunication is often
suasive route. But an important (11), who discovered that affect is more subtle and disguised, but it is this part
characteristic of peripheral processing is effective when it is processed subcon- that endures and ultimately is most
that the attitude changes that result are sciously: Bornstein found that conscious effective at changing attitudes. It is this
weak and relatively transient, and the processing of affective elements weak- relationship-building metacommunica-
peripheral route also is one in which ens their potency, because it allows the tion that is emotionally persuasive.
repetition of emotive cue-based advertis- subject to evaluate rationally and count- If you think about when you meet
ing is more influential than the actual er-argue against the influence. someone, you’ll realise that you might
message. So Ehrenberg’s reinforcement This implies that the less attention be influenced to meet them again by
model in fact corresponds very closely to consumers pay to affective elements in what they say, but you are influenced to
the less strongly persuasive peripheral advertising, the better they will work. become friendly towards them by the
processing. This might seem to confirm Christie Norheilm (12) has confirmed way they say things. Extending this
Jones’s opinion that it is a ‘weak’ model, this experimentally. She has found that analogy to marketing, brands can easily
but recent findings show otherwise. if ads are processed deeply, repeated get sales using the content of their adver-
exposure causes affective response to tising – by demonstrating added value,
Decision-making and emotion first rise and then fall sharply. But when cutting price, improving performance,
Traditional models suggest that behav- ads are processed in a shallow, inatten- and so on. But brands build enduring
iour change is driven by changes in tive fashion, affective responses are relationships and create loyal consumers
attitudes. Early models like Lavidge and enhanced, with no downturn from repe- only by the ‘relationship-building’ meta-
Steiner’s (6) had decision-making driven tition. communication in their advertising. For
by affect (feelings and emotions), but There is also experimental confirma- example, Colgate didn’t become a super-
affect operated only as a consequence of tion that repetition at low attention has brand just by preventing tooth decay like
cognition (thinking). Zajonc (7) success- an effect on decision-making. D’Sousa every other toothpaste. It became a
fully contradicted this in 1980, showing (13) found evidence of ‘small but signifi- superbrand because, through years of
that affect is generally pre-cognitive, not cant’ increases in brand awareness and advertising, it has built up a relationship
post-cognitive. More recently, Damasio brand choice arising from repetition of with people, so that they now trust and
(8) has shown that cognition is ‘hard- radio ads in a divided-attention situa- like it as a brand. This is exactly how
wired’ via the emotions, and that tion. All this supports the Low- brands like Andrex, Olay, Persil, Stella
feelings are therefore capable of driving Attention Processing (LAP) Model (14), Artois, Orange, BMW and many others,
decisions in the face of negative cogni- which suggests that advertising that have become so strong.
tion. This has since been validated by operates emotionally can be processed
Shiv & Fedhorikhin (9): by constraining without active attention and can exert a Measuring rational and
decision time they found that subjects significant influence on choice, often emotional persuasion
chose chocolate cake in place of fruit without the consumer realising it. Part of the problem with emotional per-
salad, ignoring the sensible guidance of Whether you call it reinforcement, suasion is that it is really hard to
their ‘thinking’ brain, and giving way to peripheral processing or LAP, advertis- measure. It is relatively easy to measure
their emotions – exactly how busy par- ing that works in this way is not weak, it rational persuasion – you can pretty
ents act when shopping for groceries is simply emotionally persuasive. much just ask people if they feel more
with their children. What this suggests inclined to buy the brand. But if you ask
is that real-life decisions are very vulner- Emotional persuasion them if they feel inclined to form a rela-
able to advertising that operates The true importance of emotional per- tionship with the brand they are likely
emotionally. suasion emerges from findings by Paul to think you are nuts.
Damasio (10) also found that, while Watzlawick (15). Watzlawick identifies Working in partnership with OTX,
cognitive processing depends on work- two distinct levels for communication: a we have devised a research system that
ing memory and is enhanced by content level and a relationship level. solves this problem: the CEP™ Test
attention, affective processing is inde- The former he terms communication, (patent pending). The CEP™ Test uses a
pendent of working memory and the latter metacommunication. Rational set of ten dimensions to measure the
attention. He established that emotions persuasion takes place in the ‘content’ Cognitive Power™ and the Emotive
and feelings are formed subconsciously area of communication, is easily Power™ of an advertising execution.
and autonomically (independent of analysed and classified, but is the fastest Cognitive Power measures the rational
will). But an even more important find- to fade in memory. In contrast, the ‘rela- persuasion of the advertising – how
© World Advertising Research Center 2006 July/August 2006 • Admap 47

3. advertisingresearch
FIGURE 1 Bottom right is a cosmetic ad. It gets over
CEP TM
test results a lot of information but it is almost entirely
lacking in empathy. This advertising isn’t
100 going to score any relationship-building
points with the consumers. In contrast, the
80 star award for relationship-building goes to
one of last year’s British Airways ads, top
60
BA Andrex
left. But what would happen if they decid-
Honda Disney ed to abandon their wonderfully soothing
40 World
all-important operatic music track?We
Emotive power
Sony
Stella could tell them! ■
Artois 20
Guinness 1. ASC Ehrenberg: Repetitive advertising
–100 –80 –60 –40 –20 20 40 60 80 100 and the consumer, JAR 14, April 1974,
BP pp. 25-34.
–20
2. N Barnard & ASC Ehrenberg:
Burger
chain Advertising: strongly persuasive or
–40
nudging?, JAR 37, 1, 1997.
3. KA Braun: Postexperience advertising
–60
Cosmetic effects on consumer memory, Journal of
Consumer Research 25, 4 1999.
–80
4. JP Jones: Advertising: strong force or weak
force? Two views an ocean apart, Inter-
–100
national Journal of Advertising 9, 3, 1990.
Cognitive power
5. RE Petty & JT Cacioppo: Attitudes and
Persuasion: Classic and Contemporary
well it will achieve sales. Emotive Power In the UK we tested a corporate ad for Approaches. Boulder, CO: Westview Press,
measures the emotional persuasion of BP. It spoke a lot about global warming 1996.
6. RJ Lavidge & GA Steiner: A model for
the advertising – how well it will build and what BP was doing to stop it, and
predictive measurements of advertising
the brand relationship. scored very well on Cognitive Power, but
effectiveness, Journal of Marketing 25, 4,
In the first six weeks of the launch we the Emotive Power score was below aver-
1961.
performed approaching 120 tests of age. This indicates the ad won’t make
7. RB Zajonc: Feeling and thinking:
brand communication material. We consumers like BP any better. This result
preferences need no inferences. American
have found that the system works not was reflected in very small increases in
Psychologist 35, 1980.
only on TV, print, poster, radio and cine- favourability, mostly among existing 8. AR Damasio: Descartes' Error. New York,
ma, but on internet ads, promotional users. Sony’s spectacular release of thou- NY: GP Putnam's Sons, 1994.
banners – in fact, any form of brand sands of coloured balls in San Francisco, 9. B Shiv & A Fedorikhan: Heart and mind in
communication. What is more, we have for the launch of their Bravia TV, caused conflict: the interplay of affect and
found that high scores on cognitive much excitement amongst creatives. It cognition in consumer decision making,
power or emotive power are validated in doesn’t explain why the Bravia is better, Journal of Consumer Research 26, 1999.
90% of cases by positive shifts in but it is emotionally persuasive, as we 10. A Damasio: The Feeling of What Happens.
favourability between those who have can see from the very high score it London: Heinemann, 2000.
and have not seen the advertising. But achieves on Emotive Power. What’s 11. RF Bornstein: Exposure and affect:
low scores show no shift. Some results more, there is a substantial upward shift overview and meta-analysis of research,
are shown in Figure 1. in favourability. 1968-1987, Psychological Bulletin 106, 2,
Take as an example six of these adver- What about Guinness’ recent ‘evolu- 1989.
tisements we tested during our tion’ ad? Very creative, but only average 12. CL Nordheilm: The influence of level of
development and validation phase. In on Emotive Power. Non-users rated it processing on advertising repetition effects,
the US, we tested a highly creative and below average on both scales, users Journal of Consumer Research 29,
quite well-liked ad for a burger retail rated it above. And when we looked at December 2002.
chain. But the scores for Cognitive the favourability shifts there was a big 13. G D'Sousa: Can repeating an
Power and Emotive Power were way shift amongst users and no shift at all advertisement more frequently than the
below average. And the advertising among non-users. So the message for competition affect brand preference in a
showed no shift at all on favourability Guinness is that if they want to rein- mature market?, Journal of Marketing 59,
between those who recognised and did force their user base, this ad works fine. 2, 1995.
not recognise it. In contrast, Disney The Honda Diesel ad also created a 14. RG Heath: The Hidden Power of
World’s 50th Anniversary ad scored stir, two minutes long and totally icono- Advertising. Admap Monograph No. 7,
well above average on Cognitive Power clastic. And it creates a stir in Emotive World Advertising Research Center, 2001.
and Emotive Power, in fact it was the Power as well, nearly as high as Andrex. 15. P Watzlawick, JB Bavelas & DD Jackson:
highest-scoring of all ads we tested. And But that’s not a recent Andrex ad, it is Pragmatics of Human Communication.
it also scored the highest shift in the very first Andrex Puppy ad ever New York, NY: Norton & Co. Inc, 1967.
favourability, among both users and made. It is still as emotively persuasive
non-users. as it was 20 years ago. robert.heath@value-creation.co.uk
48 Admap • July/August 2006 © World Advertising Research Center 2006