he

Fundamental Templates

of

Quality

Ads

Jacob Goldenberg * David Mazursky * Sorin Solomon

The JerusalemSchool of Business Administration,The Hebrew University of Jerusalem,Mt. Scopus,
Jerusalem91905, Israel
msgolden@mscc.huji.ac.il* msmazur@mscc.huji.ac.il* sorin@vms.huji.ac.il

Abstract
Creative ideation is a highly complex process, which is difficult to formalize and control. Evidently, even in a complex
thinking context certain patterns of creativity may emerge.
Relying on such observed patterns may help in "organizing"
the creative process by promoting routes that have been
proven to lead to productive ideas and avoiding those that
do not.
The present research suggests that successful advertisements share and are characterized by such abstract patterns
termed creativity templates. The theoretical rationale for the
emergence of such templates and the empirical studies that
detect the key creativity templates underlying quality ads
indicate that the templates are identifiable, objectively verifiable, and generalizable across multiple categories. Studies
1 and 2 were designed to identify and describe the templates.
Six major creativity templates were derived by inference
from a sample of 200 highly evaluated print ads drawn from
award-winning ad contests such as The One Show (Study 1).
Judges found that 89% of the ads could be explained by the
six creativity templates. Following a formal description of the
templates and their versions, a study comparing 200 awardwinning and 200 nonwinning ads (Study 2) is reported. It
was found that the two groups differed systematically in the
number and distribution of creativity templates: 50% of the
award-winning ads as opposed to only 2.5% of the nonwinning ads could be explained by the templates. Further validation of the template approach was obtained by manipulating presence or absence of templates in an experimental
setting. In Study 3 groups of individuals were trained in
template-based idea generation, an association technique, or
not trained at all, prior to an ad-ideation task. Another group
subsequently rated the ideas. Findings indicate that a priori
knowledge of the templates was associated with the generation of higher quality ads in terms of creativity, brand attitude judgments, and recall (Study 4), with some variation in
terms of feeling responses which included humor, emotion,
and annoyance.
The findings of the reported studies and several real-life
applications conducted in leading ad agencies, indicate that
the template taxonomy is a trainable, resource-saving, and
effective tool. It simplifies and improves the decision-making

0732-2399/99/1803/0333/$05.00
1526-548X electronic ISSN

process involved in designing advertising strategies. It can

be applied either by hiring trained personnel employed by
consulting firms, or by training the agency's own personnel
to routinely evaluate past and current ads, and engage in
creative activity.
The template approach represents a step forward in defining a comprehensive model of the antecedents of outcome
reactions to advertising stimuli. Improved understanding of
the wide spectrum of reactions connecting the basic templates with end-user reactions is likely to be beneficial both
for academicians and for practitioners. Such a framework can
serve as a basis for a synthesis between the activity of creative professionals whose focal interest is the generation of
ads, managers, whose main responsibility is strategy formulation, and academic activity, which focuses mainly on
the consumer reaction-end of the advertising process. Hence,
in addition to academicians, the relevant target audience is
likely to include a wide array of communication-related personnel such as creative professionals and planners in advertising agencies, consultants, and brand managers.
In addition, it is postulated that the template taxonomy
provides the means to achieve "creativity expertise". Unlike
the divergent thinking approaches, in which the required expertise is not necessarily related to the creativity process itself (e.g. individuals can be trained to be better moderators
in brainstorming), the creativity template approach is trainable and has the capacity to measure and directly improve
creativity outcomes. The template taxonomy facilitates the
focused cognitive effort involved in generating new ideas,
the capacity to access relevant information, and enables high
memorability of the reduced set of information needed to
perform the tasks.
The fact that templates are less transient than the ideas
produced does not mean that templates are permanent or
that they are insensitive to changes over long term frameworks. Indeed, advertising reflects social norms and trends,
and as such, long term social trends are expected to reshape
the templates and provide conditions for the evolution of
new templates. Nonetheless, the dynamics of template
changes are expected to be much slower than the dynamics
of changes in ad hoc idea generation.
(Advertising Creativity;Advertising Strategy;Creativity in Marketing;MarketingIdeationProcesses)

MARKETINGSCIENCE? 1999 INFORMS

Vol. 18, No. 3, 1999, pp. 333-351

THEFUNDAMENTALTEMPLATES
OF QUALITYADS

Introduction
Creativityin advertising frequentlyinvolves methods
that encourage the generationof a large number of ad
concepts (Batra,et al. 1996)on the assumptionthat the
rewardsof producinga large numberof ideas will outweigh the costs (Winston1990).The generationof new
ideas in this manner tends to be highly unformalized
and unsystematic. Often, such methods are based on
the divergentthinkingapproach(e.g., focus groups,free
association, and other projective techniques; see
O'Guinn,et al. 1998)whereby judgment is suspended
and ideas emerge by associative thinking in a "limitation free" environment.
However, even in a divergent thinking context certain patternsof creativitymay emerge. Creativeteams
often seek ways to become more productive as they
progress from one creativitytask to another. Common
patternsrelevant to different domains are sometimes
identified (cf. Boden 1992; Dasgupta 1994; Weisberg
1992). These may then be applied on an ad hoc basis
within a given advertisingcontext,or even transported
to other contexts.Such patternswill be more stableand
less transientthan the abundanceof randomideas that
emerge in the process of associative thinking. They
may also help in "organizing"the creativeprocess by
promotingroutes that have been proven to lead to productive ideas and avoiding those that do not. Nonetheless, even if they prove productive, such patterns
tend to be idiosyncraticand they are often not verbally
definable (Weisberg1992). As such, they are likely to
lack permanenceand generality.In the present paper,
it is posited that certain patterns are identifiable,objectivelyverifiable,and generalizableacrosscategories.
It is suggested that these patterns, termed creativity
templates,underlie the generation of quality ads because they facilitate focused creativity, and lead to
more effective outcomes.
Letus portraythese notions with an example.Figure
la shows an ad for the 1989 French Open Tennis
Championship sponsored by Penn (The One Show,
1989). The ad features a croissant-shapedtennis ball
or, viewed differently,a croissantwith a (Penn)tennisball surface.The patternof this ad served in generating
an ad featuring a hockey puck shaped tennis ball for
the Canadian Open (Figure lb, The One Show, 1991)
and a moisturering of a tea cup for Penn'ssponsorship

334

Figure la

Ad for the FrenchOpen Tennis Championship

..

...

cf-

.........................

.
.

...

.
*-:j::-:-:;
:

.i.....:

....

S...~~~~~~~~~~~~~~~~~~~~~~~~~~~~............
--:::::-Xo
.~~.
.~~~~~~~~~
- ~~
. ....:...'.'...
.";.'::..
,":...
i

:*
........
. .. . ........

!i:i
..* ..........

.......

.. .:

. .. .... .

,...

:.. . -..x,j;.-.,.inl.:
?.; .........
:

!!:'"
ii.i .. .......:':.:}.

s-.

..

......~................
.......--........
.,--.-.. ..
.. .. ;;=..
: .. . ..

. . ....

. ..

.. .. .::^.--;

(F:IIiguIIIIKr

of theBritishLipton Championships
ic The..,'S.
.
*~~~~~~~~~~~~~~~~~~~~~~~~.
.
.............iX
..........gw
ii
isamodified

tenms

designed

ball
con-.................

to

American

_-3,"

f.ag

thestars.as

withY

tennis

z:

symboliZe.a. .

, .,w

eS.

b.:

..

f::'
:r:::::..
'-x
......................................
-:sC
.....
-nE:
............
:.,
:w
.......... ... ...... .:.

ballsfreplacing

]_

.
: i Uiz
........
_E;
is~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.................
... i;

~~~~~~~~~~~~~~~~~~~~~~~~~~~..
"~~~~~~~~~~~~~~~~~~~~~~~
ggEb,,~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
~~~~~~~~~~~~~~~.
.. ...,:

ood2g::s: .:..'"

,X'

.......... . ..........
.......

d8,

SSj

?',

..

.. . ...;a
ee,

...... .........
.. '"''j
..''''!

i~~~~~~~~~.........

,W.;:X.O,

W.X

.'

fff
w1-U
.;...........
~~~~~~~~~~~~~~~~..

Identification

of

the

Creativity

Templates

*.......i:
.
.-;.
:- .....
i;~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~..
try
Conssetwit
th rgnl ad an it prnil
..-...
i..

.?

pattern th US

;.'....!..i.

pe Tenni Chmponhi

might

of the British Lipton Championships (Figure lc, The

One Show, 1992).The common patternin all three ads

is a modified tennis ball designed to symbolize a country. Consistent with the original ad and its principle
pattern, the U.S. Open Tennis Champiornshipmight
consider using a hamburger-shapedtennis ball or the
American flag with tennis balls replacingthe stars, as
ad ideas.
Identification of the Creativity Templates

The common pattern of the tennis ball ads can be depicted schematically. From analysis of the French tennis tournament ad (Figure.1a), a scheme depicting the
possible links between the tennis tournament and
France can be constructed. Figure 2a provides the
breakdown of the tennis tournament (left-hand s'ide)
into some of its internal components such as a player,

court, or a ball. The message theme in this ad (righthand side) is the location which in the case of France,
can be represented by various symbols such as the
MARKETINGSCIENCE/VOl. 18, No. 3, 1999

GOLDENBERG, MAZURSKY, AND SOLOMON

The FundamentalTemplatesof Quality Ads

Ad for the CanadianOpen

Figure lb

the FrenchOpenTennisChamSpecificSchemeUnderlying
pionshipAd

Figure2a

.....................
.p
,'

Tennis"

Location?-

'tournament

.... ...........''1
l
.....
.
:... ........
.....,....

France

..,:
* ... .....

Symbols set:
Louvre
Eiffel Tower
Croissant

Product space
......

... ... ......

...

.. . . ... ..... .. .

. ...........--,.
.......

Cour
Player

a croissant-shapedball

Ball

Figure lc

Ad for the BritishLiptonChampionship

Louvre, the Eiffel Tower, or a croissant. The advertised

event and the main theme (France as location) are then
unified by matching their shapes. In a similar manner,
schemes can be constructed for the ads that stress the
importance of other locations. Note, however, that
these schemes depict specific combinations of events
and locations.
Generalization of this operation may be achieved by
inferring a general scheme. A scheme can be considered general only to the extent that it can be widely
applied in a variety of products, events, and messages.
The repetitive appearance of a scheme in different domains reveals a creativity template. Thus, the transformation from a specific scheme (Figure 2a) to a general
scheme (Figure 2b) extends the notion of common patterns to the notion of creativity templates. The general
scheme, shown in Figure 2b, consists of two parts: The
first part, denoted as the productspace, is formed by the
internal components of the product and the objects that
interact with it (P1, P2, P3 in Figure 2b). The tennis ball
is a major internal component of the set of components

of
Version
the Replacement
Schemeunderlying
General
Template
thePictorial
Analogy

Figure2b
Product

_Message

Symbolsset

Product space

SymbolI Symbol 2 Symbol3

Pi
P2
P3

MARKETING SCIENCE/VOl.

18, No. 3, 1999

operator
bya linking
Matching

335

GOLDENBERG, MAZURSKY, AND SOLOMON

The FundamentalTemplatesof Quality Ads

that featurein a tennis championship.The second part,

denoted as the symbolsset, is formed by the symbols
that feature in the consumer's representationof the
message. In the tennis tournamentexamples,the croissant (as a symbol of France)was chosen in the creation
of one ad, hockey puck (as a symbol of Canada) was
chosen in another, and finally, the moisture ring of a
tea cup (as a symbol of the UK)was selected for a third
ad.
The elements chosen from the two parts of the general scheme (the product space and the symbols set)
are then unified through a linking operatorwhich
matches their shape, color, or sound. Note that the
product space and symbols set andthe specificationof
their matching mechanism can be tracedin additional
domains. Although at first glance it may appear remote, the ad for Nike-Air sneakers, shown in Figure
3a, has the same fundamental scheme as the tennis

3a
Figure

Version
ofthe
AnExample
fortheReplacement
Nike-Air:
USA
andKennedy,
(Wieden
Template
Analogy
Pictorial
withpermission.)
1995.Reprinted
4 IV

tournamentads. Figure 3b depicts the scheme underlying the Nike-Air ad. This scheme termedReplacement
is a version of a PictorialAnalogytemplate. Replacement is obtainedwhen a product (e.g., sneaker)or one
of its parameters,replacesa symbol consistentwith the
meaning of the conveyed message (e.g., firemansheet).
Conceptually, the Nike-Air application is more abstract than the tennis tournamentapplications which
involved simple duplicationof a common pattern.The
general scheme no longer involves identical information, nor does it necessarilyinvolve the same product.
Yet, it is identifiable,objectivelyverifiable,and generalizable across differentads. As such, it is defined as a
creativitytemplate.
The Theoretical Rationale for the Creativity
Template Approach
The creativitytemplate approachcontends that a substantialpart of creativebehavior is guided by abstract
fundamental schemes. In some instances creativity
teams may define explicit ideation rules that are consistent with templates, although in many other instances consistency with templates may be implicit.
Even when the creative execution process involves an
unstructuredidea generationcontext,many ideas will
be definable in terms of creativity templates. These
templatesserve as paths that the self-organizedsystem
tends to follow (Kelso 1997) when new ideas are
formed. In the context of ideation SparsenessTheory
(e.g. Minsky 1985) holds that almost every evolutionary search for ideas is likely to yield certaincommon
themes.
Attempts to identify relational structures in other
domains have produced several frameworksthat are
conceptually analogous to the template approach.

Figure3b

Ad
the Nike-Air
SpecificSchemeUnderlying

\ Nike Air

r~~~~~~~~~M
_r

_0_%*6

Sidewalk

336

..

<

Protection

Productspace

softbetween
The Air Essential.Something
youandthepavement.
NIKE-.
FROM
INTROD'UCING
TWONEWWAILNGSHOCES
NIKEWTH
THEYRE
VERY
SAFEPIACES
TOLAND.
IN THEHEEL,
AIR'CUSHION[ING

'

Foot
Sneaker

SymbolsSe:
FIreman-shet Air-bag
Spings

a sneaker-shapedfireman-sheet

MARKETINGSCIENCE/VOl. 18, No. 3, 1999

GOLDENBERG, MAZURSKY, AND SOLOMON

The FundamentalTemplatesof Quality Ads

Such structures have been developed in disciplines

such as Linguistics (Eco 1986; Chomsky 1978), Anthropology (Levi-Strauss 1974), Random Graphics (Palmer
1985), Venture and Transitional Management
(Kauffman 1995), Psychology (Simon 1966), and Artificial Intelligence (Minsky 1988). However, the background, schemes, and implications of the structures
developed in these areas are essentially different. For
example, creativity may not be as central as it is to
advertising.
In light of the central role of creativity in advertising
the emergence of templates is expected since they ensure the balance between surprise and regularity. Regarding surprise, Hayes (1978, p. 232) noted that an act
is perceived as creative when ". . . most people could
not or would not have arrived at the same solution".
However, Simon (1966) stressed the importance of regularity in the creativity process and the need to form
a solution plan to direct their effort because people
seem to get lost easily in executing their solution plans.
In the sender-receiver communication process surprisingness is indeed useful but not at the expense of altering the overall intended meaning of the message. If
some regularity in the idea generation process is identified then creative ideas may be evoked while ensuring that their overall meaning is preserved.
Findings in the area of cognitive psychology provide
support to the conclusion that the detection and use of
templates may even result in enhanced surprisingness.
For example, according to Perkins (1981), adherence to
a cognitive frame of reference involves sensitivity to
the "rules of the game" and by functioning within a
frame, a better position is achieved to notice or recognize the unexpected. Finke et al. (1992) noted that
restricting the ways in which creative cognitions are
interpreted forces people to think about conceptual implications in more atypical ways, which promote creative discovery.
Indeed, the concept of structured creativity is already embedded in a number of current techniques
such as morphological analysis (e.g., Urban and
Hauser 1993 in the context of new product development; note also the HIT procedure, Tauber 1972 in that
context). This method is akin to the notion of creativity
templates in that it calls for identifying the parameters
of the problem, listing all the possible combinations of

MARKETING SCIENCE/VOl.

18, No. 3, 1999

parameters, examining their feasibility, and deciding

on the best alternative. However, morphological analysis does not provide specific and generalizable guidelines on how to combine these parameters. A step toward providing structured guidelines was introduced
by Altshuler (1986) in his attempt to uncover latent
logical patterns underlying creative ideas. By a backward analysis of problem-solution relationships, he
succeeded in identifying a number of such patterns
which he labeled "standards". These standards represent common phenomenological patterns. The notion
of creativity templates extends the view of common
patterns by allowing them to be more abstract and
hence more widely applicable across ads for different
products and services.
Some other approaches, such as Resonanceand those
involving the Janusian concepts, are more directly tailored to the context of advertising creativity. The resonance approach involves dual or multiple meanings
surrounding a single word or phrase, such as "ForgetMe-Knots" in an ad showing men's ties arranged to
form a floral bouquet (see Mcquarrie and Mick 1992).
The Janusian approach involves "the capacity to conceive and utilize two or more contradictory concepts,
ideas or images simultaneously" (Rothenberg 1971, p.
195); for instance, Blasko and Mokwa (1986) cite:
"We're first because we last" and "We've got the inside
of outside protection." Although, unlike the creativity
templates described in this paper, these methods do
not lend themselves to schematization, they do provide important rules for creativity. An interesting step
toward generating a broader taxonomy of figuration
modes was presented recently by Mcquarrie and Mick
(1996). The rhetorical perspective proposed by them
contends that the manner in which a statement is expressed may be more influential than its content. Finally, it is interesting to note a specific area of advertising, namely, humor, in which certain dimensions of
ads were found as successful predictors of humorous
ads. Alden et al. (1993), using Raskin's psycholinguistic theory of humor to explain why certain ads are perceived as more humorous than others, found that ads
that employed a contrast between everyday life and
the unexpected were generally perceived as more humorous than those employing a contrast between everyday life and the impossible. By extending these approaches, that typically emerged from a specific

337

GOLDENBERG, MAZURSKY, AND SOLOMON

The FundamentalTemplatesof Quality Ads

pattern of creative execution, the creativity template

taxonomy provides fundamental generalizable structures for the generation of quality ads.
In reviewing the relevant research, the distinct contribution of several well-known taxonomies of advertising strategies, which focus on other perspectives of
ad generation, should be assessed vis-a-vis the creativity template approach. For example, Simon (1971) proposed a framework which includes ad strategies such
as information, argument, repeated assertion, command, and symbolic association. Similarly, Burke et al.
(1990) proposed a taxonomy in which positioning (defined as the featured benefits and the distinctiveness
of a brand relative to other brands), and message emotion (which pertains to the emotional tone of the ad),
are key ad strategies. The main distinction between
these frameworks and the creativity template taxonomy lies in the level of the cognitive representation of
the framework factors. The advertising strategies represent summative factors and intended consequences
(e.g. emotional response), and the creativity templates
represent the schemes that antecedeand give rise to these
strategies.For instance, a specific well-defined template
may evoke an emotional response, but the emotion itself does not offer the scheme nor the means to elicit
this response. Thus, the aforementioned advertising
strategy frameworks focus on the decision between
different consequences (e.g., emotion, positioning); in
contrast, the creativity template approach focuses on
the cognitive activities that lead to these consequences.
In the present research, the notion of creativity templates in advertising is conceptualized, formulated,
and examined empirically. Studies 1 and 2 were designed to identify and describe the templates. We begin by deriving six creativity templates from a sample
of 200 highly evaluated ads and examining their distribution in this sample (Study 1). Then, a formal description of these templates is provided. Next, a study
comparing 200 award-winning and 200 nonwinning
ads, is reported (Study 2). The purpose of Study 2 was
to examine whether templates appear uniquely in high
quality ads. It was found that templates appear substantially more frequently in the award-winning ads
than in the nonwinning ads. Further validation of the
template approach was obtained by manipulating the

338

presence or absence of the templates in an experimental setting. In Study 3 individuals were trained in
template-based idea generation, in an association technique, or not trained at all, prior to an ad ideation task.
Another group of individuals, blind to the training
procedure and hypotheses, subsequently rated the
ideas. Findings indicate that a priori knowledge of the
templates was associated with higher quality ads in
terms of creativity, brand attitude judgments, and recall (which is examined in Study 4), with some variation in their capacity to trigger feeling responses.

Study 1: Detection of Creativity

Templates
Objective. The objective of the first study was to
identify general creativity templates of ads among
high quality ads and to examine the frequency and
distribution of these templates.
Ads and Screening Procedure. The awardwinning ads and contest finalists of NY, The One
Show, and USADREVIEW, all for the years 1990-1995,
served as a pool of highly evaluated ads. A set of 500
ads was selected using convenience sampling from this
pool and presented in a random order to three senior
creative experts (all of whom had at least 12 years of
experience in the advertising field) who were asked to
select the highest quality ads from that set. The three
experts performed the selection task individually and
independently. The criterion of quality was in accordance with that which is used by judges of The One
Show contest combining newness and significance.
The instruction given to each expert was to sort out the
highest quality ads and subsequently, order the pile
containing the highest quality ads from highest to lowest. Upon completion, the piles containing the highest
quality ads chosen by the experts were compared. The
interjudge agreement rate measured by the overlap in
the "best" ads across the sort piles was 90% and disagreements were resolved by discussion. The 200 top
ads served as the sample for subsequent analysis.
Inferring Templates. The search for templates was
conducted by inferring the linking operatorsin each ad
and identifying the relevant sets and their spaces, in a
manner consistent with the inference of the pictorial

MARKETING SCIENCE/VOl.

18, No. 3, 1999

GOLDENBERG, MAZURSKY, AND SOLOMON

The FundamentalTemplatesof Quality Ads

analogy template described earlier. For example, an ad

showing a lady barking at burglars and scaring them
away, in an ad for a security lock (see Table 1), led to
the identification of a template version termed absurd
alternativesby the following procedure. First the product (lock) and the message theme (safety) were identified. Then the linking operator was inferred-the
lady voice was replacedby dog barks. An unanswered
question was, what elements are linked by the linking
operator? In one end of this link, the lady reading a
book threatened by a burglar is a situation which provokes the need for safety (the message). In the other
end, the dog could serve as an alternative to a lock in
enhancing safety. Other options, serving the same
function (e.g., security guards, alarm systems) are also
available. Accordingly, the lady represents an element
in a situation set whereas the dog and the other options
represent elementsin an alternativeoptionset. The absurd
alternatives template version of the extreme situation
template is obtained by the unrealistic nature of the
solution created by the linking operator.
The repetitive appearance of this scheme in various
ads and regarding different products and messages,
defines a general scheme, or template. A total of six
key templates and their 16 versions were identified. A
detailed description and formulation of the six templates is given in Table 1 using one example for each
template. The following templates and their versions
were identified (see Figure 4 for overall structure):
I. The Pictorial Analogy Template.The pictorial analogy template portrays situations in which a symbol is
introduced into the product space, as discussed in detail in the Introduction. This template has two versions:
The replacementversion and the extremeanalogy version.
In the extreme analogy version the symbol is taken to
the extreme whereas in the replacement version it is
merely transplanted.
II. The Extreme Situation Template.The extreme situation template represents situations that are unrealistic
in order to enhance the prominence of key attributes
of a product or service. This category includes three
versions: The absurd alternative version, the extremeattribute version, and the extreme worth version. The extreme situation template is exemplified and described
in Table 1 using the absurd alternative version. The

MARKETING SCIENCE/VOl.

18, No. 3, 1999

extremeattributeand extremeworth versions portray situations in which either the attribute or the worth of a
product (or service) is exaggerated to unrealistic proportions (e.g., a jeep driving underneath the snow to
demonstrate its all-weather driving capacity: Cliff
Freeman & Partners, NY, 1994).
III. The Consequences Template. The consequences
template indicates the implications of either executing
or failing to execute the recommendation advocated in
the ad. There are two versions of this template: The
extreme consequences version (exemplified and described in detail in Table 1) and the inverted consequences version. The inverted consequences version
warns against the implications of not executing the recommendation of the ad (e.g., an ad promoting a brand
of vitamin showing an otherwise highly energetic person, unable to get out of bed in the morning).
IV. The Competition Template.The competition template portrays situations in which the product is subjected to competition with another product or event
from a different class. The selection of the other product or event is guided by its expected superiority over
the advertised product, for example: (1) a race between
an advertised car and a bullet (Della Femina
Travisanod & Partners, Los Angeles 1989), or (2) a person contemplating whether to continue eating the (advertised) cereal or to answer a ringing phone. There
are three versions of the competition template: The attribute in competition version, the worth in competition
version, and the uncommonuse version. The difference
between the first two versions relates to whether the
competition pertains to a product attribute or whether
it challenges the worth of the product. The competition
template is exemplified in Table 1 by the uncommon
use version.
V. The Interactive ExperimentTemplate.1The interactive experiment template induces realization of the
benefits of the product by requiring the viewer to engage in an interactive experience with the medium in
which the ad appears. This can be achieved either by
actually engaging in an experiment (the activation ver'The notion of interactive experiment is different from the notion of
"demonstrations" often used by copywriters, in that demonstrations
despite their function in enhancing involvement do not involve
physical action in the manner described here.

339

GOLDENBERG, MAZURSKY, AND SOLOMON

The FundamentalTemplatesof Quality Ads

Table 1

Examples, Descriptions, and Formulationof Template Versions

I. The PictorialAnalogyTemplate:
The ReplacementVersion

Example

Description

Examplesand detailedformulationof the scheme

underlyingthe replacementversion of the pictorial
analogytemplate(composed of a symbols set, a
productspace, and a linkingoperator)are presentedin
the Introductionelaboratingthe notion of creativity
templates (see Figures1-3).
Note descriptionin the Introduction.

Formulation

Note formulationin the Introduction.

Scheme

Note scheme in the Introduction.

II.The ExtremeSituationTemplate:
The AbsurdAlternativeVersion
The commercialfor locks showing an old lady scaringaway burglarsby
barkingat them (Suissa MillerAdvertisingCompany,USA,1993,
Cannescontest award)conveys the message that a safe and peaceful
evening can be achievedeitherby buyinga certainlock or by barking.

The idea of this version is to presenta tongue-in-cheeksuggestion to

the viewer:"Youdon't have to buy our product.Thereare alternative
options for achievingthe same results, such as . . ." The alternative
option is presentedin a seemingly serious mannerbut, contraryto the
declaredpositionof the advertiser,the viewerwill drawthe conclusion
that such an alternativeis absurdand ridiculous.
The followingelements typicallyappearin this version:
1. An unexpectedshift in the consumer'sframe of mind into an
imaginarystatus or into a differentproductcategory(but unlikely,to a
competitivebrand).
2. The absurdityand extremeunrealismof the alternativeoption are
obvious and recognizableby the consumer:Anyattemptto makethe
alternativemore realisticwould only weakenthe claim of the ad.
The specific scheme of the lock commercialmentionedin text consists
of two sets: A set of altemativeoptions and a set of situations.An
alternativeoption is an object or an action (a dog, in this case) which
can be used to achievethe product'sattribute(safety). The alternative
option does not have to be realisticalthoughit is assumed that the
target audiencewill be familiarwith it. A situationis a common-use
scenario of the productin time and place (in our case a peaceful
evening in the home of an old lady).The linkingoperatorlinks one
element from the situationspace (the lady) with one element in the
alternativespace (barking).
Product

Situations set:
SI S2 S3 S4

340

Message

Alternative options set:

ALTI ALT2 ALT3

MARKETING SCIENCE/VOl.

18, NO. 3, 1999

GOLDENBERG, MAZURSKY, AND SOLOMON

The FundamentalTemplatesof Quality Ads

Table 1

Continued
I. The PictorialAnalogyTemplate:
The ReplacementVersion

Specific
Scheme

II.The ExtremeSituationTemplate:
The AbsurdAlternativeVersion

Note scheme in the Introduction.

?

Locks

Situations set:
1) Sleeping,2) Burglar
breaking in during a quiet
evening 3) Nobody at home

"Quiet evening"spaceIog"
Familyconversation;
WatchingTV;
Lady reading a book

Ill. The ConsequencesTemplate:

The ExtremeConsequencesVersion
Example

Description

A commercialfor car loudspeakersshowing a bridgeon

the verge of collapse when the loudspeakersof the car
parkedon it are turnedon at high volume. The
message is that the music can be playedso loud that
even the sturdyfoundationsof the bridgeare
threatenedby its impact(BBD,Los Angeles, 1994, Cleo
awardwinner).
The idea of this version is to presentan extreme
consequence of an emphasizedproductattribute.The
absurdityof the consequence, even though presented
in a serious manner,is eminentlyobvious to the
viewer.Therefore,even a negativeresult (the collapse
of a bridge)is conceptualizedas an indicationof the
qualityof the product.The followingelements usually
appearin this version:
1 Consequencesbased on a true fact: The extreme
situationis created by takinga key attributeof the
productto the extreme (e.g., the sound emitted by the
loudspeakerscauses objects-even a sturdybridgeto vibrate).
2. The absurdityand extremeunrealismof the
consequences are obvious and recognizableby the
viewer.

MARKETING SCIENCE/VO1. 18,

NO. 3, 1999

Barking at burglars

Safety

Alternative options set:

1) Dog 2) Securityguard3)
Alarmsystem

Vace
Barking
Leash
Dogfood

IV.The CompetitionTemplate:
The Uncommon-UseVersion
A commercialfor jeans showing a couple in a broken-downcar being
towed by a pairof jeans tied to the rescuingcar.

The idea of this templateis to emphasizea productaftributeby applying

it to solve a problemin a contexttotallydifferentto its intendeduse.
The followingelements typicallyappearin this version:
1. A problematicscenario or issue.
2. Ambiguityas to the productto be the subject of the ad when the
problemor dilemmais presented.

341

GOLDENBERG, MAZURSKY, AND SOLOMON

The FundamentalTemplatesof Quality Ads

Table 1

Continued

Formulation

Ill. The ConsequencesTemplate:

The ExtremeConsequencesVersion

IV.The CompetitionTemplate:
The Uncommon-UseVersion

The specific scheme of the loudspeakercommercial

consists of two sets: a set of situationsand a set of
consequences. A consequence is a phenomenon,
action, or behaviorwhich results from the product
attributeappearingin the message. The consequence
has to appearfamiliarand not unreasonableto the
targetaudience (e.g., vibrations).It does not have to be
absurdor extreme.The linkingoperatoracts on the
productand a selected item in the consequences set by
takingthe consequence to an extreme.

The specific scheme of the jeans commercialconsists of two sets: a set

of situationsand a set of problems.The problemsuspends the natural
flow of events in the situation.The situationin our exampleis a
couple drivinga car. The problemis the breakdownof the car. The
viewerexpects to see "howit is going to continuefrom here".The
problemwill be solved by using the product;it is thereforeimportant
to inventthe problemby thinking"backwards"
so that the product
attributecontainedin the message will provideits solution.The linkis
the use of the productas a solution by exploitingthe attribute(the
strengthof the jeans).

General
Scheme

Product

Message

Product

Message

Linking operator
Situations set:
SI S2 S3 S4

Consequences set:
CONI CON2 CON3

Situations

set:

Linking operator
Problems set:
PR1 PR2 PR3

Specific

Scheme
Scheme

~~~~~~~ ? [ <<?_(

LoudLoudspeakers

Situations set:
1) Listening to music in a car
while driving 2) in a trafic
jam

Volume
~High

-----*

Jeans

Consequences set:
1) Vibrations 2) Disturbance to the other drivers
3) Other drivers listen to the music

Strong

a
with
Towingacar
Twn
the Jeans
Situations set:
Driving a car 2)
Sitting in a pub
1)

Extreme vibrations cause damage

Example

342

Problems set:
1) The car breaks
down 2) A huge
obstacle appears on the
road

V. The InteractiveExperimentTemplate:
The ActivationVersion

VI.The DimensionalityAlterationTemplate:
The Time LeapVersion

An exampleof the activationversion is an ad containinga

large blackpatch.Whenthe viewer performsthe action
suggested in the ad, he/she would become aware of
the necessity of an anti-dandruffshampoo (DDB,
NeedhamSan-Paulo,1995).

A commercialfor life insuranceshowing a wife arguingwith her husband

for canceling his life insurance.The whole scene takes place after he
dies, and portraysthe wife communicatingwith her late husbandin
the setting of a seance (a Cannesawardwinnerin 1993).

MARKETINGSCIENCE/VOl.

18, NO. 3, 1999

Table 1

Continued
VI.The DimensionalityAlterationTemplate:
The Time LeapVersion

V. The InteractiveExperimentTemplate:
The ActivationVersion
Description

Formulation

General
Scheme

The consumer is requiredto performa task or experiment The idea of this templateis to presentan ordinarysituation(in this
example,an argumentabout whetherto continueinvestingin the
in orderto receivethe message conveyed by the ad.
product).The entertainingeffect is achieved by shiftingthe scenarioto
The message is containedin the compellingresult.
the past or the future.
Most of the ads in this categoryconvey a message
emphasizinga need or a problemthat can be resolved
if the productis used. The followingelements typically
appearin the activationversion:
1. An experimentrequiringphysicalaction.
2. The experimentis executableon the spot.
The experiment'sresults highlighta generalneed rather
than a uniquequalityof the specific brand.
The specific scheme of the life insurancead consists of two sets: a set
The specific scheme of the anti-dandruffshampoo ad
of episodes introducingthe message claim (e.g., EP2-a wife arguing
consists of two differentsets: the senses set and the
with her husband)and a times set (past, future).First,the episode
experimentset. The relevantsenses set is drawnfrom
the list of the five senses. The experimentset consists
space is selected (e.g., wife, husband).Thenan operatorlinksan
of test scenarios to ascertainneed for the product.The
elementfrom the time set and an element drawnfrom the episode
space (e.g. the husband'slife status is transferredinto the future).
linkingoperatorrequirementis that the experiment
Note that the inventedsituationin the differenttime frame has to be
representedin the experimentspace will be performed
relevantto the productand its attributesand, therefore,in this case,
physicallyby interactingwiththe media (newspaper,
the futureis more appropriate.
radio,etc.).
Message

Product

Timeset:
Past, Future

p so es set:
EPI EP2 EP3...

Timeleapoperator

Ep'sSpace

Product

Message

pisodes set:
EPI EP2 EP3..

Time set:
Past, Future

Ep'sSpace

Timeleapoperato

~~~~~~~~~~~~~~~~~
~~~~~~~~~~~~~~~E
~~~~~~~~~~~~~~~E

EP2'
EP2EP29"

Specific
Scheme

Anti-dandruff
X_

Removal

Shampoo

of dandruff

Eemts:
set:
Experiments
1) Asking people
2) Testing own scalp for
dandruffsuac

ees:set:
Senses
1) Sight
(onlyone)

I) Mirror

o
2) Asample

M~~~~~~~AnRKEIGSIEC/O.18fOf,199.
4

|1)

between
Argument

over the newspaperto

wife andlatehusband

product

3 1

Titneset:
Past,Futuremoe

thescalp
Scratching
evaluateneedforthe

ExpeIment pace

Lifeinsurance <

>-

important

Ep sodes set:
Dueto injurya familyneeds
2) A discussionbetweena
coupleon a necessityof life

Bps Spa
I)wife

) husband

GOLDENBERG, MAZURSKY AND SOLOMON

The FundamentalTemplatesof Quality Ads

Figure 4

The CreativityTemplate Taxonomy

Pictorial
analogy

A_su

Extreme e
Situaton

acme
Rep

anaog

CompWition

onsequences

me

ltr

Interative
etiperiment

AttributT i

conuence competitio

Dimensionality
Alteratoon

onI

ew

leap
paamte

at1trbt coseuencJoXpetib,0|i
iExtrem
C
worth

i_
|

Uncommo
use

Divislo

$ Tme leap|

sion) or by just imaglning the performanceof such an

experiment (the imaginry eerinment verson). The interactive experimenttemplate is exemplified in Table
I by the activationversion.
VL The DimensionalityAlterationTemplate.The dimensionality alteration template manipulates the dimension of the product in relation to its environment.
connectionverIt has four versions:The new parameter
sion, the maltiplicationversion, the division version, and

the timeleapversion. This template is exemplified beconlow by the time leap version. In the newparameter
nectionversion of the template, previously unrelated
parametersbecome dependent (e.g., the speed of a new
aircraftis demonstrated by reducing the size of the
ocean).The multiplicationand divisionversions are executed by multiplying the product and comparimgthe
duplicates,or dividing the productinto its components
and creatingsome form of relationshipbetween them.
This templateis exemplifiedin Table1 by the time leap
version.
Template Distribution. Following template infer344

ence, the distribution of templates among the high

quality ads was examined. The template-adclassification was performed independently by two trained
judges (different from the prior judges) who had experience of at least ten years in the advertising field.
They were taught to identify linking operators, and
then were given a list of possible spaces. Eachtemplate
was illustratedby using five examples. Subsequently,
they were given an exercise to test their ability to correctly classify the template. Each judge was asked to
classify a set of ads (consistingof two ads per template
version as well as nontemplates).The judges correctly
classifed more than 95% of the ads in this exercise.
Following the training task the judges classified the
ads. The inte udge agreement rate in the assignment
of ads into the six templates was 94%, and disagreements were resolved by discussion. Of the 200 highly
evaluated ads, 89%could be accountedfor by the templates. Figure5 summarizesthe distributionof the ads
by their creativitytemplate.

RKETINGSCIENCE/VOL 18, No. 3 1999

GOLDENBERG, MAZURSKY, AND SOLOMON

The FundamentalTemplatesof Quality Ads

Figure 5

print ads appearing in the same magazines and newspapers as Group 2 ads and belonging to product categories comparable to Groups 1 and 2. In both cases
selection was based on convenience sampling. None of
the ads in Group 3 were award winners or finalists at
least within the three years following publication as
verified by examining the leading contest albums.

The Distributionof Ads by Templates: Study 1

70
60

50

400

-I

30
20-

z~

10

1
Pictorial
analogy

[2 2
Consequence

3
Extreme
situation

4
No
Creativity

Competition

Dimensionality
alteration

7
Interactive
experiment

templates

Study 2: Examination of the

Creativity Template Taxonomy
Study 2 was designed to examine the extent to which
creativity templates categorize winning ads. Two
classes of ads were sampled. One class consisted of
award-winning and contest-finalist ads. The second
class was composed of an array of undistinguished ads
drawn from a set of magazines comparable to those in
which the award winners and finalists originally appeared. In view of the high frequency of creativity templates in the group of highly evaluated ads found in
Study 1, our hypothesis was that creativity templates
would be more frequent in award-winning and finalist
ads than in nonwinning magazine and newspaper ads.
Stimuli. To facilitate comparison between the
findings of Study 1 and Study 2, the group of ads examined in Study 1 was termed Group 1 and the two
groups of ads in Study 2 were termed Groups 2 and 3.
Group 2 consisted of 200 ads appearing in an ad collection (The One Show Album, 19912). Of the 14,000
ads submitted to the contest, 700 were winners and
finalists. A subset of 200 ads was drawn from the winners and finalists set. Group 3 consisted of 200 selected

2Given the wider sampling frame used in Study 1 than that of Study
2, the overlap in the selected ads was minor (only two ads), and its
effect on the comparison was negligible.

MARKETING SCIENCE/VOl.

18, No. 3, 1999

Procedure. All 400 ads were analyzed according to

the creativity template taxonomy. The judges were
given instructions consistent with the schemes described in Table 1. The categorization procedure for
Groups 2 and 3 was conducted independently by two
trained judges (with similar qualifications to those
serving as judges in Study 1). The ads were presented
in a random order and the judges were blind to the
group assignment. The interjudge agreement rate was
95.5%, and disagreements were resolved by
discussion.
Results. Table 2 shows the numbers of ads in
Groups 2 and 3 that fit the creativity templates. For
comparison, the explainable proportion of Group 1
(from Study 1) is also included.
The proportion of creativity templates decreased
from 89% in Group 1 (the highest quality ads), 50% in
Group 2, to 2.5% in Group 3 (nonwinning ads).3 Comparison of the distributions of template-matching ads
in the three groups indicated a significant difference
between groups 2 and 3 (Mann-Whitney U:p < 0.01)
and only marginally significant difference between the
two award-winning groups (Groups 1 and 2, MannWhitney U: p < 0.10). It is interesting to note that in
Groups 1 and 2 the pictorial analogy template and the
consequencestemplate combined accounted for the majority of the template-matching ads. In Group 1 the
pictorial analogy template accounted for 38% of the
template-matching ads and the consequences template
accounted for 21%. In Group 2 the pictorial analogy
template accounted for 40% of the template-matching
ads and the consequences template accounted for 24%.
3The decrease in template-based explanation from Study 1 to Study
2 is not surprising given the more stringent quality requirements
imposed in Study 1. Even though the prime objective of this research
was to identify templates in the highest quality ads, search for new
templates was performed in the groups included in Study 2 but it
did not lead to discovery of other major templates.

345

GOLDENBERG, MAZURSKY, AND SOLOMON

The FundamentalTemplatesof Quality Ads

Table 2

Distributinof Templates by Ad QualityGroup

Template
PictorialAnalogy
Replacement
Extremeanalogy
ExtremeSituation
Absurdalternatives
Extremeattribute
Extremeworth
Consequences
Extremeconsequences
Invertedextremeconsequences
Competition
Attributein competition
Worthin competition
Uncommonuse
InteractiveExperiment
Activation
Imaginaryexperiment
DimensionalityAlteration
New parameterconnection
Multiplication
Division
Time leap
No CreativityTemplate
Total

Contest
Highly
Evaluated Winning
Ads
Ads

Non
Winning
Ads

Group2
(Study 2)
40
26
14
11
3
5
3
24
19
5
8
3
3
2
3
3
0
15
4
5
5
1
99
200

Group3
(Study 2)
2
1
1
1
0
0
1
0
0
0
2
1
1
0
0
0
0
0
0
0
0
0
195
200

Group1
(Study 1)
(I)
68
44
24
24
(II)
7
10
7
37
(III)
17
20
19
(IV)
5
10
4
11
(V)
8
3
19
(VI)
4
6
5
4
22
200

Study 3: Examining the Impact of A

Priori Knowledge of the Creativity
Templates on Judgments
Overview. Following Studies 1 and 2 which detected
and examined the appearance frequencies of templates
in various ad quality levels, Study 3 was designed to
examine how the utilization of creativity templates in
the creative execution process affects outcome judgments. In the first part, three groups of individuals differing in the training that they received were asked to
create ads. The first group was requested to generate
ads based only on a brief (without additional training),
the second group, in addition to the brief, was trained
to utilize the free association method in generating
new ads, and the third group, in addition to the brief,

346

was trained to utilize a creativity template. In the second part, the ads were rated by other individuals on
several key advertising outcome scales. All the participants in Study 3 held an undergraduate degree and
none of them were or had previously been employed
in advertising related jobs.
Part 1: Stimuli and Procedure. Three groups each
consisting of 20 participants (all indicated familiarity
with brands in the three examined products, were
blind to the group assignment, and were paid for participation) were randomly assigned to one of the three
idea generation groups. The three groups did not differ
in age, education and occupation. Training both in free
association and in the creativity template was conducted by experienced moderators. Training time was
comparable in the two training groups (less than two
hours including practice tasks and idea generation).
Template training involved the absurd alternatives
version of the extreme situation template (applied later
in the study for anti-dandruff shampoo ads), the interactive experiment template (for diet products ads), and
the replacement version of the pictorial analogy template (for sneakers). In all cases template training involved examples drawn solely from products different
from those used in the study.
The three groups were then asked to generate ad
ideas for the three product categories. The number of
ideas per category ranged between 32-46. To mimic a
real-world creative execution screening procedure that
typically involves deciding among the highest ranking
ideas, all ideas were submitted to a creative director
who was blind to the objective of the study and whose
task was to screen the best five ideas in each category.
The top-ranking ideas in each category were then used
in constructing a questionnaire for Part 2 of the Study.
Part 2: Stimuli, Procedure, and Judgments. Three
versions of the questionnaire were generated each containing the best 15 ideas presented in a random order
(five per product category) relevant to the specific
training condition (i.e., one version containing "no
training" ideas, second version composed of "free association" ideas, and a third version comprising
"template-based" ideas).
A different group of 36 individuals (paid for participation, and blind to the study objective and to the

MARKETING SCIENCE/VOl.

18, No. 3, 1999

GOLDENBERG, MAZURSKY, AND SOLOMON

The FundamentalTemplatesof Quality Ads

origin of ideas) were randomly assigned to one of the

three questionnaire versions. The three subgroups did
not differ on the same demographic variables examined in Part 1. Each participant rated 15 ideas on five
different scales. The first two scales assessing ad quality are frequently included as major items in advertising response scales (e.g., Edell and Burke 1987;
Mitchell and Olson 1981): One scale reflects creativity
and the other attitude toward the brand. They were
also chosen in accordance with Finke's (1990) suggestion (in creativity research) to assess ideas both by their
creativity and practical value. The other three scales
were extracted from previous studies focusing on ad
feelings. While studies on feeling responses typically
involve relatively large inventories (e.g., Holbrook and
Batra 1987, Edell and Burke 1987, Richins 1997), the
current purpose was to capture the major dimensions
proposed by Edell and Burke 1987 (see also Richins
1997), namely upbeat, warm, and negative, by key
scales, given the load involved in rating 15 ideas. To
this end, the list of responses loading on the three key
feeling dimensions was examined. The scales "humor", "emotional", and "annoy" were chosen based on
their loading on the three key dimensions and their
classification as important in practice by a senior creative director. Instructions followed Edell and Burke
(1987). All five scales were rated on five point scales (1
representing lowest value and 5 highest value).
Results. A MANOVA was performed with training as between-subjects factor (three levels of training-"no training", "free association" training, and
"template" training) and the product as a withinsubjects factor (three products-anti-dandruff shampoo, diet products, and sneakers) for the five judgbrand
ments (the judgment factor)-creativity,
attitude, humor, emotion, and annoyance. Table 3 displays the mean judgments. All main effects (judgment
type, training, and product) and their two-way and
three-way interactions were significant in multivariate
tests (all Wilks' Lambda values were significant at p <
0.001). In view of the present study's focal interest on
the effects of the template approach, the analysis subsequently concentrated on contrasts and simple effects.
The following results were obtained.
First, "template training" was found to be superior

MARKETING SCIENCE/VOl.

18, No. 3, 1999

to "no training" and "free association" training in all

the comparisons pertaining to the two ad quality measures (creativity and brand attitude, see Table 3). This
finding is based on simple effect analyses performed
for each judgment within the judgment factor with two
orthogonal contrasts, one assessing the value of training (by comparing the combined training groups with
the nontraining group) and the other, comparing between the two training methods-the template approach and free association. The first contrast showed
that, on the whole, training is effective. Except for one

3
MeanJudgments:
Experiment

Table3

Extreme
Situation
Shampoo)
Template
(Anti-Dandruff
BrandAttitudeHumor EmotionAnnoyance
Creativity
TYPE
TRAINING
2.80
Notraining

Freeassociation2.08
Template
training4.02(2)**

* 2.40
2.62(1)*

1.83

1.57

2.28

1.45

1.62

1.63

4.13(2)**

3.57(2)**

1.79

1.60

Template
(Diet)
Interactive
Experiment
BrandAttitudeHumor EmotionAnnoyance
Creativity
TYPE
TRAINING
No training

2.57(l)**

Freeassociation 2.77
Templatetraining3.82(2)**

2.38(1)-

1.90

1.90

2.07(l**

2.77

1.93
1.85

1.97

1.43
1.60

3.97(2)**

1.30(2)**

Pictorial
Template
(Sneakers)
Analogy
BrandAttitudeHumor EmotionAnnoyance
Creativity
TYPE
TRAINING
No training

2.24(l)**

Freeassociation 2.81
Templatetraining3.60(2)**

2.49(1)**

1.32(1)* 1.46

1.68(1)**

2.86

1.69

1.68

1.27

3.50(2)**

2.18(2)** 1.77

1.13

at p < 0.05.
*Significance
**Significance at p < 0.01.

groupsandthenontrainbetweenthetwotraining
(1) Denotesdifference
inggroup.
betweenthe templatetraininggroupandthe free
(2) Denotesdifference
associationtraining
group.

347

GOLDENBERG, MAZURSKY, AND SOLOMON

The FundamentalTemplatesof Quality Ads

case (creativity judgment in the case of the absurd alternative version of the extreme situation template), all
comparisons indicated that training improves ideation
quality. The second contrast showed that the high
value of training is primarily attributable to template
training. In all cases template training was superior to
training in free association (all six comparisons were
significant at p < 0.001 level).
Second, the above training effect does not generalize
to the feeling responses. Training in two of the three
templates bolstered humor reactions: Training in the
replacement version of the pictorial analogy template
yielded ads that were rated higher than free association training (F(1,177) = 102.01, p < 0.001), and training in the absurd alternative version of the extreme
situation template yielded ads that were rated higher
than free association training (F(1,177) = 6.25, p <
0.001). However, training in the activation version of
the interactive experiment was not found to invoke humor relative to the other training method (F(1,177) <
1, n.s.). In addition, none of the templates was found
to enhance emotional reactions. In fact, training in the
activation version of the interactive experiment even
hindered emotional ratings (F(1,177) = 13.69, p <
0.001). Finally, template training did not reduce annoyance compared to free association training, although a combined training effect of decreasing annoyance in two of the three comparisons was observed
(F(1,177) = 9.61, p < 0.001 in the case of activation,
and F(1,177) = 18.49, p < 0.001) in the replacement
version of the pictorial analogy template.
Third, a noticeable finding emerges from the comparison between the no training and free association
conditions. No clear indication was found that the free
association method heightens creativity or brand attitude. Although this method is widely applied in advertising practice, the contention that it necessarily enhances effectiveness was challenged by several
researchers (Perkins, 1981; Weisberg 1992). Some researchers claim that free association as well as other
frequently used projective techniques may even reduce
effectiveness even though they overcome group effects
which typically characterize focus group methods
(Diehl and Stroebe 1987, O'Guinn et al. 1998;
Dominowski 1995). Nonetheless, in the present study

348

the free association method was shown to reduce annoyance reactions in the case of two of the three products (F(1,177) = 8.53, p < 0.01 for diet products, and
F(1,177) = 4.97, p < 0.05 for sneakers).4

Study 4: Examining the Impact of A

Priori Knowledge of the Creativity
Templates on Recall
Another form of assessing ad quality was a recall test.
Three groups of 18 individuals were exposed to the
same ads as in Study 3. Each group obtained the list
of 15 ad ideas consisting of five "template-based" ads,
five "free association" ads, and five "no training" ads,
in a random order. A day later, they were visited again
and asked to recall as many ad ideas as they could
from the list seen in the previous day. The results
showed that for all three product classes among the 54
respondents template-based ideas had the highest recall rate. Overall, 28.5% template-based ideas were correctly recalled, 20% were based on free association
training, and 14.8% ads were drawn from the "notraining" set.5 McNemar tests confirmed that the distinct order of the ideation methods (template training,
free association, and no training) was statistically significant at the p < 0.01 level).

Discussion and Conclusions

The present research provides the theoretical reasoning and shows empirically that definable templates can
be detected in high quality ads. Viewed from a micro
perspective, individual advertisers may adopt idiosyncratic templates and use them in generating new ideas.
4An additional study was conducted comparing 24 ads generated by
advertising professionals(art directors and copywriters) trained in the
absurd alternatives version of the extreme situation template with
28 ads generated by professional counterparts. A competing training
condition was not included in this study given the familiarity of
professionals with the majority of the frequently used alternative
methods, which could potentially render the comparison unreliable.
The ads created by the trained professionals were rated higher by
five judges who were unaware of the template approach (mean =
4.46) than the ads created by the nontrained professionals (mean =
2.83, t(50) = 5.48, p < 0.001).
5This recall rate is consistent with an average of 20% obtained in
day-after recall tests (e.g., Wells et al. 1992).

MARKETING SCIENCE/VOl.

18, No. 3, 1999

GOLDENBERG, MAZURSKY, AND SOLOMON

The FundamentalTemplatesof Quality Ads

The present research lends support to a broader perspective which contends that the templates may be
widely applied across products, messages, and target
audiences. It serves to enhance the understanding of
the emergence of quality ads as well as creativity in
marketing communication.
The findings of the present study indicate the superiority of template-matched ad ideas in creativity
judgments, brand attitudes, and recall. However, a differential impact was detected regarding the feelings
invoked by the templates. This is not surprising in light
of the cognitive processes that different templates are
expected to invoke. For example, the interactive experiment (activation) template requires physical activity and effortful processing. Enhanced involvement is
associated with high cognitive activity (e.g., Assael
1998) and thus, the interactive experiment is likely to
be useful when the strategy is designed to cause a particular behavioral change which is less responsive to
peripheral cues. The Absurd Alternatives and Replacement version are more useful when humor is the intended strategy. Hence, while generally, the templates
appear to have a positive impact on recall leading to
potential sales increase, the ways to achieve this goal
are diversified.
These findings are a step toward defining a comprehensive model of the antecedents of outcome reactions
to advertising stimuli. Improved understanding of the
wide spectrum of reactions that connects the basic templates with end-user reactions is likely to be beneficial
for both academicians and practitioners. Such a framework would create a synthesis between the activity of
creative professionals whose focal interest is generating the ads, managers, whose main responsibility is
strategy formulation, and the academic activity, which
focuses mainly on the consumer reaction-end of the
advertising process. Obviously, further research is required to shed more light on this important aspect of
creativity templates.
Three important features of the template approach
need to be emphasized. First, the templates are useful
in guiding the creativity execution process; however,
they do not prescribethe outcome ideas. In other words,
they provide the framework for generating ideas although within the template constraints various ideas
may be generated. Second, templates are less transient

MARKETING SCIENCE/VOl.

18, No. 3, 1999

than the ideas produced, but this does not mean thal
templates are permanent or that they are insensitive tc
changes over long term frameworks. Indeed, advertising reflects social norms and trends, and as such, long
term social trends are expected to reshape the templates and provide conditions for the evolution of new
templates. Nonetheless, the dynamics of template
changes are expected to be much slower than the dynamics of changes in ad hoc idea generation. Third,
from a theoretical viewpoint, it is also expected that
the set of templates will always remain small: Only
under this condition will templates maintain their generalizability and ensure adherence to the overall intention of the message and consistency with the chosen
marketing strategy.
It is postulated that the template taxonomy provides
the means to achieve "creativity expertise". Unlike the
divergent thinking approaches, in which the required
expertise is not necessarily related to the creativity process itself (e.g., individuals can be trained to be better
moderators in brainstorming), the creativity template
approach is trainable and has the capacity to directly
improve creativity outcomes.6 In fact, training individuals in creativity templates may result in higher levels
of "creative expertise" (Alba and Hutchinson 1987).
The template taxonomy facilitates the focused cognitive effort involved in generating new ideas, the capacity to access relevant information, and improved
memorability of the reduced set of information needed
to perform the tasks.
Our research identifies an infrastructure of certain
regularities that may assist in screening and constructing creative ideas. If we can define these regularities a
posteriori, we can reconstruct a priori ad skeletons,
which consist of their main parameters and can be fed
in only by those ideas that conform to these parameters. In this newly generated schematic world most of
the ideas are likely to be perceived as creative, even
though the well-defined rules and the exhaustive
search used to obtain them do not easily reconcile with
61t should be noted that while Study 3 successfully demonstrated the
value and potential of the template approach, it was applied on versions that constitute less than a third of the highest quality ads of
Study 1. It is suggested that future research will validate the approach in its entirety and screen the most effective templates for
training advertising creatives.

349

GOLDENBERG, MAZURSKY, AND SOLOMON

The FundamentalTemplatesof Quality Ads

what we traditionally have viewed as "pure creativity". Yet, in view of the numerous problems we are
called upon to solve in our day-to-day lives, and if we
accept that creativity is assessed by the way it is perceived by consumers, it seems that we ought to reappraise our fundamental approaches to creativity and
even reevaluate their operational definition. Our experience with using templates both in the reported
studies and in real-life applications among several
leading ad agencies, indicates that creative activity
prescribed by the well-defined, template-based processes, is a trainable, resource-saving, and effective
tool. It simplifies and improves the decision-making
process involved in designing advertising strategy.
The template approach can be applied either by hiring
personnel that is experienced in templates and employed by a consulting firm, or by training the agency's
own personnel to routinely evaluate past and current
ads, and engage in creative activity.
In addition, the present investigation concurs with
an emerging stream of research which deemphasizes
the traditional treatment of visual and verbal modes in
advertising as functionally distinct entities. Some of
the qualities of pictures which, in the past, were believed to characterize verbal information, and some of
the qualities of verbal information which were previously more closely associated with pictures, are being
reexamined. One direction of this research is visual
rhetorics. Scott (1994) challenged the assumption that
pictures are merely reflections of reality, claiming that
images represent complex figurative arguments. Relatedly, although bearing on verbal information,
Unnava et al. (1996) argued against the concentration
of consumer research on visual imagery as the only
type of imagery, claiming that words differ in the degree to which they provoke imagery or influence reading and listening. The creativity template approach is
in accord with this research trend in that it treats the
message and its delivery as a whole rather than decomposing it into the functions carried by the visual
versus the verbal modes. The Nike-Air and the barking
lady examples serve to illustrate the complex figurative arguments that are conveyed by pictures. Moreover, in the anti-dandruff ad, imagery-provocation
and informativeness are entirely independent of their

350

verbal and visual components; neither the visual component nor the verbal component can be understood
as separate information entities but, in combination,
they achieve a high level of imagery.7
References
Alba, Joseph W., J. Wesley Hutchinson. 1987. Dimensions of consumer expertise. J. ConsumerRes. 13 411-454.
Alden, Dana L., Wayne D. Hoyer, Chol Lee. 1993. Identifying global
and culture-specific dimensions of humor in advertising: a multinational analysis: J. Marketing 57(2) 64-75.
Altschuler, G. S. 1986. To Find an Idea: Introduction to the Theory of
Solving Problemsof Inventions. Novosibirsk, USSR, Nauka.
Assael, Henry. 1998. Consumer Behaviorand MarketingAction. South
Western College Publishing, Cincinnati, OH.
Batra, Rajeev, David A. Aaker, John G. Myers. 1996. AdvertisingManagement. Prentice Hall, Englewood Cliffs, NJ.
Blasko, Vincent, J., Michael P. Mokwa. 1986. Creativity in advertising: a Janusian perspective. J. Adv. 15 (4) 43-50.
Boden, Margaret. 1992. The CreativeMind. Abacus, London, UK.
Burke, Raymond R., Arvind, Rangaswamy, Jerry Wind, Jehoshua
Eliashberg. 1990. A knowledge-based system for advertising
design. Marketing Sci. 9 (3) 212-29.
Chomsky, Noam. 1978. Syntactic Structures. Peter Lang Publishing,
New York.
Dasgupta, Subrata. 1994. Creativity in Invention and Design-Computational and Cognitive Explorations of TechnologicalOriginality.
Cambridge University Press, Cambridge, England.
Diehl, Michael, Wolfgang Stroebe. 1987. Productivity loss in brainstorming groups: Toward the solution of the riddle. J. Personality Soc. Psych. 53 (3) 497-509.
Dominowski L. Roger. 1995. Productive problem solving. S. Smith,
T. Ward, R. Finke, eds. CreativeCognition Approach.Massachusetts Institute of Technology, Cambridge, MA, 71-95.
Eco, Umberto. 1986. Semioticsand the Philosophyof Language.Indiana
University, Bloomington, IN.
Edell, Julie A., Marian Chapman Burke. 1987. The power of feelings
in understanding advertising effects. J. Consumer Res. 14 (December) 421-33.
Finke, Ronald A., Thomas B., Ward, Steven M. Smith. 1992. Creative
Cognition. MIT Press Cambridge MA.
Hayes, John R. 1978. Cognitive Psychology:Thinkingand Creating.The
Dorsey Press, Chicago, IL.

7This research was supported by grants from the K-Mart Foundation

and the Davidson Center. The authors wish to thank Richard Staelin,
editor of Marketing Science, the area editor, and the anonymous reviewers for numerous helpful suggestions and comments. The authors also thank Asim Ansari, Beth Hirschman, Jacob Hornik, Gita
V. Johar, Amnon Levav, Yaacov Schul, and the creative directors and
planners from Gray Advertising, McCann-Erickson, Ogilvy and
Mather, Rapp-Collins, and Symbol Peres for their cooperation and
valuable comments.

MARKETING SCIENCE/VOl.

18, No. 3, 1999

GOLDENBERG, MAZURSKY, AND SOLOMON

The FundamentalTemplatesof Quality Ads

Holbrook, Morris B., Rajeev Batra. 1987. Assessing the role of emotions as mediators of consumer responses to advertising. J. Consumer Res. 14 (December) 404-20.
Kauffman, Stuart. 1995. At Home in the Universe. Oxford University
Press, Oxford, England.
Kelso, J. A. Scott. 1997. Dynamic Patterns: The Self-Organization of
Brain and Behavior.MIT Press, Cambridge, MA.
Levi-Strauss, Claude. 1974. Structural Anthropology. Basic Books,
New York.
Mcquarrie, Edward F., David Glen Mick. 1992. On resonance: a critical pluralistic inquiry into advertising rhetoric. J. ConsumerRes.
19 (September) 180-97.
. 1996. Figures of rhetoric in advertising language. J. Consumer Res. 22 (March) 424-437.
Minsky, Marvin. 1985. Extraterretrials:Science and Alien Intelligence.
Edward Regis, ed. Cambridge University Press.
1988. The Society of Man. Simon and Schuster.
Mitchell, Andrew A., Jerry C. Olson. 1981. Are product attribute
beliefs the only mediator of advertising effects on brand attributes? J. ConsumerRes. 18 (August) 318-322.
O'Guinn, Thomas, C., Chris T. Allen, Richard J. Semenik. 1998. Advertising. South-Western, Cincinnati, OH.
Palmer, Edgar M. 1985. GraphicalEvolution:An Introductionto Random
Graphics.John Wiley and Sons; New York.
Perkins, D. N. 1981. TheMind's Best Work.Harvard University Press,
Cambridge, MA.
Richins, Marsha. 1997. Measuring emotions in the consumption experience. J. ConsumerRes. 24 (September) 127-146.

Rothenberg, A. 1971. The process of Janusian thinking in creativity.

Archives of GeneralPsych. 24 (March) 195-205.
Scott, Linda M. 1994. Images in advertising: the need for a theory of
visual rhetoric. J. ConsumerRes. 21 (September) 252-273.
Simon, Herbert A. 1966. Scientific discovery and the psychology of
problem solving. R. G. Colodny, ed. Mind and Cosmos:Essays in
ContemporaryScience and Philosophy,Vol. 3, University of Pittsburg Press, Pittsburg, PA.
Simon, Julian L. 1971. The Management of Advertising. Prentice Hall,
Englewood Cliffs, NJ.
Tauber, Edward M. 1972. HIT: heuristic ideation technique-a systematic procedure for new product search. J. Marketing 36 5861.
The One Show. 1989. Vol. 11. Rotovision, S.A.; Switzerland.
1991. Vol. 13. Rotovision, S.A.; Switzerland.
1992. Vol. 14. Rotovision, S.A.; Switzerland.
Unnava, Rao H., Sanjeev Agrawal, Curtis P. Haugtvedt. 1996. Interactive effects presentation modality and message-generated imagery on recall of advertising information. J. ConsumerRes. 23
June 81-87.
Urban, Glen L., John R. Hauser. 1993. Design and Marketing of New
Products. Prentice Hall, Englewood Cliffs, NJ.
Weisberg, Robert W. 1992. CreativityBeyondtheMyth of Genius.W. H.
Freeman, New York.
Wells, William, John Burnett, Sandra Moriaty. 1992. Advertising:
Principles and Practice.Prentice Hall, Englewood Cliffs, NJ, 607.
Winston, F. 1990. The management of creativity. Internat. J. Adv. 9
1-11.

This paper was receivedMarch 18, 1998, and has been with the authors 7 monthsfor I revision;processedby J. WesleyHutchinson.

MARKETING SCIENCE/VOl.

18, No. 3, 1999

351