|
Ad Variation
To Vary or Not?
Research Implications of Ad Variation in Web Advertising
Sang Yeal Lee
Doctoral Candidate
115 Carnegie Bldg.
College of Communications
Pennsylvania State University
University Park, PA 16803
Email: [log in to unmask]
Tel: (814) 867-5642
Paper submitted to the Advertising Division to be considered for
presentation at the AEJMC Convention, July 30 – Aug. 2, 2003, Kansas City,
MO, USA.
STUDENT PAPER
To Vary or Not? Research Implications of Ad Variation in Web Advertising
ABSTRACT
The tremendous growth of the World Wide Web has attracted many advertisers
who want to take advantage of the medium. However, as the novelty of the
Web disappears and users become experts of the Web, declining advertising
effectiveness has become an important issue for advertisers. This article
explores the potential impact of ad variation to maintain or increase
advertising effectiveness in Web context. The article also examines
several critical variables that may have combining effects with ad
variation and proposes testable propositions for Web advertising research.
STUDENT PAPER
To Vary or Not? Research Implications of Ad Variation in Web advertising
INTRODUCTION
Over the years, ad variation (i.e., presenting different versions of ads
for a single product, instead of presenting a single version of ad multiple
times) has been frequently used in media scheduling in traditional
media. Despite the fact that many advertisers do employ ad variation in
implementing their advertising campaigns, there have been only a limited
number of empirical studies on this topic over the last several
decades. Existing research suggests that varied ad can forestall tedium
and thereby maintain advertising effectiveness (Schumann & Clemons, 1989;
Schumann, Petty, & Clemons, 1990; Burnkrant & Unnava, 1987). Some
researchers (Unnava & Burnkrant, 1991) further suggest that even under low
frequency condition, varied ad can result in higher advertising
effectiveness than the same ad repeated multiple times.
Ad variation can provide implications especially for the Web
advertising. Ever since the Web became available to the general public,
one major concern for advertisers has been the declining advertising
effectiveness. For example, banner click through rate has been
continually declining since 1994. Industry experts suggest that the banner
click through rate in the year 2001 was less than 0.5%, compared to about 4
percent in 1994 (Pagendarm & Schaumburg, 2001). Industry experts argue
that there is a phenomenon called "banner burnout," suggesting that
advertising effectiveness in terms of banner click through rate reaches
maximum point at the first exposure. The click through rate after the
first exposure, however, tends to decline rapidly and reach less than 0.5
percent at the fourth exposure (Pagendarm & Schaumburg, 2001). Thus, it
seems that repetition of ads on the Web can have far less return on
advertising effectiveness than in traditional media. In this perspective,
increasing advertising effectiveness can be a major challenge for
advertisers.
Advertisers can meet this challenge at least in two ways. For example,
they can develop advertising technologies such as crawling banners,
interactive ads, or messenger ads, which can have a positive impact on
advertising effectiveness. However, although these technologies may be
effective, they are also annoying and intimidating to users and as a
result, Web users can easily be tired of the ads. In fact, technological
solutions can collectively create a more problematic situation since users
are already bombarded by numerous ads on the Web. Further, given the
characteristics of the medium, Web users can get used to these technologies
and accordingly, the novelty effects of the technology may eventually
disappear.
Another way is to develop advertising strategies that may not be annoying
or intimidating but still effectively contributing to increase in
advertising effectiveness. One such strategy can be ad variation. Past
studies have generally shown that ad variation can reduce the tedium
effects caused by multiple exposures to the same ad and thereby maintain
advertising effectiveness, at least to some extent (Schumann & Clemons,
1989; Schumann, Petty, & Clemons, 1990; Burnkrant & Unnava, 1987).
Given the empirical evidence, ad variation can be an effective strategy
that can increase advertising effectiveness on the Web. It is important to
note, however, that one variable often interacts with other critical
variables to influence dependent variables. Accordingly, the purpose of
this paper is to explore the potential impact of ad variation on the Web
and to identify critical variables that may have combining effects with ad
variation. In this paper, frequency of exposure, product involvement and
brand familiarity are examined and proposed to positively influence the
relationship between ad variation and advertising effectiveness. Also, the
advertising environment, the Web, adds one more critical variable, namely
user's experience with the medium. Though user's experience is a newly
emerging variable in Internet advertising, existing evidence suggests that
it can have a negative impact on advertising effectiveness. Thus, user's
experience with the medium is proposed to negatively influence the
relationship between ad variation and advertising effectiveness. Finally,
based on empirical evidence, this paper provides testable propositions that
can be applied to Web advertising research. The main emphasis in this
paper, however, is on the interaction or combining effects among variables
rather than main effects.
Frequency and Advertising Effectiveness
Frequency of exposure is a major determinant of advertising effectiveness
and thus it has received a significant attention from the researchers in
advertising and marketing for the last several decades. However, there
still remains ambiguity regarding the exact function of frequency because
of a couple of reasons. First, as Pechmann and Stewart (1988) pointed out,
the functions of frequency depend on many other variables in different
contexts. Thus, isolating the impact of frequency alone may not be of much
theoretical and practical value because there are many other variables that
may positively or negatively interact with frequency. Second, the impact
of frequency often varies depending upon dependent variables. For example,
it has been widely accepted that frequency has a diminishing return on
advertising effectiveness after a certain point of exposure because of
tedium effects. After the tedium point, however, consumer's affect may
rapidly decline while consumer's memory may gradually decline if there is
additional frequency. The primary proposition of ad variation is that
varied ad can forestall tedium caused by multiple exposures to the same
ad. Thus, in order to understand the effects of ad variation, it is
necessary to look at how repetition impacts advertising effectiveness.
Theoretical Perspectives on Frequency
The general belief that frequency of exposure, i.e., repetition, will
increase memory of the message can be explained in terms of accessibility
of the information. Accessibility can be defined as the potential that
knowledge stored in memory will be retrieved in later use (Higgins,
1996). Althaus and Kim (2001) argue that accessibility effects can occur
"when a stimulus activates a knowledge construct, which temporarily
increases the accessibility of that particular knowledge construct in
memory" (p. 3). One factor that influences the accessibility of knowledge
construct is the frequency of activation (Shrum & O'Guinn, 1993; Althaus &
Kim, 2001). Specifically, frequency of activation increases the likelihood
that the activated construct will be used in future processing of
stimulus. According to the construct accessibility explanation then,
repeated exposure to the same ad can result in increase in accessibility of
the information relevant to those particular ad, and thereby increase the
likelihood that the stored information can be activated more easily.
Past research has demonstrated that frequent activation of a construct in
memory can facilitate accessibility of construct through the priming
process. For example, Higgins and King (1981) suggest that people can be
unaware of spreading activation of knowledge construct, triggered by
stimulus primes. They also suggested that the frequency with which a
construct has been activated in the past could be a determinant of
accessibility in the future. Thus, it can be argued that frequent
activation of a stimulus activates an accessible knowledge construct in
memory, which in turn primes individuals' encoding of subsequent processing
of the same stimulus.
The link between frequency of stimulus activation and priming via
construct accessibility provides a powerful explanation for frequency
effects. One theory that explains such a link is "repetition
priming." According to repetition priming, when people experience a
stimulus twice, that stimulus can be more efficiently processed at the
second occurrence. This is based on the assumption that the initial
processing of the stimulus must leave a residual trace in memory that
facilitates subsequent recognition of the same stimulus stored in memory
(Mitchell & Brown, 1988). In other words, when the same stimulus is
repeated, it acts as a retrieval cue for the earlier processing and
activates the residual trace stored in memory. As a result of repetition
priming, people respond to a repeating stimulus faster and more accurately
than when the stimulus is novel (Ratcliff & McKoon, 1996; Mitchell & Brown,
1988). Past research in psychology has demonstrated repetition priming
effects in word identification (e.g., Sloman, Hayman, Ohta, Law, & Tulving,
1988), picture recognition (Mitchell & Brown, 1988), and phonological
representation of stimulus (Pallier, Sebastian-Galles, & Colome,
1999). The implication of repetition priming in advertising context is
that, when the same ad is repeated, the components of ad (texts, graphics,
or sound) stored in memory in prior processing can facilitate subsequent
processing of the same ad.
Encoding specificity provides a similar explanation. The basic concept of
encoding specificity was proposed by Tulving and his colleagues to account
for episodic memory performance. Simply put, the hypothesis holds that a
retrieval cue will be effective only if information in the cue matches with
the memory trace of the target event at the time of its original encoding
(Tulving & Thomson, 1973). In other words, memory performance will be
enhanced when there is a close correspondence between the stimulus and the
memory acquired in prior events. Thus, according to encoding specificity,
accuracy of memory is a function of degree of overlap between information
stored from encoding and information present at retrieval (Tulving &
Thomson, 1973). Emphasizing the importance of encoding conditions at input
stage, Tulving and Thomson (1973, p. 369) provided the encoding specificity
principle as following:
"Specific encoding operations performed on what is perceived determine what
is stored, and what is stored determines what retrieval cues are effective
in providing access to what is stored."
The implication of encoding specificity hypothesis in advertising
environment is that when ads are repeated multiple times, repeating the
same ad will be better remembered than repeating different versions of the
same ads.
The impact of frequency can be explained in terms of information processing
also. Specifically, transfer-appropriate processing hypothesis suggests
that the stimulus, at least in the initial exposures, can be better
remembered if the same stimulus is repeated. That is, the processing of a
stimulus can be facilitated if it overlaps with processing of the stimulus
on a previous presentation (Ratcliff & McKoon, 1996; Schacter,
1992). Ratcliff and McKoon (1996) argue that changes in stimulus form may
"reduce the amount of priming because these changes mean that there are
differences in early perceptual processes (p. 406)" and thereby decrease
memory performance. Further, according to Young and Bellezza (1982), a new
encoding of a stimulus, such as another version of the same ad, may create
a new chunk in memory and, therefore, there is a possibility that encoding
a different stimulus can introduce interference between the first stimulus
and the second.
On the surface, given the foregoing discussion, it seems to suggest that
the ad repeated in the same format will be better remembered than the ads
in varied formats. Despite the robustness of these theories, however, they
do not fully explain the advertising response function, which involves a
non-linear response. That is, it is a well-known fact that consumers can
get tired of ads after certain level of exposure, and thereby the
effectiveness of advertising decreases. In fact, past research generally
suggests a non-linear advertising response as a function of frequency.
Frequency Models in Advertising
Over the last several decades, many models have been proposed to explain
advertising effectiveness as a function of frequency and three models seem
to appear more frequently in advertising literature.
Figure 1: Frequency Models in Advertising
Concave Curve
S-Shape Curve
Inverted U-Curve
Ad Effectiveness
Frequency
The two factor theory proposed by Berlyne (1970) suggests that the impact
of frequency is mediated by two factors: habituation and tedium. According
to this theory, repetition can take the form of an inverted-U curve in
which there are two separate opposing psychological processes operating
simultaneously - "positive habituation" and "negative tedium." Berlyne
(1970) argues that positive habituation can lead to an increase in affect
with diminishing returns of each additional exposure, while as a result of
repetitive exposures, tedium sets in, which would decrease affect rather
rapidly. Similar explanations of an inverted-U curve function of
repetition were proposed by Cacioppo and Petty's (1979) dual processing
model of attitudes and by Pechmann and Stewart's (1988) two-stage learning
model. Pechmann and Stewart (1988), for example, explain the inverted-U
curve response of advertising in terms of "wearin" and
"wearout." According to them, wearin occurs during approximately the first
three exposures, whereas at the third exposure, positive thoughts finally
outnumber negative thoughts. The second stage begins with approximately
the fourth exposure where message recipients become bored. As a result,
message recipients generate negative repetition-related thoughts,
undermining the persuasive impact of the ad.
Evidence of an inverted-U curve response as a function of frequency also
comes from psychology. Past research has also demonstrated that mere
frequency can affect attitudes and behaviors of message recipients. The
mere exposure effect (Zajonc, 1968), for example, posits that a mere
increase in frequency of exposure to stimuli would increase positive affect
toward those stimuli. A meta-analysis performed by Bornstein (1989)
indicated that the mere exposure effect is a "robust" phenomenon in human
cognition, and that preferences could be formed without conscious awareness
of preference-formation. The effects of mere exposure, however, are not
necessarily linear. Zajonc (1972) reported that there is a saturation
point where the effects of mere exposure reach a "plateau." A follow-up
investigation by Kail and Freeman (1973) supported Zajonc's study by
indicating a decrease in affective response that resulted in the inverted-U
curve formulation. In a review of more than 200 studies, Bornstein (1989)
also concluded that the data showed a non-linear effect of mere
exposure. Specifically, he argues that, after about ten to twenty
exposures, the effects of mere exposure begin diminishing, supporting an
inverted-U curve response as a function of frequency.
Krugman (1970) provides a different perspective on the effects of
frequency. He proposed the three-hit theory, which posits that an ad
reaches maximum effectiveness at the third exposure. Krugman argues that
frequency one, two and three each have "special qualities." The first
exposure is unique because it elicits "cognate response" to understand the
nature of the stimulus. The second exposure is more evaluative, and
results in "personal cognate response." The third exposure is the "true
reminder" since the viewer has already gone through his/her cognitive
process. He further argues that "there is no such thing as a fourth
exposure psychologically; rather, fours, fives, etc., are just repeats of
the third-exposure effect." Thus, the three-hit theory can be thought of
as an S-shaped response curve. Early studies indicated that attitudes,
purchase intentions as well as positive cognitive responses peak at the
third exposure (Cacioppo & Petty, 1979, 1980; Calder & Sternthal, 1980;
Belch, 1982).
Compared to these two models, researchers recently suggest that an inverted
U-curve or S-shaped curve is rare in the real world and the more frequently
observed model is a concave model (Sissors & Bumba, 1996; Cannon & Riordon,
1996). According to this model, the effects of frequency will begin to
have a diminishing return after a certain point of exposure, but after this
point, the impact of frequency would decline gradually or even maintained
if there is additional frequency. Thus, unlike the inverted-U model or
S-shaped model, the concave model does not suggest a rapid decline in
advertising effectiveness after the tedium point.
Two conclusions can be drawn from the frequency models discussed so
far. First, one commonality of these models is that there will be a point
where advertising effectiveness eventually begins to decline due to
tedium. Secondly, advertising effectiveness after the tedium point varies
depending upon dependant variables. For example, the inverted-U curve
response proposed by Berlyne (1970) and Zajonc (1972) seems to apply
primarily to affective measures such as attitude toward ad and attitude
toward brand, while the concave response model and the S-shaped response
proposed by Krugman (1970) can be more suitable for explaining the impact
of frequency on memory.
The Impact of Frequency on the Web
Most frequency models pertaining to tedium are built on an assumption that
consumers are "exposed" in a somewhat captive state with ads such as
television or radio commercials. However, this may not be a realistic
assumption in Web advertising. Because Web users have freedom to choose
content, features and spaces, they are not in a captive
state. Accordingly, the impact of frequency on the Web should be different
from other traditional media and, therefore, it should be treated
differently. Although one may argue that Web advertising can be treated as
the same as conventional media, employing conventional reach and frequency
notions (Cannon, 2001), given the fact that the Web provides distinctive
medium characteristics such as interactivity, the Web should be treated
qualitatively differently from traditional media (Pavlou & Stewart, 2000).
Evidence of this "qualitative difference" can be found in industry
research. For example, industry experts generally agree that frequency on
the Web can result in so called "banner burnout," which suggests that
advertising effectiveness in terms of banner click through rate reaches
maximum point at the first exposure, but after the first exposure, the
click through rate declines rapidly, reaching less than 0.5% at the fourth
exposure (Pagendarm & Schaumburg, 2001). The notion of banner burnout
provides a clearly different explanation on the impact of frequency from
existing frequency models that suggest advertising effectiveness would
increase until a certain point of exposure. More specifically, it implies
that tedium can occur much faster on the Web than in traditional media.
Another piece of evidence comes from the medium environment. The sheer
number of ads on the Web may be incomparable to any other medium. Web
users are bombarded by numerous ads, often the same ads, every time they
log on the Web. Past research has shown that consumers can get annoyed by
the same ad repeated multiple times (Bauer & Greyser, 1968). Further, Web
ads are often embedded in an already crowded space or pop up in the middle
of the screen, resulting in interruption of user's information
processing. In this sense, annoyance or intimidation can be caused by Web
advertising in general, which in turn may have a negative impact on
advertising effectiveness of a specific ad regardless of the content or
creativity in that ad.
From this perspective, it seems that Web advertisers have to account for
tedium in advertising placing not only for their ads but also Web
advertising in general. Further, depending on the communication
objectives, advertisers often have to repeat their ads more than the level
of frequency that supposedly leads to a tedium point, where the
effectiveness of advertising may begin to rapidly decrease. Accordingly,
the need to develop strategies that can cope with the tedium can be an
important issue in Web advertising.
Effects of Ad Variation
When tedium sets in, additional frequency can result in a diminishing
return on advertising effectiveness. Advertisers can handle this situation
in several ways. For example, they can stop repeating the ad, endure
diminishing return on advertising effectiveness, or try to find strategies
to cope with the tedium. One strategy that can cope with the tedium is ad
variation. Past research has suggested that varied ads can forestall
tedium caused by repeated exposures (McCullough & Ostrom, 1974; Mitchell &
Olson, 1981; Schumann & Clemons, 1989; Schumann, Petty, & Clemons, 1990;
Unnava & Burnkrant, 1991). For example, Unnava and Burnkrant (1991)
demonstrated that varied ad executions can enhance memory for brand name
over repeated same-ad executions in print media. Furthermore, McCullough
and Ostrom (1974) found that even slightly varied ads resulted in higher
liking for the product.
According to Schumann and Clemons (1989), an ad can be varied in terms of
"cosmetic" and "substantive" variations. Cosmetic variation refers to
"peripheral" changes to the product or ad such as changes in color, print
style, format, or characterization. Substantive variation refers to
central changes to the product being advertised such as changes in the
message or arguments and/or the copy found within the ad. Schumann, Petty,
and Clemons (1990) offer more refined definitions of ad
variation. According to them, cosmetic variation means changes in
"nonsubstantive" features of the ad, while the basic product message is
kept the same. On the other hand, substantive variation can be defined as
a change in message content, i.e., arguments, attributes.
Past research has in general focused on two theoretical propositions in
explaining the effects of ad variation: encoding variability and
differential attention hypotheses. Specifically, these hypotheses have
been primarily used to explain how ad variation copes with a situation in
which tedium is likely to occur. For example, contrary to encoding
specificity and transfer-appropriate processing hypotheses discussed
earlier, encoding variability hypothesis holds that when people are exposed
to a stimulus more than once, their memory performance can be improved if
the stimulus is presented differently (Burnkrant & Unnava, 1987, 1991;
Young & Bellezza, 1982). Thus, if people are exposed to two different ads
under the same brand name, these ads will leave two different chunks of
memory traces instead of one, and the information stored may provide more
contextual cues for later retrieval. Burnkrant and Unnava (1987) argue
that "the likelihood of retrieving the stimulus is believed to be directly
related to the number of traces involving that stimulus (p. 173)." In
other words, the encoding variability hypothesis predicts that the
probability of recall is directly related to the amount of the cues
relevant to the target information (Glenberg, 1979). Accordingly, encoding
variability hypothesis suggests a superior memory in varied ad conditions
than in non-varied ad conditions.
Another alternative explanation why varied ads may be more effective than
repeating the same ad can be offered by a different level of attention that
people pay to different ads. According to differential attention
hypothesis, when people are exposed to the same ad multiple times, their
attention level progressively declines due to a decrease in motivation to
process (Unnava & Burnkrant, 1991). That is, regardless of their prior
level of attention or elaboration on the ad, people may think that "I saw
that ad before" and as a result, they will pay less attention to the ad
than when the ad is varied. Consequently, repeating the same ad can result
in inferior performance in memory due to less attention and motivation to
process the information. On the contrary, however, when an ad is varied
and presented more than one format, people pay more attention because
varied ads can be processed as different stimuli, and more importantly,
people will be less likely to experience tedium. Thus, according to the
differential attention hypothesis, varied ads can induce greater attention
that would result in superior memory.
Applying encoding variability and differential attention hypotheses,
researchers have demonstrated superior memory in ad variation conditions
(Unnava & Burnkrant, 1991; Burnkrant & Unnava, 1987; Schumann, Petty, &
Clemons, 1990). For example, Unnava and Burnkrant (1991) examined the
effects of repeating varied ad executions on brand name memory. They found
that attention and encoding variability contributed independently to brand
name memory.
It is proposed in this paper that the predictions proposed by encoding
variability and differential attention hold true in Web advertising
environment also, especially considering that tedium can occur much faster
on the Web. Thus, the first proposition is developed as following:
P1: Ad variation will have a positive impact on memory.
The potential impact of ad variation on attitudes can be explained by the
elaboration likelihood model. Compared to the same ad, when ads are
presented in a varied form, people pay more attention to those ads and can
have more motivation to process the information on the ads since varied ads
are considered new stimuli. According to the elaboration likelihood model,
the information processed under high motivation will be processed through
central route, creating positive attitudes toward a brand or
product. Based on ELM perspective, Schumann and others (1990) demonstrated
that varied ads can result in significantly positive attitudes. This leads
to the following proposition:
P2: Ad variation will have a positive impact on attitudes.
As often the case in the real world, many variables can interact with each
other to influence advertising effectiveness. Frequency of exposure is one
such variable. Frequency has been considered an important variable in
examining the impact of ad variation largely because of the fact that
multiple exposures to an ad can lead to tedium among consumers. The
primary proposition of ad variation pertaining to frequency is that varied
ad can forestall tedium and thereby delay declining advertising
effectiveness. However, what happens when the frequency is low, i.e.,
below the level of tedium point? It is argued in this paper that repeating
the same ad under low frequency condition will be more effective than
varied ads. In other words, when consumers are exposed to an ad for the
first time, it will take a certain level of exposure for an ad to wearin or
to get familiarized with the information in the ad. Accordingly, when
consumers are exposed varied ads in the initial exposures, the information
on each ad will be processed as new information, which will result in less
memory as encoding specificity and transfer-appropriate processing suggest.
On the other hand, when the ad is not varied, the level of frequency or
additional frequency would not make much difference as the banner burnout
suggest. That is, frequency after the initial exposures would not make any
significant difference on advertising effectiveness. Thus, the following
proposition is developed (See Figure 1 for graphical representation):
P3: Ad variation will interact with frequency to influence memory.
Specifically, frequency will have a positive impact on memory in varied
condition while it will have no significant impact in no-varied condition.
Research has shown that after a tedium point, the level of attitudes or
purchase intentions decline while recall does not decline (Cacioppo &
Petty, 1979, 1980). This relates to industry wisdom that an intimidating
ad repeated multiple times often tends to be better remembered, although
such an ad can lead to negative attitudes. Thus, while ad variation will
have a positive impact on attitudes as frequency increases, no-varied ad
that is repeated multiple times will have a tedium effect. In other words,
while varied ads will forestall tedium in high frequency condition, no
varied ad will have a negative impact on attitudes due to tedium. This
leads to the following proposition (Figure 2):
P4: Ad variation will interact with frequency to influence attitudes.
Specifically, frequency will have a positive impact on attitudes in varied
condition while it will have negative impact on attitudes in no-variation
condition.
Figure 1:
P3 Figure 2: P4
Low Freq. High
Attitudes
No- variation
No- variation
Variation
Memory
Low Freq. High
Variation
Effects of Brand Familiarity
Brand familiarity has been generally known to have a positive impact on
advertising effectiveness. Past studies on ad variation have employed a
novel brand (Schumann, Petty, & Clemons, 1990; Motes, Hilton, & Fielden,
1992), a fictitious brand (Unnava & Burnkrant, 1991), or an existing brand
(Burnkrant & Unnava, 1987). Past studies, however, have not fully
accounted for the impact of brand familiarity as a potentially moderating
variable, although it can have moderating effects on ad variation.
Compared to a novel brand, a familiar brand can provide a mnemonic
advantage since people have pre-existing information associated with a
familiar brand. For a familiar brand ad, therefore, people could readily
associate the information in the ad with the information they have already
acquired through prior ads, or product experience. However, when the
brand is novel, there will be no prior knowledge or memory trace specific
to that brand since a new brand will be completely a new stimulus and, as a
result, it would be inappropriate to apply encoding variability and
differential attention hypotheses for a novel brand. This argument
provides an important implication regarding the impact of brand familiarity
on ad variation. That is, when the brand is novel, repeating the same
information will be more effective than varying ads as encoding specificity
and transfer-appropriate processing suggest. Thus, the following
proposition is developed (Figure 3):
P5: Ad variation will have a positive impact on memory when the brand is
familiar, while ad variation will have no significant when the brand is
unfamiliar.
Similar predictions can be made for attitudes. Specifically, for a
familiar brand ad people could readily associate the information in the
varied ad with the information they already have and therefore more
information would lead to more favorable attitudes. For an unfamiliar
brand, however, ad variation would make no difference since varied ads
would be processed as new information. Thus, the following proposition is
developed (Figure 4):
P6: Ad variation will have a positive impact on attitudes when the brand is
familiar, while ad variation will have no significant impact when the brand
is unfamiliar.
Figure 3:
P5 Figure 4: P6
Familiar
No-Variation Variation
Attitudes
Familiar
Memory
Unfamiliar
Unfamiliar
No-Variation Variation
Effects of Product Involvement
Product involvement has been an important variable in advertising research
because of its impact on advertising effectiveness. Product involvement
has been examined from various perspectives such as personal relevance
(Zaichkowsky, 1985), brand choices (Bei & Heslin, 1996), enduring
involvement (Bloch & Richins, 1983), commitment (Beatty, Kahle, & Homer,
1988) and product knowledge (Lichtenstein, Bloch, & Black, 1988). For
example, research has demonstrated that people who are exposed to an ad
under high personal-relevance condition tend to evaluate products based on
strengths of argument in ads, while people under low personal-relevance
condition tend to rely on peripheral cues such as attractiveness (Petty,
Cacioppo, & Schumann, 1983).
When product involvement is high, people will pay more attention to the
relevant information and their level of motivation to process that
information will increase (Burnkrant & Sawyer, 1983; Clarke & Belk,
1978). Zaichkowsky (1985) suggests that low involvement would lead to a
relative lack of information seeking behaviors while high involvement would
lead consumers to be more interested in acquiring information about the
brand and product attributes. From this perspective, product involvement
can lead consumers to greater elaboration on the information contained in
ads. Accordingly, the information processing on varied ads will results in
more thoughts about the brand and product, which will lead to greater
memory and attitudes. However, when product involvement is low, people
will pay relatively less attention to the information and hence, less
motivation to process information. Therefore, ad variation would not make
any difference under low involvement conditions. This leads to the
following two propositions (Figures 5 & 6):
P7: Ad variation in high product involvement condition will have a positive
impact on memory, while ad variation in low involvement will have no
significant impact.
P8: Ad variation in high product involvement condition will have a positive
impact on attitudes, while ad variation in low involvement will have no
significant impact.
Figure 5:
P7 Figure 6: P8
Memory
High Involvement
Low Involvement
High Involvement
No-Variation Variation
Low Involvement
Attitudes
No-Variation Variation
User Experience as a Moderating Variable for Advertising Effectiveness
Unlike other traditional media such as television and radio where message
recipients have less freedom to choose information, the Web provides users
with unprecedented freedom to choose content, time, and even duration of
information exposure at their own will. This is an important
characteristic of the Internet that differentiates Internet "users" from
"audiences" in traditional media. In fact, the Internet adds a peculiar
individual-difference variable, namely user's experience with the medium,
which can have a significant impact on advertising effectiveness. For
example, as the Internet gets older and more users become "experienced"
users, some users display a tendency to avoid banners on the
Internet. Benway and Lane (1998) investigated whether Web users had a
tendency to avoid banners in a task-oriented environment. They found that
the banners not only failed to grab users' attention, but also banners were
frequently ignored. Using an eye-tracking device, Dreze and Hussherr
(1999) also found that Internet users avoid banner ads. They found that
experienced users paid significantly lesser attention to banner ads than
novices.
The tendency among Web users to avoid banner ads can be explained by
cognitive theories. Automaticity theory, for example, posits that people
acquire skills as a function of repetitive information and operating
processes (Bargh, 1994). Automaticity theory also suggests that memory (in
terms of explicit and implicit) of stimuli is a key factor for people
acquiring automatic skills (Bargh, 1994). Furthermore, automatic
processing typically develops when people deal with the same stimulus
consistently over many trials. Thus, Web users' processing of information
and browsing skills may become automatic as the Web gets mature, or as an
individual becomes an experienced user. As a result, experienced users
would automatically know where the banners and the information they want
are located in the Webpages. One important point here is that product
advertising is not the information web users are looking for when they log
on to the Web: they want contents that may satisfy their information
needs. Thus, while experience users will effectively screen out ads,
novice users will have less ability to screen out ads. As a result, while
ad variation would not make significant difference for experienced users,
it will make a positive impact for novice users. Thus, the following
proposition is developed (Figure 7):
P9: Ad variation will have a positive impact for novice users on memory.
Figure 7: P9
Low experience
Memory
No-Variation Variation
High experience
Conclusion
Given the fact that tedium can be a more serious problem in Web advertising
environment, it is an imperative to develop strategies that can effectively
cope with tedium. While technology oriented ads can be effective in the
initial stage, they can be annoying and intimidating to users and thus may
not be an effective tool in the long run for advertisers.
This paper explored the role of ad variation in Web advertising
environment. Essentially, based on encoding variability and differential
attention hypotheses, it is proposed that ad variation can have positive
effects on advertising effectiveness. The impact of ad variation, however,
can be moderated by other variables. For example, building on encoding
specificity and transfer-appropriate processing, it is proposed that
frequency can interact with ad variation to influence advertising
effectiveness, suggesting that when the level of frequency is low, ad
variation would have a negative impact on advertising effectiveness. It is
also proposed that that brand familiarity and product involvement can
positively moderate the impact of ad variation. Finally, it is proposed
that user's experience with the medium can negatively influence the impact
of ad variation.
Clearly, there could be many other variables that may positively or
negatively moderate the impact of ad variation on advertising
effectiveness. In providing solutions for declining advertising
effectiveness on the Web, it is important to note that while technology
oriented solutions may create more annoyance and negative attitudes toward
Web advertising in general, user driven solutions could be less annoying
but still effective.
Table 1. Theoretical Explanations on Ad Variation
Theory
Key Proposition
Condition
Outcome
Encoding Specificity
Memory will be enhanced when there is a close correspondence between the
stimulus and the existing memory on that stimulus.
No-Variation
Positive impact on memory
Transfer-Appropriate Processing
The processing of a stimulus can be facilitated if it overlaps with
processing of the stimulus on a previous presentation.
No-Variation
Positive impact on memory
Encoding Variability
Memory performance can be improved if the stimulus is presented differently.
Variation
Positive impact on memory
Differential Attention
People pay more attention when the stimuli are different.
Variation
Positive impact on memory
Mere Exposure
Mere exposure to a stimulus will increase affect toward that stimulus.
No-Variation
Positive affect
ELM
Varied ad will be processed through central route
Variation
Positive affect
Sources:
Althaus, S., & Kim, Y. (2001, April). Modeling the impact of news discourse on
Presidential approval: A reassessment of priming effects during the gulf
war. Paper presented at the annual meeting of the Midwest Political Science
Association. Retrieved October 10, 2001, from
http://spcom-asus.ad.uiuc.edu/Users/salthaus/gulfprime.pdf.
Bargh, J. A. (1994). The four horsemen of automaticity: awareness,
intention, efficiency,
and control in social cognition. In Wyer, R. S, & Srull, T. K. (Eds.), The
Handbook of Social Cognition, Vol 1: Basic Processes, 2nd Edition, (pp. 1-
40). Hillsdale, NJ: Lawrence Erlbaum Associates.
Bauer, R., & Greyser, S. (1968). Advertising in American: The consumer
view. Boston,
MA: Harvard University Press.
Beatty, S., Kahle, L., and Homer, P. (1988). The involvement commitment
model: Theory
and implication. Journal of Business Research, 16, 149-167.
Bei, L., and Heslin, R. (1996). The consumer reports mindset: Who seeks
value: The
involved or the knowledgeable? In M. Brucks and D. J. Macinnis (Eds.),
Advances in Consumer Research, 24, 151-158.
Belch, G. (1982). The effects of television commercial repetition on
cognitive response and
message acceptance. Journal of Consumer Research, 9, 56-65.
Benway, J. P., & Lane, D. M. (1998). Banner blindness: The irony of
attention grabbing on
the World Wide Web. Proceedings of the Human Factors and Ergonomics Society
42nd Annual Meeting, 1, 463-467.
Berlyne, D. (1970). Novelty, complexity, and hedonic value. Perception and
Psychophysics,
8, 279-86.
Bloch, P., & Richins, M. (1983). A theoretical model for the study of
product importance
perceptions. Journal of Marketing, 47, 69-81.
Bornstein, R. (1989). Exposure and affect: Overview and meta-analysis of
research, 1968-
1987. Psychological Bulletin, 106, 265-289.
Burnkrant, R., & Sawyer, A. (1983). Effects of involvement and message
content on
information-processing intensity. In Richard J. Harris (Ed.), Information
Processing Research in Advertising. Hillsdale, NJ: Lawrence Erlbaum
Associates, pp. 43-64.
Burnkrant, R., & Unnava, H. (1987). Effects of variation in message
execution on the
learning of repeated brand information. Advances in Consumer Research, 16,
173-176.
Cacioppo, J., & Petty. R. (1979). Effects of message repetition and
position on cognitive
response, recall, and persuasion. Journal of Personality and Social
Psychology, 37, 97-109.
Calder, B. J. & Sternthal, B. (1980). Television Commercial Wearout: An
information
processing view. Journal of Marketing Research, 18, 173-6.
Cannon, H. (2001). Addressing new media with conventional media planning.
Journal of Interactive Advertising, 2 (1). Retrieved, February 6, 2002,
from http://jiad.org.
Cannon, H. M., & Riordan, E. A. (1996). Beyond effective frequency:
Advertising media
planning in an era of integrated marketing communications. In G. B. Wilcox
(Ed.), Proceedings of the 1996 Conference of the American Academy of
Advertising, (pp. 28-32).
Clarke, K., & Belk, R. (1978). The effects of product involvement and task
definition on
anticipated consumer effort. In H. Keith Hunt (Ed.), Advances in Consumer
Research, 5. Ann Arbor, MI: Association for Consumer Research, 313-318.
Dreze, X., & Hussherr, F. (1999). Internet advertising: Is anybody
watching? Retrieved
April 1, 2001, from http://sbaxdm.usc.edu/Publications/advertising-new.pdf.
Glenberg, A. M. (1979). Component-level theory of the effect of spacing of
repetitions on
recall and recognition. Memory & Cognition, 7, 95-112.
Higgins, E. T. (1996). Knowledge activation: Accessibility, applicability,
and salience. In E.
Tory. Higgins & A. W. Kruglanski (Eds.), Social psychology: Handbook of
basic principles (pp.133-168). New York: Guilford.
Higgins, E., & King, G. (1981). Accessibility of social constructs:
Information-processing
consequences of individual and contextual variability. In N. Cantor & J.
Kihlstrom (Eds.), Personality, Cognition, and Social Interaction (pp.
69-121). Hillsdale, NJ: Erbaum.
Kail, R. V., & Freeman, H. R. (1973). Sequence redundancy, rating
dimensions and the
exposure effect. Memory and Cognition, 1, 454-458.
Krugman, H. E. (1972). Why three exposures may be enough. Journal of
Advertising
Research, 12 (6), 11-14.
Lichtenstein, D., Bloch, P. & Black, W. (1988). Correlates of price
acceptability. Journal of
Consumer Research, 15 , 243-252.
McCullough, J., & Ostrom, T. (1974). Repetition of highly similar messages
and attitude
change, Journal of Applied Psychology, 59, 395-397.
Mitchell, D., & Brown, A. (1988). Persistent repetition priming in picture
naming and its
dissociation from recognition memory. Journal of Experimental Psychology,
14, 213-222.
Mitchell, A., & Olson, J. (1981). Are product attribute beliefs the only
mediator of
advertising effects on brand attitude? Journal of Marketing Research, 18,
318-332.
Motes, W., Hilton, C., & Fielden, J. (1992). Language, sentence, and
structural variations in
print advertising. Journal of Advertising Research, 32 (5), 63-77.
Pallier, C., Sebastian-Galles, N. & Colome, A. (1999). Phonological
representations and
repetition priming. In Proceedings of the 6th European Conference on Speech
Communication and Technology 4, pp. 1907-1910, Budapest, Hungary. Retrieved
October 14, 2001,
from
http://cogprints.soton.ac.uk/documents/disk0/00/00/09/28/cog00000928-00/p035.pdf.
Pagendarm, M., & Schaumburg, H. (2001). Why are users banner-blind? The impact
of navigation style on the perception of Web banners. Journal of Digital
information, 2 (1). Retrieved, January 5, 2002, from
http://jodi.ecs.soton.ac.uk/Articles/v02/i01/Pagendarm/.
Pavlou, P., & Stewart, D. (2000). Measuring the effects and effectiveness
of interactive
advertising: A research agenda. Journal of Interactive Advertising, 1 (1).
Retrieved, January 5, 2001, from http://jiad.org.
Pechmann, C., & Stewart, D. (1988). Advertising repetition: A critical
review of wearin and
wearout. Current Issues and Research in Advertising, 11, 285-329.
Ratcliff, R., & McKoon, G. (1996). Bias effects in implicit memory tasks.
Journal of
Experimental Psychology: General, 125, 403-421.
Schacter, D. (1992). Understanding implicit memory: A cognitive
neuroscience approach.
American Psychologist, 47, 559-569.
Schumann, D., & Clemons, D. (1989). The repetition/variation hypotheses:
Conceptual and
methodological issues. Advances in Consumer Research, 16, 529-534.
Schumann, D., Petty, R., & Clemons, D. (1990). Predicting the effectiveness
of different
strategies of advertising variation: A test of the repetition-variation
hypotheses. Journal of Consumer Research, 17, 192-201.
Shrum, L. J., & O'Guinn, T. (1993). Processes and effects in the
construction of social
reality. Communication Research, 20, 436-471.
Sissors, J. (1983). Confusions about effective frequency: Semantics and
measurement
problems. Journal of Advertising Research, 22 (6), 33-37.
Sissors, J., & Bumba, L. (1996). Advertising Media Planning. Lincoln, IL:
NTC Business
Books.
Sloman, S., Hayman, C., Ohta, N., Law, J., & Tulving, E. (1988). Forgetting
in primed
fragment completion. Journal of Experimental Psychology: Learning, memory,
and Cognition, 14, 223-239
Tulving, E., & Thomson, D. (1973). Encoding specificity and retrieval
processes in
episodic memory. Psychological Review, 80, 352-373.
Unnava, H., & Burnkrant, R. (1991). Effects of repeating varied ad
executions on brand
name memory. Journal of Marketing Research, 28, 406-416.
Young, D., & Bellezza, F. (1982). Encoding variability, memory organization
and the
repetition effect. Journal of Experimental Psychology, 8, 545-559.
Zaichowsky, J. (1985). Measuring the involvement construct. Journal of
Consumer
Research, 12, 341-352.
Zajonc, R. (1968). Attitudinal effects of mere exposure. Journal of
Personality and Social
Psychology, 9, 1-27.
|
