How Web Banner Designers Work: The Role of Internal ...
Transcript of How Web Banner Designers Work: The Role of Internal ...
HAL Id: sic_00288393https://archivesic.ccsd.cnrs.fr/sic_00288393
Submitted on 21 Jun 2008
HAL is a multi-disciplinary open accessarchive for the deposit and dissemination of sci-entific research documents, whether they are pub-lished or not. The documents may come fromteaching and research institutions in France orabroad, or from public or private research centers.
L’archive ouverte pluridisciplinaire HAL, estdestinée au dépôt et à la diffusion de documentsscientifiques de niveau recherche, publiés ou non,émanant des établissements d’enseignement et derecherche français ou étrangers, des laboratoirespublics ou privés.
How Web Banner Designers Work: The Role of InternalDialogues, Self-Evaluations, and Implicit
Communication TheoriesDidier Courbet, Marie-Pierre Fourquet-Courbet, Marc Vanhuele
To cite this version:Didier Courbet, Marie-Pierre Fourquet-Courbet, Marc Vanhuele. How Web Banner Designers Work:The Role of Internal Dialogues, Self-Evaluations, and Implicit Communication Theories. Journal ofAdvertising Research, World Advertising Research Center, 2007, 47 (2), pp.217-229. �sic_00288393�
How Web Banner Designers Work: The Role
of Internal Dialogues, Self-Evaluations, and
Impiicit Communication Theories
MARIE-PIERRE
FOURQUET-COURBET
University of Avignon,
France
marie-pierre.fourquet@
univ-avignon.fr
DIDIER COURBET
University of
Nice-Sophia Antipoiis,
France
MARC VANHUELE
HEC School of
IVlanagement, Paris,
France
Tliis research was supported by
the Centre National de ta Re-
cherche Scientifique (CNRS-
France), program "Cognitiqiie-
Societe de Vmformatioii."
In-depth interviews with web banner designers, combined with retrospective protocols,
reveai implicit theories ofthe communication process that they apply during their
creation process. These theories take the form of reactions of imaginary audiences
with whom web banner designers engage in imaginary dialogues. The dialogues reveal
the evaluation standards held by internet users, advertisers, and different colleagues.
CREATIVE COMMUNICATION DESIGNERS work with
little or no support from formal theories of com-munication, but Kover (1995) shows that copywrii-
ers nevertheless share implicit theories about thefunctioning of communication that lack a scien-tific foundation but that people use to explainreality. They are implicit because the people whohold them are not necessarily aware of them andoften cannot express them in a precise way (Furn-ham, 1990; Schneider, 1973). Kover elicits implicittheories through open interviews with copywrit-ers and shows that the theories operate within thecontext of an internal dialogue between the cre-ative person and an imaginary communicationrecipient, during which the creator tests potentialmessages on the virtual audience and refines themon the basis of the imaginary feedback obtained.This process continues iteratively until the de-signer achieves an emotional agreement. WhereasKover's study is based on interviews with copy-writers from traditional advertising agencies, weconsider the extent to which his findings mightapply to the production process of another groupof communication professionals: creators of webbanners.
Several reasons exist to doubt the applicabilityof Kover's (1995) findings to web banner design-ers. At a theoretical level, the notion of implicittheories stems from Moscovici's (1984) work onthe social interpretations shared by a group, butsuch representations may not yet exist in the webdesigner world for two reasons. First, the businessis young and may not have a strongly developed
culture or set of processes to help newcomersassimilate. Second, web banner designers are avery heterogeneous group, including computerexperts, graphic designers, traditional advertisingcreative personnel, and so forth. At a more prac-tical level, designing web banners involves differ-ent cognitive processes tban those associated withcopywriting, as well as different sensory-motoractions, because it occurs on the computer andtherefore may depend more on trial-and-error cy-cles than copywriting does.
In addition to examining whether implicit theo-ries and internal dialogues are part of the webbanner creation process, we examine how thesetwo elements integrate in the actual creative pro-cess. Recent work (El-Murad and West, 2004; Lon-ergan, Scott, and Mumford, 2004; Mumford, 2003)outlines the steps of the creative process, whichmay or may not map onto Kover's (1995) internaldialogue idea. Does a creative process correspondto one long dialogue or a series of short ones withinterruptions? Is the dialogue with one imaginaryviewer or several?
Our project therefore has clear implications fordifferent types of practitioners. Manuals aboutweb advertising (e.g., Adams, 2003; Zeff and Aron-son, 1999) discuss the results of the creative pro-cess, but rarely the process itself. In a professionalcontext, most designers remain secretive abouttheir working methods, so web banner creatorsmay find that our work stimulates their ownproductivity, because they can discover how theircolleagues work and reflect on their own work
DOI: 10.2501/S0021849907070213 J u n e 2 0 0 7 JOURORL OF HOUERTISIDG RESERRCH 1 8 3
HOW WEB BANNER DESIGNERS WORK
processes. Our study also could help ad-vertising executives and even advertisersdevelop better working relationships withcreative personnel through a better under-standing of work methods and the mentalmodels they hold of their own work.
In the next section, we provide an over-view of recent frameworks describing theoperation of creative processes in advertis-ing. Subsequently, we describe the meth-odology of our study. The results appearin three different sections: In the first sec-tion, we show that Kover's (1995) implicittheories apply to web banner creation andthat internal dialogues take place duringthe different phases of the creative process.In the next two sections, we present the keydifferences between our results and previ-ous work by describing how our infor-mants engage in dialogues with not one butfive imaginary people and showing that ideaevaluations, which typically appear in theo-ries of creativity as the end phase of thecreative process, actually exist during eachof the creative steps.
CREATIVE PROCESSES IN THE
PRODUCTION OF ADVERTISING
Prior research on creative thinking agreesthat it involves some form of problemsolving through generating novel and per-tinent insights (Stemberg and Lubart, 1991)and proposes several models of creativ-ity, often based on similar ideas, but thatuse different terminology (see Johar, Hol-brook, and Stem, 2001; Mumford, 2003).In their review of the trends in creativeresearch in an advertising context, El-Murad and West (2004) distinguish threeprimary theories. According to Kris's(1952) primary process cognition, a per-son switches between two types of pro-cesses: dreaming in a quasi-hypnotic stateand reasoning in an abstract and logicalway. Creative people switch easily be-tween the two processes, using the firstto discover new combinations of mental
elements and the second to elaborate thesenewly discovered combinations. Thetheory of associative hierarchies (Med-nick, 1962) pertains to people's abilitiesto make associations between previouslyunrelated facts; highly creative peoplemake more such associations by usinganalogical transfer mechanisms. Finally,Mendelsohn's (1976) defocused attentiontheory takes as criteria the number ofelements a person can keep in mind atone time, because more available ele-ments allow for more original combina-tions, whereas focused attention makesfewer elements available.
The creative process itself also consistsof different steps, though most authorsrefer to Wallas's (1926) four stages. Thepreparation stage consists of a series ofpreliminary analyses to solve a problem,followed by incubation, or a temporarywithdrawal to let the project "stew." Inthe illumination stage, the person achievesa sudden insight, then examines it in theverification and evaluation stage. This last
stage has received relatively little atten-tion in the literature compared with cre-ation stages.
Kover (1995, p. 599) observes that copy-writers perceive the advertising commu-nication process as consisting of two steps:"(1) breaking through to attract interest,and (2) delivering a message." Their en-gagement in internal dialogues helps themdevelop and mentally test messages thatmove the audience through these twosteps. However, it is not clear how theprocesses that Kover describes fit withexisting models of the creative process.
METHODOLOGY
Our objective is to gain an understandingof (1) the implicit representations and theo-ries that web banner designers hold re-garding the communication process inbanner advertising and (2) the cognitiveprocesses involved in web banner cre-
ation and how they fit into existing theo-ries about creative production. Becauseour objective has an exploratory nature,we undertake a qualitative research ap-proach; Russ (2003) recommends this ap-proach in particular for examining creativeprocesses.
Data collection
We interviewed 31 French web bannerdesigners for approximately an hour anda half each, using a face-to-face, semi-direct format with open-ended questions.The interviewer asked opening questionsand helped informants develop a clearrepresentation of their ideas in depththrough reformulations, relaunched ques-tions, listening signals, the use of silence,and so forth. We combine this type ofinterview with retrospective protocols onthe creation process (Reis and Gable, 2000),during which designers demonstrated theircreations on a computer screen and ex-plained step-by-step how they had devel-oped them by thinking out loud. We usedan interview guide inspired by Kover(1995) for tbe first five interviews, thencomplemented it to include coverage ofthe creative process (see the Appendix).
For our sample selection (see Table 1),we attempted to maximize the diversityof informant profiles in terms of workenvironment (advertising agency, internet-oriented agency, freelance), size of tbeorganization (small, average, large), geo-graphical location (Paris area, provinces),level of education (self-taught, higher ed-ucation in computer graphics, advertis-ing), and age. The hermeneutic perspectivewe use for the analysis zooms in on com-monalities in responses across respon-dents, not their differences (see Young,2000).
Interpretation
Through our analysis, we attempt to id-
entify shared views about the creative
1 8 4 JOUROHL or eOUERTISIflG RESERRCH June 2 0 0 7
HOW WEB BANNER DESIGNERS WORK
To draw the attention of net surfers
Informant Characteristicsvisual (animation, flickering, alternating
Company type Advertising/communication agency: 11 moving and motionless elements, con-Web agency 9 trasts in colors, forms, and type) and se-
mantic techniques (Kover, 1995). TheGraphic design: 8
communication rules that designers refer
; 9.̂ ..̂ r.'..iT to seem to have a universal character, in
Region Paris region: 4 that they apply in all situations and for all
Provinces: 27 audiences and products (e.g., "flickeringattracts the viewer's attention").
Company organization/size Freelance: 4 c J • ,. ,. u \ ,. c!T. r.....f. ° ; Some designers try to shock net surf-
Less than 10 employees: 17 ers to get their attention, such as with aMore than 20 employees: 10 specific graphic atmosphere or the use of
codes that are out of touch with the se-Informant background Web design: 4
mantic universe of the product, but most
'^L^,P^]9,^°!^P^}^L'^.^^}^'^}..^. use traditional approaches (originality.
Computer systems: 5 playfulness, teasing) to generate trust and
ir^aditional advettising-'lO^ ^'S'^^' *° * ^ '•^^^'^^'" * ^ * ^ '^ ' * ' "S '^"^advertiser's website will not be a waste
Traditional graphical design: 6 <: ,.• -ru • u- ..• i ci >.s....". f? of time. Their objective also often centers
Other: 2 on generating positive emotions to make
the receiver more open to the message
itself. One informant expressed this very
idea:
process. After we filtered out irrelevant
information, two coders independently an-
alyzed the interview transcripts (1,203 re-
sponses, reactions, and remarks by the
respondents 56,780 words) using the con-
tent analysis (Spiggle, 1994). After a first
reading of all the transcripts, the coders
assigned the text data to mutually exclu-
sive categories that represent themes and
subthemes, according to the methodol-
ogy proposed by Weber (1990). We eval-
uate, on the one hand, the relative
importance of each theme and subtheme
compared with the entire corpus and, on
the otber hand, the importance of each
subtheme compared with the topic to
which it is attached. Intercoder agree-
ment regarding themes is 85 percent, and
disagreements were resolved after discus-
sion. We achieved the saturation point
(Glaser and Strauss, 1967) for the analy-
sis at the 25th interview. From the 26tb
interview onward, we uncovered no new
information.
IMPLICIT THEORIES
A specific characteristic of web banners,
compared with standard advertisements,
is their drive to generate clickthroughs
to the website of the advertised brand
(Chandon, Chtourou, and Fortin, 2003).
Although their communication objectives
differ from those of copywriters, web ban-
ner designers nevertheless maintain an
implicit theory about the communication
process that includes the same two steps
discovered by Kover (1995) for tradi-
tional advertising; namely, they distin-
guish the challenges of drawing attention
from those of delivering the message. Sim-
ilar to advertising copywriters, banner
designers represent their audience as in-
dividual persons with whom they en-
gage in a dialogue.
j make highly interactive banners zuith a
message that changes when the viewer uses
the mouse to move the cursor on the ban-
ner. Letting the user play makes him eager
to know more about the brand behind the
banner.
Once the imaginary recipient has allot-
ted attention to the message, the de-
signer tries to make him or her understand
the content (second stage). To cause the
viewer to progress through tbis stage suc-
cessfully, the message must be simple on
a formal level (e.g., appropriate typogra-
phy, such as Arial font) and explicit on
the semantic level (coherence between text
and image). Despite the significant
amounts of iconicity of advertising on
the internet, according to our informants,
consumer understanding occurs primar-
ily because of the text:
June 2 0 0 7 JOUROHL OF flDUERTISIIlG RESEHRCH 1 8 5
HOW WEB BANNER DESIGNERS WORK
The communication rules that designers refer to seem to
have a universal character, in that they appiy in aii situ-
ations and for aii audiences and products.
Because of the long download time, we
eannot make beautiful commercials for the
internet yet. The images in banners re-
main rather basic. Jt is therefore the text
that communicates the message.
Once they understand the message, in-ternet users move to the third stage, ac-cording to our informants: memorizationof the message. To facilitate memoriza-tion, designers emphasize the key ele-ments of the message (e.g., colors that areeasily remembered, visual shifts, funnyImages). These elements get repeated inboth the same and different banner adver-tisements, but there seems to be a thresh-old of maximum repetition that must notbe crossed to avoid obstructing the view-er's surfing or psychological rejection ofthe brand.
Web banner designers are either more op-timistic or more pessimistic about the im-pact of their work than were Kover's (1995)copywriters. Most of our informants thinkthe internet has at least as much, if not more,influence than traditional advertising me-dia (e.g., television, radio, press) becauseof its capacity to individualize communi-cation and establish a privileged relation-ship with recipients by adapting messagesto their psychological and demographicprofiles:
Communication takes place in a more in-
timate setting than with other media. The
individual is alone in front of the com-
puter screen that stands physically close to
him.. This is the perfect medium for cus-
tomer relationship management.
They also perceive the internet as a faster,
more interactive, and richer multisensory
medium than traditional media:
The advertiser often thinks that a web
surfer is a passive viewer. That's an error:
because this person participates physically,
thinks continuously before moving the
mouse around. Watching, listening, feel-
ing, the posture before the screen, they all
make the internet a multisensory medium.
These designers sometimes appear tohave an exaggerated opinion of the powerof their creations, because they ignore otherfactors that determine banner success (e.g.,product positioning, attractiveness of theoffer, media plan). Thanks to the "feed-back" that the internet allows (number ofclicks, sales, site traffic), the designer canquickly assess the quality of the banneradvertisements he or she created and mod-ify them accordingly.
Some designers, however, consider webbanners' influence weak. According tothem, because the internet surfer is cog-nitively more "active" than recipients oftraditional media, he or she also has greaterability to circumvent advertising. Thisavoidance stimulates the use of more in-trusive forms of web advertising (e.g.,pop-ups) that are considered potentiallyharmful for the advertiser, because theyobstruct the web navigation and therebycause net surfers to seek consciously toavoid them. In addition, as the recipientgets more experienced, he or she becomesaccustomed to banner advertising, whoseimpact thus decreases.
iVIULTIPLE DIALOGUES
Our analysis of the interviews shows that,similar to copywriters, web banner design-ers "hear" intrapsychic voices and dia-logue with them in working memory.Informants spontaneously mentioned theexistence of these voices, not as a result ofsuggestive questioning by the interview-ers. All their imaginary interaction pro-cesses follow the direction of a voice thatrepresents the designer's ego during cre-ation. According to Kover (1995), when acopywriter's ego interacts intrapsychi-cally with "internahzed others," a "psy-chic dissociation," if not a "creative trance,"ensues. Our results corroborate Kover'sdescription of an internal dialogue as com-ing close to "madness" or "automaticwriting."
However, in contrast with Kover (1995),who refers to only one other actor (i.e.,"an internalized target person," p. 601),the intrapsychic voices mentioned by ourinformants represent five distinct actors(in addition to the designer):
I am in another world. I do not feel time
passing and do not know where I am. The
advertiser, the web surfer, and one or two
colleagues join me in this world, jfeel com-
pletely disconnected. We have a question-
answer session. On this basis, J can create
banners very quickly. I close my eyes and
see how the banner takes shape all by itself,
as a result of the responses.
j try to see what each person would say; I
check with each of them in turn.
The first voice the designer hears is thatof a recipient surfing the internet whoreacts to the message being created. If thedesigner knows the advertised productwell, he or she becomes the referent. Thedesigner then dissociates him- or herselfinto two separate entities and dialogueswith the other self, who plays the role of
1 8 6 JOORRHL OF RDOERTISIRO RESERRCH June 2 0 0 7
HOW WEB BANNER DESIGNERS WORK
Web banner designers "hear" intrapsychic voices and
dialogue with them in working memory.
a typical recipient of the web banner ad-vertisement. When he or she knows littleabout the product, the designer creates arecipient using mental representations ofa typical person. This dissociation alsoappears in Kover's (1995, p. 601) work.According to our respondents, this firstvoice might
tell me something like "this banner is in-
teresting because it is the first time I see
one like that," [then] J immediately check
with the advertiser to have his opinion.
The second voice is that of the adver-tiser. The designer imagines what the ad-vertiser would say about the designedbanner advertisements on the basis of in-formation received during briefings, theobjectives of the campaign, and his or herknowledge of competitive banner adver-tising activity:
Suddenly, J hear the advertiser shouting
that j am a no-good, that J do not under-
stand what he wants. The web surfer [says]
that the advertiser does not know what he
is talking about and that he finds him
funny.
The ego of the people interviewed alsointeracts with two imaginary colleagues.Kover, James, and Sonner (1997) demon-strate that professional colleagues serveas an important target audience for cre-ative people; we confirm this role andshow that colleagues are present in de-signers' minds during the creation pro-cess itself. The opinions of these twoimaginary fellow members get mentallycreated from continuous monitoring ofinternet advertising.
The first colleague incarnates an excel-lent banner advertising designer who rep-resents the standards and practices in theprofession. Interaction with this fellowmember of the design community pro-vides an evaluation of the value of thechoice options in the design:
My pal will say that this is just junk.
The second colleague knows all the latestfashions and recent trends in the field ofadvertising creation on the internet:
You are a real "has been" with this banner.
Finally, the fifth voice represents the as-pirational self-image of the designer thathe or she hopes others will hold about himor her and that will give him or her a uniqueposition in the world of banner designers:
I want to leave my mark on each on my
creations. In my mind I try to establish a
compromise between what the advertiser
has told me and my personal creative style;
whenever possible, J try to make advertise-
ments that are light, nice, or humorous.
Kover (1995, p. 602) describes how thedialogue "stops when the copywriter con-nects with the other, when the other andthe copywriter communicate and the com-munication strikes an appropriate emo-tional cord in both," in which case, thetension of that dialogue is relieved. Ourinformants also feel the need to connectwith several others and try to get positivereactions from all, which implies that com-promises must be made and that tensionusually remains, even if the best possiblesolution has been found:
Jt is rare that all those whom J imagine
agree. But I have to make sure that at least
three or four are not too frustrated.
THE CREATIVE PROCESS
How banner designers work
Barmer designers perceive little differencebetween designs created by experience orby intuition. The process of self-observationthat they spontaneously establish whenengaging in their job does not involveverbalizing the rules, nor does it explainany processes in detail. Therefore, part ofthese processes likely is based on auto-matic, procedural know-how and createdunconsciously, effortlessly, and without in-tentional control (Kirsner et al., 1998).Through repetition and intensive use, heu-ristics and procedures become automaticcreation "practices" that are difficult toverbalize. However, designers do not de-preciate their ignorance of their own rea-soning; on the contrary, they regard it asvaluable, a kind of artistic genius andtalent:
Once I know the objectives I have to reach,
things develop all by themselves in my
head; J am unable to tell you how I do it;
that's probably what we call "genius."
This talent is appreciated much more thanscientific or academic knowledge. The re-sults of the content analysis even show acertain depreciation of clearly thought-out acts of creativity, in favor of intuitiveacts based on a kind of creative illumina-tion and synonymous with artistic talent:
What you learn in school are just the
basics and the errors to avoid. You cannot
acquire talent in school. It is about intu-
ition. You are born with it or not.
At the same time, some stages of thecreative processes seem more active anddeliberate than the literature on creativity
June 2 0 0 7 JOURORL OF ODOEOTISIOG RESERRCR 1 8 7
HOW WEB BANNER DESIGNERS WORK
Through repetition and intensive use, heuristics and
procedures become automatic creation "practices" that
are difficuit to verbaiize.
might suggest. Our informants use a range
of well-developed techniques to search
for an original association or develop one
on the basis of preselected elements. In
addition, they deliberately drop projects
for some period of time, during which
they search for exposure to other media
(e.g., television, CD-ROMs) to "discover"
analogies:
/ look a lot at program sponsoring on
television because these messages are sim-
ilar to ours.
During or immediately after the intra-psychic verbal interactions, the designersenter a materialization phase in whichthey generate ideas through a very quickcognitive or sensory-motor process. Somedesigners see their creation first in mentalimagery, whereas others prefer to put itdown immediately, with paper and pencilor directly on the computer. Once visual-ized or graphically realized, the creatorsmove to evaluate the resulting production.
Continuous evaluations
Our content analysis shows that a de-signer uses evaluation criteria based onfive factors: (1) implicit theories aboutinfluence over the internet surfer, (2) thedesigner's perceptions of his or her per-sonal creation style, (3) the knowledge heor she has acquired during previous cre-ations that succeeded or failed, (4) currentadvertising trends and recent messagescreated by others, and (5) the perceivedexpectations of the advertiser. The corre-spondence between these factors and the
internal voices we detailed in the previ-ous section seems obvious. Implicit theo-ries about the reactions of different partnersduring the imaginary dialogues inter-twine with these evaluation criteria, andsome dialogue partners clearly embodythe criteria.
Contrary to the majority of academicwork on the process of creativity, we findthat the assessment process does not takeplace only at the end of the creative act.Instead, the designer self-evaluates his orher message as it takes form, so even veryearly in the process, the constructed eval-uation standards serve as guides (in apositive sense as well) throughout the pro-cess. In addition, the evaluation stan-dards are not binary (acceptable versusunacceptable), but rather form a latitudeor spectrum of acceptance. A designerimagines what, in reference to Weber(1949), might be called an "ideal-typicalbanner ad," and when the creation comesclose enough to this ideal type, the de-signer agrees to stop the design process.The standards thus represent bench-marks, but provide arguments and sup-port during the final presentation to theclient. This practice can be considered sim-ilar to an anticipatory move in the subtlegames copywriters play to keep control oftheir creative work (Kover and Goldberg,1995). For example, as our respondentsacknowledge:
From the first brief of the client onwards, I
try to determine what the criteria are on
which no compromises can he made: for
instance that the brand name has to he
very visible. But I also look out for criteria
that seem less stringent and on which I
can make the client change his mind....
Next, during the creation process itself, I
check systematically if my banner corre-
sponds to these criteria.
During the brief I identify already the
arguments that I will eventually use to
"sell" the banner that 1 have not even
created yet. Of course, the brief orients my
creative work, but it also helps to develop
my final arguments and to prepare ques-
tions to the objections that will be made.
If the evaluation standards are not ex-plicit, the designer imagines them usinghis or her preliminary experience and rep-resentations of the target audience.
During the creative process, the de-signer also systematically evaluates theaesthetic aspect of the message. Our in-formants insist that this assessment mustbe fast, not intellectualized or elaborated;that is, the message must have immediateappeal. When a concept receives a nega-tive evaluation, it might be abandoned orimproved during a new round, usuallyby invoking implicit communication theo-ries ("my message does not attract atten-tion; I need to use an attention grabbingtrick"). When it receives a positive evalu-ation, it gets developed in further detail.
The process of creating the messageoften ends when the message meets theacceptance criteria of the advertiser, suchas the designer imagines them. Therefore,when the designer's imagined voices donot agree, the advertiser's voice domi-nates and has the last word. The voice ofthe advertiser therefore enables designersto limit or even erase cognitive disso-nance that could cause stress and preventthem from leaving the process of psychicdissociation, which they need to do togenerate ideas. Therefore, the concept of"satisficing" (Simon, 1960) applies to web
1 8 8 JOURORL or ROUERTISIRO RESERRCR June 2 0 0 7
HOW WEB BANNER DESIGNERS WORK
The assessment process does not take place only at
the end of the creative act. Instead, the designer
self-evaluates his or her message as it takes form.
banner production and specifically to thedecision-making process involved withstopping the creation process. The copy-writer or designer does not systematicallyseek to create the most influential mes-sages, but rather to create messages whosejustification will be accepted by the client.This justification is based on the aestheticand semiotic techniques that will help thebanner advertisement meet its objectives:
Of course, I do all I can to create a banner
that generates clickthroughs, but advertis-
ing is not a hard science. I am happy when
the client is happy.
CONCLUSION AND DISCUSSION
Despite the growing penetration of broad-band internet access that allows for in-creasingly rich banners, advertising onthe internet still consists of relatively sim-ple messages. In addition, internet adver-tisers often pay more attention to thegeneral format of a banner and mediaplan than to message creativity, with thegoal of getting a high clickthrough rate.We show that the designers of these ban-ners nevertheless take the same mentalapproach to their creative activity as doadvertising copywriters for traditional me-dia, and they appear as personally in-volved as copywriters (e.g., concerns aboutsel f-presenta tion).
Overall, we replicate Kover's (1995) find-ings about the existence of implicit theo-ries of communication and the use ofinternal dialogues with an imaginary au-dience to enhance message acceptance.The implicit theories elicited from banner
designers roughly correspond with thosediagnosed by Kover: attract interest andattention and then deliver the message.Some elements are, of course, specific tothe internet medium, so for example, ac-ceptance means a clickthrough. We alsofind elements, like memorization of themessage, that Kover does not describe,but that may be relevant for traditionaladvertising as well. In general, we diag-nose mental representations with moredimensions than those described by Kover(1995), and though some additional di-mensions may be linked to the specifici-ties of internet advertising, others may bedue to the more advanced set of ques-tions we use.
Unlike Kover (1995), we observe multi-ple dialogues with different imaginary au-diences: the internet user, the advertiser,two imaginary colleagues, and the aspira-tional alter-ego of the designer. In addi-tion, we show how the use of intrapsychicdialogues and implicit theories fit into theoverall creative process, as described inexisting theories about creativity. A third,more practical difference from Kover'swork is our finding that banner designersseem much more concerned than copy-writers about the advertiser's evaluation.In our context, the most significant otherin the internal dialogues is the advertiser,and getting approval is much more im-portant than it was for copywriters (Koverand Goldberg, 1995), possibly because mostbanner designers meet their clients per-sonally. In contrast, copywriters rarely in-teract directly with the advertiser, butrather rely on an account executive to
serve as an intermediary. Most bannerdesigners also must personally presentand defend their work to the client. Fi-nally and possibly because we find theinvolvement of five different audiences,unlike Kover's respondents, our infor-mants talk about the compromises theyhave to make.
We also add two important insights totheories on creative processes. First, eval-uations are not restricted to the end phaseof the process, but occur throughout. Veryearly in the creative process, banner de-signers consider the possible reactions oftheir imaginary audiences and develop arationale to get their creative ideas ac-cepted. These simulated reactions, basedon a set of evaluation criteria developedearly in the design process, often in theform of an ideal-type banner, seem tostimulate creativity and the search for newideas. Second, banner designers seem moreactive and deliberate in their search forassociations and analogies than creativitytheories would suggest. Overall, internaldialogues and evaluations of ideas aboutand incremental versions of the bannersintertwine completely, and the evaluationstandards are inseparable from the differ-ent imaginary audiences, because eachmember of a target audience embodies aset of evaluation criteria (Courbet andFourquet-Courbet, 2005).
We also recognize possible limitationsof our study. For example, we interviewFrench banner designers, and though ad-vertising as a business is global, advertis-ing cultures have national dimensions.French advertising, for instance, often aimsto entertain the audience with symbolism,humor, and drama (Zandpour, Chang, andCatalano, 1992), whereas American adver-tising usually focuses on databased argu-ments about product benefits. Because theytry to obtain different types of audienceimpacts, creative designers from differentcultures may operate differently. But it is
June 2 0 0 7 JOURRHL OF HRUERTISinG RESERRCR 1 8 9
HOW WEB BANNER DESIGNERS WORK
also possible that our findings may trans-fer to other cultures, because our studydescribes how implicit theories about com-munication operate in the form of imagi-nary dialogues that reveal how idealizedothers react to incrementally developedmessages. Creative persons in differentcultures certainly may follow a processwith the same structure and form, even ifthe content of the dialogues and the re-spective importance of the different part-ners differ. Examining cultural differenceswould, in any case, be an interesting topicfor further research.
Although banner designers resist for-mal models and theories of their workingmethods, our study may help them take astep back and reflect on their personalhabits and practice to move their work toa superior level. Most work in isolationand remain rather secretive about theirworking methods; some may not evenrealize that there is a method to theirwork. Understanding how designers workalso may help advertisers and accountexecutives develop more productive rela-tionships with them. Finally, we hope thisstudy will be used for the education andtraining of those involved in communica-tion. Our most surprising result is proba-bly the finding that there exists a creativeprocess for generating convincing mes-sages that is independent of the mediumand type of message. That is, banner de-signers are not that different from cre-ative writers in a traditional advertisingworld.
MARIE-PIERRE FOURQUET-COURBET (Ph.D.) is an associ-
ate professor of communication science at the Univer-
sity of Avignon (France). Her research interests
inciude the infiuence ofthe media (television, inter-
net) and of different types of communications (adver-
tising, poiiticai communication, olfactive marl<eting).
She also studies the strategies and representations
of the communication designers. She has written over
25 articles for schoiarly journals or books in commu-
nication, marketing, or sociai psychology and is co-
author of La Tetevision et ses Inftuences (2003,
Bruxeiies: De Boeck).
DiDiER CouRBET (Ph.D.) is an associate professor of
communication science at the University of Nice-
Sophia Antipoiis (France). His research interests focus
on the psychoiogicai processes underiying creativity,
impiicit measures of advertising persuasiveness, and
advertising in a sustainabie deveiopment context. He
has written over 30 articles for schoiarly journals or
books in marketing communication or sociai psychol-
ogy He is author of Puissance de ta Tetevision (Paris:
L'Harmattan, 1999) and co-author of La Television et
ses influences (Bruxeiies: De Boeck, 2003).
MARC VANHUELE (Ph.D., management. UCiJ\) is an
associate professor of marketing and an associate
dean for research at the HEC Schooi of Management,
Paris, France. His research focuses on price cognition,
CRM, and the effects of ambient advertising.
REFERENCES
ADAMS, R. www.advertising: Advertising and Mar-
keting on the World Wide Web. New York: Watson-
Guptill Publications, 2003.
CHANDON, J.-L., M . S. CHTOUROU, and D. R.
FoRTiN. "Effects of Configuration and Expo-
sure Levels on Responses to Web Advertise-
ments." [ournal of Advertising Research 43, 2
(2003): 217-29.
CouRBET, D., and M.-P. FOURQUET-COURBET.
"Les processus psychologiques lors de la crea-
tion publicitaire." Hermes 41 (2005): 67-74.
EL-MURAD, J., and D. C. WEST. "The Definition
and Measurement of Creativity: What Do We
Know?" Journal of Advertising Research 44, 2
(2004): 188-201.
FuRNHAM, A. "Commonsense Theories of Per-
sonality." In Everyday Understanding: Social and
Scientific Implication, K. j . Gergen and G. R.
Semin, eds. London: Sage, 1990.
GLASER, B. G., and A. L. STRAUSS. The Discovery
of the Grounded Theory: Strategies for Qualitative
Research. New York: Adline, 1967.
JoHAR, G. V., M. B. HoLBROOK, and B. B. STERN.
"The Role of Myth in Creative Advertising
Design: Theory, Process, and Outcome." jour-
nal of Advertising 30, 2 (2001): 1-25.
KIRSNER, K., C . SPEELMAN, M . MAYBERY, A.
O'BRIEN-MALONE, M . ANDERSON, and C. MAC-
LEOD, eds. Implicit and Explicit Mental Processes.
Upper Saddle River, NJ: Lawrence Erlbaum
Associates, 1998.
KOVER, A. J. "Copywriters' Implicit Theories of
Communication: An Exploration." journal of Con-
sumer Research 21, 4 (1995): 596-611.
, and S. M. GOLDBERG. "The Games Copy-
writers Play: Conflict, Quasi-Control, A New
Proposal." Journal of Advertising Research 35, 4
(1995): 52-62.
-, W. M. JAMES, and B. S. SONNER. "TO
Whom Do Advertising Creatives Write? An
Inferential Answer." Journal of Advertising Re-
search 35, 4 (1997): 52-65.
KRIS, E. Psychoanalytic Explorations in Art. New
York: International University Press, 1952.
LoNERGAN, D. C , G. M. SCOTT, and M. D.
MUMFORD. "Evaluative Aspects of Creative
Thought: Effects of Appraisal and Revision Stan-
dards." Creativity Research Journal 16, 2&3 (2004):
231-46.
MEDNICK, S. A . "The Associate Basis of the
Creative Process." Psychological Review 69, 3
(1962): 220-32.
1 9 0 L OF BDUEBTISIflG RESERBCH June 2 0 0 7
HOW WEB BANNER DESIGNERS WORK
MENDELSOHN, G. A. "Associative and Atten-
tional Processes in Creative Performance." jour-
nal of Personality 44, 2 (1976): 341-69.
Russ, S. W. "Creativity Research: Whither Thou
Goest." Creativity Research joumal 15,2&3 (2003):
143-45.
WEBER, M . The Methodology of the Social Sciences,
E. Shils and H. Finch, trans, and eds. New
York: The Free Press, 1949.
Moscovici, S. "The Phenomenon of Social Rep-
resentations." In Social Representations, R. Farr
and S. Moscovici, eds. Cambridge, U.K.: Cam-
bridge University Press, 1984.
MuMFORD, M. D. "Where Have We Been, Where
Are We Going? Taking Stock in Creativity Re-
search." Creativity Research journal 15, 2&3 (2003):
107-20.
REIS, H . T , and S. L. GABLE. "Event-Sampling
and Other Methods for Studying Everyday Ex-
perience." In Handbook of Research Methods in
Social and Personality Psychology, H. Reis and C.
M. Judd, eds. Cambridge, U.K.: Cambridge Uni-
versity Press, 2000.
SCHNEIDER, D . J. "Implicit Personality Theory:
A Review." Psychological Bulletin 73, 5 (1973):
294-309.
SIMON, H . The New Sciences of Management De-
cision. New York: Harper and Row, 1960.
SPIGGLE, S. "Analysis and Interpretation of Qual-
itative Data in Consumer Research." journal of
Consumer Research 21, 3 (1994): 194-203.
STERNBERG, R. ]., and T. I. LUBART. "An Invest-
ment Theory of Creativity and its Develop-
ment." Human Development 34, 1 (1991): 1-23.
WALLAS, G . The Art of Thought. New York:
Harcourt Brace, 1926.
WEBER, R. P. Basic Content Analysis, 2nd ed.
Newbury Park, CA: Sage, 1990.
YOUNG, C . E. "Creative Differences between
Copywriters and Art Directors." journal of Ad-
vertising Research 40, 3 (2000): 19-26.
ZANDPOUR, F., C . CHANG, and J. CATALANO.
"Stories, Symbols and Straight Talk: A Compar-
ative Analysis of French, Taiwanese, and U.S.
TV Commercials." journal of Advertising Re-
search 32, 1 (1992): 25-38.
ZEFF, R., and B. ARONSON. Advertising on the
Internet. New York: John Wiley & Sons, 1999.
June 2 0 0 7 JOURRHL OF RRUERTISinG RESERRCR I 9 i
HOW WEB BANNER DESIGNERS WORK
APPENDIXInterview Guide
1. Introduction to internet advertising
Whaf are fhe specificities of infernef andbanner advertising, compared with adver-tising in other media? Do the strategicobjectives differ? What are the strategiesfor media planning? What are the linkswith message creation?
2. Effects on the receiver
What is the profile of the average internetuser? Why do they surf the internet? Howdo they react to banners? How is thatdifferent from reactions to advertising inother media? What goes on in fhe mindof the internet user when he or she en-counters internet banners? How does ad-vertising on fhe internet influence itsaudience? If you forget about what istypically told to advertisers, what do youthink are the real effects of internet adver-tising for products or brands? Up to whatpoint can the client's objectives be met?
3. Message creation on tiie internet in
generai
What are, according to you, the basics ofmessage creation for the internet? Whaf isa good banner? What is a poor banner?Why? For the client, what is a good andwhat is a poor banner? How do you know?For the internet user, what is a good andwhat is a poor banner? How do you know?What characterizes a good banner designer?
4. Creations of tiie informant
Overall, what do you fry to accomplish?Do you have a creative philosophy youfollow? General rules in the way youwork? Recurring practices? A personalstyle? Can you explain this in more de-tail? How do you think about the internetuser? What reactions do you try to ob-tain? How do you reach this goal?
[Af the computer screen, the designershows some of his or her work.]
Can you show me some banners youhave created and explain them to me?[Wait for spontaneous explanations; if nonegiven, continue questioning.] What is theobjective of this banner? How did youadapt your creation and the different ele-ments of the message to the communica-tion objectives of the advertiser? What doyou think goes on in fhe internet user'smind from the moment he or she visitsthe webpage with your banner until theeffect that you aim at is obtained? Canyou analyze each of the elements of yourbanner, explain the objective of each, andits effect on fhe internet user? [If neces-sary, repeat question about objectives andhow they relate to visuals, words, linksbetween text and image, animation, move-ments, color, shapes, typography, hyper-text links, light, interactivity, organizationof different elements, etc.]
5. Protocoi: The creation process
Try to remember the moment you createdthis banner and put yourself back intothat moment. Explain in detail, step bystep, how you designed if. You can closeyour eyes if necessary.
What are the strong and weak points ofthe banner? Why? How did you "sell" thebanner to the advertiser? Which argu-ments did you develop? How did theadvertiser react?
6. The creation process usuaily foilowed
What is the information that you have inmind right before you start the creationprocess? Where does this information comefrom? What do you try to do? Why? Whois the banner aimed at? Who will evaluateit? Do you think about anyone else? Whenyou are creating, what is your way of
working? The different steps? How longdo they take? Where do the ideas comefrom? What do you say to yourself dur-ing the creation process itself? Tell me indetail. When you feel unable to create abanner or find a solution to a problemyou encounter, what do you do? Why?
7. Protocoi: A new creation by the
informant
I would like you to create a banner now.Tell me out loud what you think andwhat you say to yourself during the pro-cess. Do not censure yourself, but also donot add anything that you did not reallythink or say to yourself. Take your timefor this. You can use any material thatyou usually use.
Are there cases where you say differentthings? When is that? Why?
8. Evaiuating the banner created
When do you start seeing if your banneris a good (or poor) one? How exactly doyou make this evaluation? What do youlook at? Why? What do you say to your-self at that moment? What do you dothen? How do you know if the bannerwill be accepted by your client? How doyou know whaf impact the banner willhave on the internet user? In general,what is the interest of designing goodbanners? Why?
9. Reiationship with the advertiser
During the first briefing that you receivefrom the advertiser, what is it you try toknow and what information do you re-ceive? How do you present your bannersto clients? How do you develop yourarguments for the presentation? What arethe points you insist on? What do you nottalk about? Whaf goes on in your mind atthat moment?
1 9 2 JouRnm or HDUERTISIOG RESEHRCH June 2007