New Media Interactive Advertising vs. Traditional Advertising

New Media Interactive Advertising vs.

Traditional Advertising

ALEXA BEZJIAN-AVERYDePaul University

BOBBY CALDERNorthwestern University

DAWN IACOBUCCINorthwestern University

WP nw pt+l to the editor, nri

nrmyrmms YL’III~~“L’Y, Aim

Chrnm, Aqylo Lw, imd

Chvistie Nordheilm for tlviu

helpful fwdbuck 011 this

rrrmrrxvivt

This research explores the effectiveness of interactive advertising on a new mediumplatform. Like the presence in industry and the media themselves, the academicresearch stream is fairly new. Our research seeks to isolate the key feature ofinteractivity from confounding factors and to begin to tease apart those situations forwhich interactivity might be highly desirable from those situations for which traditionaladvertising vehicles may be sufficient or superior.

We find that the traditional linear advertising format of conventional ads is actuallybetter than interactive advertising for certain kinds of consumers and for certain kindsof ads. In particular, we find that a cognitive “matching” of the system properties(being predominately visual or verbal) and the consumer segment needs (preferringtheir information to be presented in a visual or verbal manner) appears to be critical.

More research should be conducted before substantial expenditures are devoted toadvertising on these interactive media. These new means of communicating withcustomers are indeed exciting, but they must be demonstrated to be effective onconsumer engagement and persuasion.INTERACTIVE MARKETING SYSTEMS are enjoying ex-plosive growth, giving firms a plethora of ways ofcontacting consumers (e.g., kiosks, Web pages,home computers). In these interactive systems, acustomer controls the content of the interaction,requesting or giving information, at the attribute-level (e.g., a PC’s RAM and MHz) or in terms ofbenefits (e.g., a PC’s capability and speed). A cus-tomer can control the presentation order of theinformation, and unwanted options may be de-leted. The consumer may request that the informa-tion sought be presented in comparative table for-mat, in video, audio, pictorial format, or in stan-dard text. Increasingly, customers can also orderproducts using the interactive system.

These new media are no fad, and while they areonly in the infancy of their development, they arealready changing the marketplace (cf. Hoffmanand Novak, 1996). The hallmark of all of these newmedia is their irlteuactivity-the consumer and themanufacturer enter into dialogue in a way not pre-viously possible.

Interactive marketing, as defined in this paper,is: “the immediately iterative process by whichcustomer needs and desires are uncovered, met,modified, and satisfied by the providing firm.” In-teractivity iterates between the firm and the cus-tomer, eliciting information from both parties, andattempting to align interests and possibilities. Theiterations occur over some duration, allowing thefirm to build databases that provide subsequentpurchase opportunities tailored to the consumer

(Blattberg and Deighton, 1991). The consumer’s in-put allows subsequent information to be custom-ized to pertinent interests and bars irrelevant com-munications, thereby enhancing both the con-sumer experience and the efficiency of the firm’sadvertising and marketing dollar.

As exciting as these new interactive media ap-pear to be, little is actually known about their ef-fect on consumers’ consideration of the advertisedproducts. As Berthon, Pitt, and Watson (1996)state, “advertising and marketing practitioners,and academics are by now aware that moresystematic research is required to reveal the truenature of commerce on the Web” or for interactivesystems more generally. Our research is intendedto address this need, and more specifically to focuson the effects of interactivity. We investigate inter-active marketing in terms of its performance inpersuading consumers to buy the advertisedproducts.

We wish to begin to understand whether inter-active methods are truly superior to standard ad-vertising formats as the excitement about the newmedia would suggest. Alternatively, perhaps thereare some circumstances for which traditional ad-vertising is more effective. Certainly it would notbe desirable to channel the majority of one’s ad-vertising resources toward interactive media untilthey are demonstrated to be superior persuasionvehicles. To this end we present an experimentalstudy comparing consumer reactions to productsadvertised through an interactive medium with re-

July l August 1998 JflURnflt OF tlOUERTlSlllG RESERRCH 23

INTERACTIVE VS. TRADITIONAL ADVERTISING

actions to products advertised in a moretraditional, noninteractive format.

CLEAN COMPARISON OF INTERACTIVETO TRADITIONAL ADSComparing interactive to traditional ad-vertising poses the difficult problem of ob-taining a valid, apples-to-apples compari-son. If the advertising differs in incidentalways, it will be difficult to attribute differ-ences to the media per se. The interactivepresentation thus should be no more high-tech than the traditional. Nor should thetraditional have superior production val-ues. Most of all, the basic product infor-mation presented should be the same inboth cases.

To obtain a valid comparison then, it isnecessary to create interactive advertisingthat differs from traditional advertising onthe core dimension of interactivity but isthe same in other respects. We conceptu-alize this dimension of interactivity as fol-lows. Interactivity is fundamentally theability to control information. Whereas intraditional advertising, the presentation islinear and the consumer is passively exposed

to product information, for interactive ad-vertising, the consumer instead actively

traverses the information. The pieces of in-formation the consumer sees depends onwhere the consumer wants to go from onestep to the next.

Depending on the design of the interac-tive system, many modes of traversal arepossible. One major possibility is the hier-archical tree organization for traversinginformation through which decisions aremade at branching points that determinesubsequent pathways. Traversal heremeans making choices at every branchpoint, as with asking people if they wishto see books or music, then fiction or non-fiction, then mystery or romance, etc. A de-sign based on hierarchical traversal is usedin this study to implement interactivity.

Traditional advertising can be concep-tualized in a parallel way in order to fo-cus, for research purposes, on its lack ofinteractivity. With traditional advertising,the consumers have no control over theorder in which they are exposed to infor-mation. The traversal mode is a simple,linearly ordered string. Ads for productsare presented one after the other in a lin-ear flow with consumers reading or view-ing predetermined ordered sequences ofinformation.

The linear flow of traditional advertis-ing media such as TV and print stand incontrast to the design of any interactivesystem. We focus in this research on a hi-erarchical traversal system and contrast itwith the linear. Other designs for interac-tive advertising (circular, networked, etc.)are of course possible and should be pur-sued in future research. All interactive de-signs, however, stand in the same contrastto the linear. Specifically, in any interac-tive format, the user has greater controlover the traversal order and the resultingsubset of information presented.

The excitement over new interactivemedia has implied that interactive sys-tems should always be superior to tradi-tional advertising. This expectation appar-ently translates into the hypothesis thathierarchical (or any other interactive) in-formation traversal is inherently superiorto the linear flow of product information.But when viewed at the level of informa-tion traversal it seems to us entirely rea-sonable to ask why this should necessarilybe the case. Might the linear traversal oftraditional advertising sometimes be bet-ter? And, if so, the best question wouldseem to be, in what cases does the addedcomplexity of the interactive system inter-fere with comprehension and persuasion?

To explore the possibility whether inter-active advertising is always superior totraditional advertising, or to begin identi-fying when traditional advertising might

be superior, we consider two additionalfactors in this study. One is a psycho-graphic personality characteristic of con-sumers. It relates to how the consumermentally represents the world-whetherthe consumer thinks “in pictures” or “inwords.” A picture or visual orientation isthought to involve a relatively more com-plex comprehension process. Thus, we an-ticipate that interactive advertising mightbe more poorly suited to these “visualconsumers” because it adds to the com-plexity of their comprehension task.

Another factor that should be importantthat we examine is the nature of the ad-vertising message itself. Advertisementsthemselves can be more visual or moreverbal. Reasoning along the same lines asabove, it would seem that more visual ad-vertisements may be more demandingand hence might actually yield betterperformance delivered via traditionaladvertising.

To summarize, in this study we willcompare interactive advertising imple-mented in a hierarchical information tra-versal design to the linear flow of tradi-tional advertising. We will equate all otheraspects of the advertising beyond this con-ceptual difference in order to obtain avalid comparison. We further examine thevisual-ness and verbal-ness of both con-sumers and the advertisements. Specifi-cally, we will measure and classify our re-spondents as relatively more visual ormore verbal using the Childers, Heckler,and Houston (1985) scale. And we willpre-test (and re-verify) our advertise-ments to be either highly visual or highlyverbal. Results will be based on typicalmeasures of recall and recognition and at-titude and purchase intentions.

THE INTERACTIVE MEDIUM EMPLOYEDThe hierarchical information traversal in-teractive advertising design used in this

24 JllURRRL OF ROUERTISIRR RESEARCH July . August 1998


study was implemented on an Apple per-sonal computer. A program was writtenfor the PC that simulates an interactivesystem. Consumers had control over theirviewing experience. They clicked iconbuttons and chose the products for whichthey wished to see advertisements.

The linear information traversal of tra-ditional advertising also employed the PCfor the sake of comparability. But the pro-gram simulated a viewing session morelike television viewing or reading a maga-zine in which the selection of advertise-ments and their ordering was predeter-mined, i.e., linear. As a reviewer pointedout, the vividness of the PC screen makesit likely more engaging than a true printadvertisement in a magazine; but wemean the comparability in the focus on thelinearity of both systems-one advertise-ment follows another and the consumerhas no input over the flow.

Thus both the interactive and the tradi-tional advertising were conducted on thePC screen. The common format was usedto facilitate comparison across the keyconceptual dimension of interactivity (notconfounded with other factors such as thenovelty of the computer for advertising).Our concern was not so much to replicateexactly any one traditional advertisingmedium but to provide a setting in whichwe could compare interactive versus lin-ear, with no other confounding differencesuch as medium. Our consumers experi-enced the traditional advertising muchlike low preproduction value TV advertis-ing using slides. The important point forthis study was that this noninteractive ad-vertising experience was comparable tothat for the interactive except for the cru-cial fact of interactivity. The key point isthat the presentation of products and ad-vertisements and their order was prede-termined in a linear sequence for the non-interactive advertising, in contrast to theinteractive advertising.

. ..,........_......... . I , I , ,Y,... ‘.~I’.-I’I-I’I~.‘l’.~j’.~l’l~.‘l~.’.,._.;;_.;,._ ~._.~~~.~.~._.~.~._.~.~~~.~._.~.~._.,~_.~._.~.~._. . .. , * , :.-. ~-~~:::~~:::~:-,

,,,., .,.,,.........-.............I’_‘.-.~.~.‘.‘I,‘,~,‘_‘,‘_‘,~.‘.~.‘_~.-.~.~ .:..‘.~.‘.‘.‘.‘.‘,‘.‘.~.~.‘.‘.~.‘.‘.,‘.~_‘_‘.‘_‘.~,‘.‘, . ;;,-,‘_‘,,‘.‘.‘_~.‘.‘.~,‘_~.-_‘.-,‘_~,‘_‘.:::::. ‘.‘.‘.~.‘_‘.‘_‘.::::::::::~.:_::l::::::j::::::::.

‘.‘_‘.-.‘.‘.-.‘_‘_ .~_‘.~_‘.-.‘.‘.‘.‘.-. a.:,-:,‘. -: ~.~.~,-.~_-.‘_~.~_~_..-. ::. .‘. , . .

‘_‘.~_~.~_‘,-.‘_~.‘_‘.-.‘_-.‘_~.~,‘_-.~_‘.-.‘.-‘_‘_‘.~_‘_‘.~.‘_~. . . . _ ,,..,. . . S@h3d~tig~ &*$ $c++ ;! ; I ; I I I ; I; :;_‘.-.‘_~:. . ..‘,‘_‘,-,’ ‘,.,‘,‘. .‘_~_~.~_‘.~.‘,~.‘_‘.-.‘_ .‘_‘. .‘.‘.‘.‘.‘_‘:, ::,-_‘_‘.~.‘_‘.~.’ ._ , . , , ~‘~~~‘~‘~-~‘~~~‘~~~-~‘~‘~‘~‘~‘. ::::::.,._.,.,._.,.,._.,._.,._._.~._.,-,._.,._. .,.. I I , ,‘.~.‘I~.~.~_‘_‘,- , , :. _‘.~.‘.~.‘_‘.~.~.-, , ,.I ,,.. . . _ _ . 5 . ,

.‘_‘_-.~.‘.~.~.‘_ . _ , , , , ,‘_‘.‘, .‘_‘.‘.~.~.‘.~.‘_~.‘.~.-.‘.~.‘_” ““.‘.’- -:: .‘.‘.-.‘.~.‘.~.‘.-.‘.‘.~.‘.-.‘,‘.~.’.‘.~.~.’.~.‘.‘.‘.‘.‘.’.’.’.. . . . . . . . . . I.....-:::. .‘.-.‘.~.~.‘.‘.-.‘_‘.~.‘_‘.~.~.~.~_’.’.’.-.’_’_~_‘. , 8..-_ _ _ ,,._.,_,._.,.,_,._.,.,.~.,.~,_. , , , , , . ._ . .-_‘.‘,~_‘.~.~_‘.-.‘.‘.‘.‘.-.‘_‘.’_’.-_’.-.’.’.‘_’ ’ ” ”.

range Juice freshly squeezedon the day of delivery

More nutrients than any otherorange juice including Thiamineand Carotene

Only the finest Florida RoyalOranges are used

Figure 1 Screen Images of Advertisements for “Less Visual”Products A and B

HOW THE ADS WERE CONSTRUCTED and drinks was administered to a smallAND SELECTED test sample of 17 consumers. The com-A pretest of 31 stimulus advertisements puter program permitted the respondentsconstructed from art photographs of foods to view the advertisements for the prod-

July . August 1998 dififlit Of ROUEATISIIIG RESERRCH 25


ucts at their own pace. Each advertise-ment showed a picture and varyingamounts of advertising copy text. Likertscales were used to measure the followingstimulus properties: product appeal, ap-petite appeal, novelty, how visual the ad-vertisement appeared, and how verbal theadvertisement appeared.

On the basis of this pre-testing, prod-ucts were selected to be approximatelyequally appealing in order to diminishany effect of initial preferences or affect.Products were also chosen that were mod-erately novel-not so new that respon-dents spent a great deal of time ponderingthe new offering and not so familiar thatrespondents made associations to familiarbrands.

Based on the extent to which the stimu-lus advertisement was considered to bevisual or verbal in conveying information,four target stimuli (products A, B, C, andD) were chosen. The advertisements forproducts C and D were perceived to berelatively more visual in nature than forproducts A and B; e.g., the product shotsfor C and D cover the entire screen and arericher in color and detail. Products A andB have less appetite appeal, as the productshots were smaller and featured productsthat appeared healthier than products Cand D. Black and white images from thecolor system of products A and B areshown in Figure 1; images for visual prod-ucts C and D appear in Figure 2.

THE EXPERIMENTAL INTERACTIVEADVERTISING EXPOSURE EXPERIENCEAn interactive shopping program wasconstructed to advertise the products.There were two product categories, bev-erages and desserts, chosen for their fa-miliarity and simplicity. These categoriesincluded our four focal products and fourfiller products.

The research participants consisted of96 people from a cross section of ages who

Figure 2 Screen Images of Advertisements for “Visual”Products C and D

were contacted in a restaurant in the mid- system at their own pace, clicking iconwest. They were recruited with a cash lot- buttons to proceed. Respondents in the in-tery as an incentive. teractive condition were given sequential

Participants navigated the interactive choices about the products that they

w

dtcITdLlS

ttt

26 JOURRRL Of RDUERTlSlRG RESEARCH July . August 1998


would see. Respondents in the linear con-dition had no choices and hence the sys-tem was not interactive. Each advertise-ment for an item included a picture anddetailed advertising copy. A single prod-uct image was projected per screen. Thesystem recorded the system choices thatthey made, the time spent on each adver-tisement, and the time spent choosing thetype of advertisement to peruse.

After looking at all the advertisements,participants were asked on-line to list thenames of all the products that they hadseen. Following this initial recall task, re-spondents’ affect toward each brand andadvertisement was measured using affec-tive and cognitive semantic differentialscales. The items measuring attitude to-ward the product are shown in Figure 3,and those that measure attitude towardthe advertisement appear in Figure 4. Par-ticipants then rated their likelihood to

Now, try to remember the ADVERTISEMENT for American Apple(candy apple). To the best of your ability, rate the

ADVERTISEMENT on these dimensions:

not persuasive 0 1 0 2 0 3 0 4 0 5 0 6 0 7 persuasive

unappealing 0 1 0 2 0 3 0 4 0 5 0 6 0 7 appealing

badO1020304050607good

unattractive 0 1 0 2 0 3 0 4 0 5 0 6 0 7 attractive

not clear 0 1 0 2 0 3 0 4 0 5 0 6 0 7 clear

unconvincing 0 1 0 2 0 3 0 4 0 5 0 6 0 7 convincing

simple 0 1 0 2 0 3 0 4 0 5 0 6 0 7 complex

overall disliking 0 1 0 2 0 3 0 4 0 5 0 6 0 7 overall liking

Figure 4 Screen Image Showing the Attitude Toward theAdvertising Measures

r-Try to remember the PRODUCT American Apple (candy apple).

To the best of your ability, please rate the PRODUCT AmericanApple on these dimensions:

bad 01020304050607good

dislikevery much 0 1 0 2 0 3 0 4 0 5 0 6 0 7 likevery much

low quality 0 1 0 2 0 3 0 4 0 5 0 6 0 7 high quality

unhealthy 0 1 0 2 0 3 0 4 0 5 0 6 0 7 healthy

awful 0 1 0 2 0 3 0 4 0 5 0 6 0 7 nice

boring 0 1 0 2 0 3 0 4 0 5 0 6 0 7 interesting

negative 0 1 0 2 0 3 0 4 0 5 0 6 0 7 positive

inferior 0 1 0 2 0 3 0 4 0 5 0 6 0 7 superior

I

Figure 3 Screen Image Showing the Attitude Toward theProduct Measures

purchase each of the items by distributing100 points across the four products in acategory.

Finally, we measured respondents on avariety of personality measures. Likertscales measured a consumer’s predilec-tion for understanding information in vi-sual and verbal format based on Childerset al. (1985). Items used to discern respon-dents as relatively more visual or moreverbal are presented in Table 1. Partici-pants also responded to requests for de-mographic information (age and gender).We also asked them to rate their level ofhunger. The entire experiment took about30 to 50 minutes to complete.

Thus, in this study we investigated therole of interactivity versus linear presen-tation of advertisements, advertising copycharacteristics (verbal or visual), and con-sumer psychographic orientations (verbalor visual) in the context of a shopping

July . August 1998 JOURRRL Uf RllUERTlSlllR RESERRCH 27


task. In the interactive condition, consum-ers had more control over the presentationof information than in the traditional, lin-ear condition. In the linear condition re-spondents clicked through the pre-setgroup of product advertisements in a pre-set order; they had no choice in the prod-ucts that were shown to them or the orderin which the products were shown. In theinteractive condition, respondents con-trolled the products that were shown tothem as well as the order that the productinformation was presented.

While the interactive condition allowedparticipants to experience and perceivecontrol, in reality the interactive and lin-ear conditions ran parallel advertise-ments in order to allow for comparabilityacross conditions. (If the participants inthe interactive condition had seen ad-vertisements different from those pre-sented to participants in the linear condi-tion, we would not be able to attribute dif-ferences in recall or persuasion to the

TABLE 1

modality of presentation or the differentadvertisements.)

WHAT THE CONSUMERS DIDThe dependent measures in Figures 3 and4 supported two factors, “attitude towardthe brand” and “attitude toward the ad”(X’ = 670.269, df = 16, p < .005), yieldingempirical results consistent with the exist-ing theoretical literature and collectivewisdom. The factor loadings are pre-sented in Table 2.

Purchase intentionUsing the respondents’ scores on theChilders et al. (1985) scale, we classifiedeach consumer as relatively more visualor more verbal and found that those per-sons scoring higher on visual-processingstyles indeed tended to rate lower on ver-bal processing and vice versa (X’ = 5.778,p = ,016).

Figure 5 shows the results on purchaseintentions for the more visual products (C

Psychographic Items Used to Discern Respondents withVisual or Verbal Orientations*

Visual Orientation. . . . . . . . . .1. I like to daydream.. . . . . . . . . . . .2. My thinking always consists of mental images or pictures.. . . . .3. When I’m learning something new I’d rather watch a demonstration than read how to

do it.. . .4. I generally prefer to use a diagram than a written set of instructions.. . . . . . .Verbal Orientation. . . . . . .1. I prefer to read instructions about how to do something rather than have someone

show me.. . . .2. I can never seem to find the right word when I need it. (reverse scored). .3. I prefer activities that don’t require a lot of reading. (reverse scored). . . . . . . . . .4. I enjoy doing work that requires the use of words.. . . . . . . . . .“These questions ore deriwdfrom Childrvs, Heckler, and Houston (1985). This swq dmonstrntes a velinbihty coef,f,cient

a = 0.88.

28 JuURnf!t Of ADllERTlSlIlG RESfRRCH July . August 1998

and D). The findings can be interpretedeither from the point of view of the re-spondent orientation, relatively verbal orvisual, or from the point of view of thelinear versus interactive contrast.

To take the former perspective, the re-spondent orientation varies on the hori-zontal axis-our participants were classi-fied as either predominately verbal or vi-sual in their orientation and preference forthe presentation of information. Verbalparticipants (those at the left of the plot)who saw the advertisements presented ina linear controlled manner did not indi-cate (statistically) differently purchase in-tentions compared with those verbal con-sumers who saw the advertisements in aninteractive manner. This finding is per-haps sensible, considering that a domi-nant feature of interactive systems is theirpictorial nature. Verbal persons are notseeking photos, but rather information viatext, and appear to simply be ignoring thegraphics content.

In contrast, the visual consumers (thoseat the right side of the plot) were different.Persons preferring visual informationwho saw advertisements in the linear (tra-ditional advertising) format stated morepositive purchase intentions than the vi-sual consumers who saw the advertise-ments via the interact ive system(F,,,, = 5.56 p = .0210). This finding sug-gests that consumers seeking visual in-formation find it more positively persuad-ing using the traditional linear vehicleformat.

Alternatively, Figure 5 can be examinedby comparing the linear versus interactiveprofiles in the plot. Purchase intentionsare enhanced in linear presentations forvisual (versus verbal) consumers, and, incontrast, they are enhanced somewhat ininteractive presentations for verbal (ver-sus visual) consumers. Thus, these resultsindicate that interactive systems are not


TABLE 2Factor Analysis Results*

Attitude toward Attitude toward

the Brand the Advertisement. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ................................bad-good product 0.893. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ................................like-dislike-verymuch product 0.958. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..................................low-quality-high-quality product 0.739. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ................................awful-nice product 0.828. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ................................boring-interesting product 0.505. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...................................negative-positive product 0.846. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..............................inferior-superior product 0.724. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...................................not persuasive-persuasive advertisement 0.832. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..............................unappealing-appealing advertisement 0.703. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..................................bad-good advertisement 0.734. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...................................unattractive-attractive advertisement 0.658. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..................................not clear-clear advertisement 0.640. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...................................unconvincing-convincing advertisement 0.835. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..................................overall liking-disliking advertisement 0.336 0.631. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ................................“Only thosr landings > 0.3 UYE noted

7

26

24

22

20

18

16

14

+ linear

+ interactive

relatively relativelyverbal visual

Respondents: Verbal vs. Visual Psychographic Orientation

Figure 5 Interactive Effect of Interactivity and VisualOrientation of Respondent on Purchase Intention for Highl:Visual Products (C and D)

uniformly superior. For visual persons, in-teractivity provides less effective persua-sion than linear advertising presentations.

It is important to note that if interactivesystems were always superior to tradi-tional linear advertising formats, Figure 5would have shown the interactive lineabove the linear line, for both sorts of per-sons. Instead, verbal persons appear to beunaffected by the system (linear or inter-active), at least for these visual products.Furthermore, visual persons demon-strated more positive purchase intentionsunder the linear, noninteractive system.

Thus we might begin to conclude thatindeed, sometimes interactive systems en-hance preference and persuasion, andsometimes traditional linear formats do asuperior job. In particular, certain system-user matches may be optimal; the result inFigure 5 reveals an interaction between vi-sual skill and interactivity. For highly vi-sual people, there was a significant in-crease in purchase intention for highly vi-sual products when information waspresented linearly. In contrast to the inter-active condition, this increase did not oc-cur for people with less visual orienta-tions. It appears that the rather verbal taskof making choices in the interactive con-ditions seems to have truncated visualprocessing. For low visual persons, limit-ing visual processing had no effect be-cause they were not likely to draw fromtheir visually oriented systems; instead,their purchase intentions were more influ-enced by the text.

Thus, there are at least some circum-stances for which interactivity does notenhance advertising effectiveness. How-ever, our explanation, regarding the vi-sual person’s processing being inter-rupted, could use more investigation. Tothis end, we examine the process measureof how much time each person spent ex-amining each advertisement.

July . August 1998 JOURRRL Of ROUERTISIRR RESEARCH 29

INTERACTIVE VS. TRADITIONAL ADVERTISING I

Time spent viewing advertisement In the interactive system, users spent less time viewingWe found that for products with highlyvisual advertisements, participants spent the advertisements, and they were less likely to pur- Imore time looking at products presentedin a linear format (M = 6.276 seconds)than when the same advertisements werepresented in the interactive system(M = 5.256; F,,75 = 4.70 p = .0333). Evenwith the novelty of interactivity, respon-dents did not spend more time with theadvertising. Instead, they spent more timelooking at advertisements that were pre-sented in the traditional format. A re-viewer made the observation that the PCinteractive mode resembles the televisionmedium, which, by association, mighthave prompted similarly passive “veg-ging” viewing. We wish to examinewhether this amount of time spent on theadvertisements also varied with partici-pant type.

Figure 6 contains the interaction be-tween the consumers’ visual versus verbalorientation and the interactivity versuslinear presentation factors on the depen-dent variable of “time spent” on the vi-

chase target products.

sual ads (C and D; F, 7s = 5.51 p = .0215).The pattern in Figure 6 is strikingly simi-lar to that in Figure 5, which immediatelysuggests that persuasion to purchase in-tention is, perhaps not surprisingly, en-hanced with time spent considering theadvertisements.

For verbal people, the manner in whichthe advertisements were presented didnot affect the time they spent on each ad-vertisement. By comparison, for visualpeople, there was a significant decrease intime spent viewing these highly visualproducts when information was pre-sented interactively in comparison to thelinear condition.

An aspect of the interactivity, perhapsthe task of making the verbal choices inthe interactive condition, truncated highly

8

zt 7 --

‘;sEns + interactiveE$ 5 --

UJ

fi=

4relatively relatively

verbal visual

Respondents: Verbal vs. Visual Psychographic Orientation

Figure 6 Interactive Effect of Interactivity and VisualOrientation of Respondent on Time Spent on Advertisementsfor Highly Visual Products (C and D)

visual processing. Perhaps the visualneeds of the visual consumers are muchmore intense than the beauty shots in theproduct advertisements. This hypothesischallenges advertising development: thevisual needs are ultra-demanding andmust be particularly appealing in order tocapture the attention of the visual-seekingconsumer.

Overall, the results on the time depen-dent variable suggest that interactivitymay interrupt the process of persuasion.Specifically, there is evidence that visualprocessing is inhibited by the use of aninteractive system.

DISCUSSION: INTERACTIVITYINHIBITS PROCESSINGThese results suggest that something un-expected may happen to people whenthey are exposed to advertising on an in-teractive system. In the interactive system,users spent less time viewing the adver-tisements, and they were less likely topurchase target products.

Media developers talk about interactivevehicles as being inherently more interest-ing and motivating. However, our resultsindicate that the consumer can still simplybuzz right through the interactive media,paying so little attention to the advertise-ments, that the message cannot functionpersuasively. Most software and Webpages certainly stand room for improve-ment for both ease of use and aestheticquality; so perhaps as these are improvedwith more creative design, interactive for-mats will be better able to retain the atten-tion of users traversing screens and mak-

30 JflURflflt Of AllUERTlSlllG RESEARCH July l August 1998


ing millisecond decisions regarding inter-est level and relevance.

Our findings are intriguing becausethey would seem to contradict new mediahype implying an entirely beneficial rolefor interactivity with respect to memory,attitude, and purchase intention. Interac-tivity indeed can be “fine,” but consumersviewing advertisements via traditionallinear formats performed at least as desir-ably (i.e., were more thoughtful in timespent and more persuaded in terms oflikelihood to purchase) in some conditionsand better in some conditions than thosemaking active decisions to obtain infor-mation via the interactive system.

As a brief aside, consider that accompa-nying the decrease in purchase intentionfor some of the interactive conditions,there were, however, no effects of interac-tivity on attitude; thus a puzzling ques-tion: how could purchase intentionschange without effecting attitude towardthe product or ad? Persuasion is thoughtto be a process from presentation and at-tention, to comprehension, generation,and retrieval of related cognitions, toyielding and retention. Apparently whena customer uses an interactive system, thelink between retrieval and yielding to thepersuasion may be broken.

To further explain, consider evidencefrom our findings. The recall order of theproducts indicates the top-of-mind char-acter of a product. The correlation be-tween recall order and the log time spentviewing the product was significantlygreater (z = 2.193, p = 0.029) for the linear(Yhnear = 0.558) than interactive conditions(rmtcrartivr = 0.067). In the linear condition,people had spent less time on the prod-ucts that came to mind quickly. This resultimplies that users spent time consideringthe products that were presented to them.In the interactive condition, there is no re-lation between time and the top-of-mindnature of the product. From this result it

The most important implication of this research is thatsometimes interactive is not better! Under certain condi-tions, interactivity interrupts the process of persuasion.

appears that, with the interactive system,the latter stages of the persuasion processmay have been inhibited.

SUMMARY AND FUTURE DIRECTIONSThe most important implication of this re-search is that sometimes interactive is notbetter! Under certain conditions, interac-tivity interrupts the process of persuasion.For the targeted products, purchase inten-tion and the time spent viewing the ad-vertisement declined when advertisingwas interactive. Moreover, there was fur-ther evidence that in particular visualprocessing was inhibited by interactiv-ity: respondents with visual orientationsappeared to be hampered by the interac-tive system as evidenced by decreases inpurchase intention for the targetedproducts and less time spent on the adver-tisements. Respondents who were rela-tively more verbal were unaffected by theinteractivity.

These results also raise the intriguingquestions how and why the visual infor-mation display sped up the visual proces-sors’ time in the interactive condition. Per-haps the decreased time was a function ofthe very control afforded the users. Whencompanies design Web pages, they are es-sentially relinquishing traversal controlto the user. Enhancing the page’s visualappeal may allow the (visual) users tocapture the essence of the incoming infor-mation more quickly and perhaps ingreater detail. However, we should alsonote that our results on the purchase in-tentions provide a cautionary conclusionwith regard to the effectiveness of newmedia advertising.

Increasingly, Web pages and the like areoffering viewers an option to obtain theinformation sought via “text only” op-tions. This option would be an importantone to study in subsequent research, be-cause it would seem to align a format op-timally with the verbal consumers’ prefer-ence for information presented in awholly verbal manner. The fact that ourfindings indicated that the interactivemethod was nearly irrelevant to the verbalconsumers is quite sensible given that theoverall impact of much interactivity is therich visual quality. If the graphics werepared down so that the content of the ver-bal information were comparatively moresalient, verbal consumers may have spentmore time on the ads and hence have beenmore persuaded to purchase.

In terms of limitations, our sample sizewas somewhat small. We would like tosee our findings confirmed on a largerscale before generalizing more broadly. Inaddition, in hindsight, we regret not mea-suring the respondents’ level of experi-ence with interactive media. Frequent us-ers of such systems might indeed have dif-ferent preferences and needs, so a scaletapping interactive experience would pro-vide a potentially useful covariate for fu-ture research.

There may be implications regardingother demographic and psychographicvariables as well. For example, gender hastraditionally been linked to visual/verbalabilities and processing preferences, so wemight expect the more typically verbal-prone users (e.g., women) to benefit morefrom, or at least not get distracted from,verbal interactivity than the more visual

July . August 1998 JOURRRL OF RDUERTISIRG RESERRCH 31


users (e.g., men). It is also known thatnovices to a category spend more time indetailed comparison, whereas expertshave already formed useful heuristics,and these differences should also havecounterparts in time spent on interactivesystems, as well as providing some resis-tance to truncated processing for the ex-perts, for example.

While we focused our investigation onvisual versus verbal-oriented people,there are many other relevant aspects ofcognitive customization (e.g., expertise) toexplore in future research. There are alsopresumably many attributes of the adver-tisements themselves that should be in-vestigated further.

The research, like the industrial use andthe media themselves, is fairly new. Manymore paths of inquiry must be followedbefore we can state simply what consum-ers are doing when confronted with inter-active formats. Until further investigationis conducted, it is wise to caution expen-ditures devoted solely to advertising onthese interactive media until they aredemonstrated to be superior persuasionvehicles, or at least not comparativelydenigrative. However, we do know:

l Sometimes interactive media do notperform as well as traditional, linear adpresentations.

l Whether the interactive method is as“effective” depends on two things:

1. whether the consumer prefers in-formation presented in a visual orverbal manner, and

2. whether the advertising content isinherently visual or verbal inimpression.

l The “effectiveness” of the interactivemedia can be measured in two ways:

1. its engagement, i.e., do consumersspend more (enough) time consid-ering the advertisements, and

2. its persuasiveness, i.e., do consum-ers report stronger positive af-fect, preferences, and purchaseintentions.

Taken together, interactive media appearto be particularly constraining for verbalpersons. A cognitive “matching” of thesystem properties (being predominatelyvisual or verbal) and the consumer seg-ment needs appears to be critical.

Investigating additional properties ofthe interactive system and consumer psy-chographic factors can only enhance thefuture effectiveness of interactive adver-tising. The new media are indeed exciting,and the potential is rich for shaping mes-sages and formats for these vehicles of thef u t u r e . a

ALEXA BEZJIAN-AVERY IS assistant professor of

marketing at DePaul University. She received her

Ph.D. from Northwestern University In 1997. She

conducts research on and teaches interactive

marketing management.

BOBBY CALDER is the Charles H. Kellstadt

DistInguIshed Professor of Marketing at the Kellogg

Graduate School of Management at Northwestern

University. His research interests include consumer

behavior, information processing. and interactive

marketing media. He teaches courses on consume!

behavior and communicating with customers.

DAWN IACOBUCCI is professor of marketing at the

Kellogg Graduate School of Management at

Northwestern University, where she teaches marketing

research service marketing. Her Ph.D. is in

quantitative psychology from the University of llllnols

at Urbana-Champaign. Her research Interests include

models for networks, and service interactlons.

REFERENCES

BERTHON, PIERRE, LEYLAND Pm, and RICHARD

WATSON. “The World Wide Web as an Adver-tising Medium: Toward an Understanding ofConversion Efficiency.” ]ournal of Advertising

Research 36, 1 (1996): 43-54.

BLATTBERG, ROBERT, and JOHN DEIGHTON. “In-

teractive Marketing: Exploiting the Age ofAddressibility.” Sloan Management Review 32,

1 (1991): 5-14.

CHILDERS, TERRY, SUSAN E. HECKLER, and

MICHAEL HOUSTON. “Measurement of Indi-vidual Differences in Visual Versus VerbalInformation Processing.” journal of Consumer

Research 12, 2 (1985): 125-34.

HOFFMAN, DONNA, and THOMAS NOVAK. “Mar-keting in Hypermedia Computer-MediatedEnvironments: Conceptual Foundations.”journal of Marketing 60, 3 (1996): 50-68.

SCHANK, ROGER. Dynamic Memory. New York:Cambridge University Press, 1982.

32 JflURnit Of RllUERTlSlNl RESERRCH July . August 1998

Documents

New Media Interactive Advertising vs. Traditional Advertising