|
Web Segmentation by Gender
Running Head: WEB SEGMENTATION BY GENDER
Not on Target:
Effects of Gender-Targeted Web Sites on Liking and Visit Intent
Shelly Rodgers
Doctoral Student
Cynthia M. Frisby, Ph.D.
Assistant Professor
University of Missouri-Columbia
School of Journalism
Department of Advertising
Contact first author:
P.O. Box 1958
Columbia, MO 65205-1958
home phone: (573) 474-9138
e-mail: [log in to unmask]
Paper submitted to the Association for Education in Journalism and Mass
Communication Conference, Advertising Division, New Orleans, 1999
Not on Target:
Effects of Gender-Targeted Web Sites on Liking and Visit Intent
Abstract
This experiment addresses the effects of gender-targeted web sites on
likability and visit intent. A 3 (male vs. female vs. neutral web site) x 2
(gender) between-subjects factorial design was used. Findings suggest that
neutral sites are preferred over gender-specific sites. In fact, both genders
rated the neutral site as more likable than either the male or female sites.
Intent to revisit the neutral site was also more likely for both genders.
Implications suggest that to get the most mileage from web sites, advertisers
should offer gender-neutral information that does not alienate either gender.
Not on Target
Effects of Gender-Targeted Web Sites on Liking and Visit Intent
Introduction
Cyber-targeting Prompts Questions
The gender cybergap is narrowing, with more women than ever becoming regular
users of the Internet. A recent survey revealed that 52% of all newcomers to the
web are women (Carey & Lynn, 1999). America Online Inc. recently reported that
its female audience has increased from 16% four years ago to 52% currently
(Machlis, 1998).
With online demographics becoming more diverse, targeting efforts are becoming
more commonplace in cyberspace (Hyland, 1999). In fact, changes in the
demographic landscape of the web are having a "profound impact" on the way
advertisers reach their target audience (Garrubo, 1999). Up to this point,
research has shown men as the dominant users of the Internet. As a result, web
sites were designed for men, and these sites were described as "content heavy"
with "little emotional appeal" (e.g., King, 1997; Kuchinskas, 1998; Peltz,
1998).
But now the tables have turned. Recognizing women as the primary decision-makers
in a household, advertisers are building and updating web sites that appeal to
the female audience. These sites offer a host of topics and products believed to
be appropriate to females, and include: beauty tips (www.covergirl.com), career
advice (www.webgrrls.com), bathing suits (www.callabay.com), facial soaps
(www.neutrogena.com), shopping (www.electra.com), relationships
(www.electra.com), fashion (www.fordmodels.com), recipes (www.kraft.com), and
games for young girls (www.happypuppy.com).
Topics that have been stereotypically males in recent years have also been
slanted to appeal to women, including: auto buyers guides (see www.elle.com),
financial information (www. cncurrency.com), and news about business and
politics (www.women connect.com). Although many male sites contain some of these
same topics, content on female sites is said to be written with a "female slant"
(Raphel, 1998). Creating an emotional experience and building close-knit
relationships are key elements that distinguish gendered sites (Eng, 1997;
Raphel, 1998). Other key elements used to attract females include chat rooms,
bulletin boards, as well as virtual shopping malls, dressing rooms and
communities (Allen, 1997; Kuchinskas, 1998; Machlis, 1998; Neff, 1997).
One such virtual community, iVillage.com, repositioned itself as a "female" site
in 1997 by designing and publishing 10 web sites to help advertisers target and
attract women users (Snyder, 1997). Using pastel colors, instead of the dark,
bold colors that were typical of male sites, was part of the strategy to make
iVillage more female-friendly. Content, which was thought to appeal to women,
was also changed to include relationship advice, health, fitness, beauty tips,
recipes and advice for working mothers (McCarthy, 1998). In an attempt to give
the site a warmer personality, photos were incorporated that featured smiling
babies, cute pets and caring mothers. The targeting strategy seemed to work
because in 1998, just one year after the "makeover," female visits to the site
had increased by 88% (Kuchinskas, 1998).
Despite its growing popularity among women, however, iVillage continues to lose
money. The company lost a reported $43.7 million last year (Afzali, 1999). In
fact, no female-targeted web site is making money right now (Wheatley, 1998).
Inability to attract advertisers has been cited as one reason for the financial
demise (Snyder, 1997). This, among other things, calls into question the
effectiveness of targeting a web site to a specific sex. Such targeting efforts
have, in past years, caused concern about alienating potential target members of
the opposite sex (Whitkowski, 1975), or producing a backlash of negative
attitudes from the targeted group (Whipple & Courtney, 1980). Anecdotal evidence
suggests that these same issues are now the concerns of online advertisers
(Hodges, 1996; Lee, 1999; Raphel, 1998).
These concerns about the impact of gender-specific web sites prompted us to
speculate just how significant gender targeting is as a factor to likability of
a web site, and how it interacts with intent to revisit a site. To explore these
issues, we formalized the following research questions.
RQ1: Does targeting by gender offend or turn off members of the targeted group
and/or members of the opposite group being targeted? Does gender targeting
alienate members of the targeted group and/or the opposite group being targeted?
RQ2: Will participants prefer "their own" gender-specific web site over each
others? Does a neutral-targeted web site have a stronger impact on participant
evaluation of the web site for both males and females?
RQ3: Which is more effective in motivating intent to revisit the site, a male-,
female- or neutral-specific site?
Literature Review
Web Sites as Advertising Vehicles
In 1996, 80% of all online ad budgets were devoted to developing web sites
(Jupiter Communications/Adspend, 1997). Fear of the unknown and fear of being
outdone by a competitor seemed to be what motivated advertisers to put up sites
in cyberspace (Raman & White, 1998). Consequently, most web sites were designed
on the basis of intuition, with little planning or execution (Ibid.).
Advertisers soon learned, however, that having a web presence meant more than
simply putting a company site. Linking and offering information that benefits
users were among the improvements for having a stronger web presence (Aldridge,
Forcht & Pierson, 1997). Search sites, like Yahoo!, and content sites, like
MSNBC, soon took shape and began to attract an audience. Based on the success of
a number of a companies (e.g., ESPN.com), web sites have shown themselves to be
effective advertising vehicles that can reach and segment a target audience
(Bush, Bush & Harris, 1998; Edwards & LaFerle, 1998).
Traditional Targeting Approaches on the Web
Reaching a target audience has been identified as one of the greatest
challenges for online advertisers (Lee, 1999). A recent survey of a national
sample of advertisers revealed that one of the top five barriers to using the
Internet as a marketing tool was difficulty in reaching and/or finding a target
market (Bush, Bush & Harris, 1998).
There is currently a debate among scholars and practitioners about whether
traditional advertising approaches will work in an online environment. Aldridge,
Forcht and Pierson (1997) argue that traditional advertising strategies to
target an Internet audience are "fruitless" and "unnecessary." Cook (1997)
argues that the web is changing how consumers participate in the marketing
communication process, in that they are both consumer and producer (i.e.,
"prosumer"). Consequently, advertisers must not rely on traditional approaches
but, instead, adapt to the "web-spun" consumer.
In a study of web site design, Masiclat and Cropp (1999) captured screen images
of actual web pages onto small cards to discern preferences in web page design.
Participants were then to sort the cards according to how attractive and useful
the sites were. What they found was a preference for sites that served a useful
purpose (e.g., search engines), with attractive sites taking second place behind
this. The authors concluded that web site designers should put their efforts
into designing useful web sites that capitalize on the strengths of the web
(i.e., interactivity), and not lean too heavily on traditional advertising
approaches (e.g., banner ads).
Other scholars disagree and argue that traditional targeting approaches are
effective to the extent that web designers partner with users to build a site
that meets their specific needs (Abels, White & Hahn, 1997). Targeting
specifications by scholars have been as broad as, distinguishing between those
interested and those not interested in a particular web site (Berthon, Pitt &
Watson, 1996), and as narrow as, segmenting based on the benefits dervied from a
product (Aldridge, Forcht & Pierson, 1997). Other scholars have suggested that
the best way to determine whether a site meets the needs of its intended
audience is to get customer feedback by way of pilot studies and focus groups
(Day, 1997).
Despite the growing number of studies that have addressed the issue of
targeting on the web, no study has empirically tested whether traditional
targeting principles work online. Thus, the current study will draw from the
traditional advertising literature to form hypotheses about what we might expect
concerning the effects of gender targeting on the web.
Traditional Advertising and Gender Targeting
Sexual identity is an important component of a consumer's self-concept. People
often conform to their culture's expectations about what that sex might prefer
or how they might behave. Of course, the standards and norms change and can
differ substantially across cultures and societies, but what remains to be seen
is how sex differences have changed. Should women prefer pink and men, blue?
And must women prefer the emotional over rational, as the stereotype would have
us believe?
Gender differences have been found for the web. An exploratory factor analysis
revealed four primary web motivations--shopping, surfing, researching and
socializing (Rodgers & Sheldon, 1999). A correlation was found for the four
subscales and gender. Females, for example, were more motivated to research and
socialize on the web, whereas males were more likely to shop. This finding
challenges the stereotype that females would prefer to shop over males, and
calls into question web sites that use shopping to attract female users.
Further, web sites that use shopping to appeal to women may, in fact, alienate
men as an important target audience.
The question, then, is not whether gender-targeted sites accomplish their goal
of attracting members of the targeted gender. Sites like iVillage.com and
women.com appear to be accomplishing this. Rather, the question addressed here
is whether gender-targeted sites alienate either of the genders, thus producing
a backlash of negative attitudes toward the brand and/or the company (and, in
this case, the web site).
Whitkowski (1975) found, for example, that female-targeted ads alienated males,
which resulted in fewer sales and public criticism.Whipple and Courtney (1980)
produced gender- targeted ads (i.e., "traditional" ads) and neutral ads (i.e.,
"progressive" ads) for five products (e.g., deodorant) and found that neutral
ads were seen as more appropriate and made the consumer more likely to purchase
the product. Neutral ads were also seen as less irritating to women and were
seen as more positive than gender-targeted ads.
It is likely that this same phenomenon will also be found for gender-targeted
web sites. Therefore,
H1: Neutral sites will be rated higher than male and female sites for
likability and intent to visit.
Studies have also found that men are more unaccepting of products that are
targeted to women than women are of mens' products (Dickens & Chappel, 1977).
For example, some women will smoke Marlboro cigarettes, typically aimed at men,
whereas men will not smoke Virginia Slims, which are typically aimed at women.
Indeed, Langer (1987) addresses this issue by saying "gender marketing in the
1980's and beyond means knowing when to address women as women, men as men, and
people as people, regardless of sex" (as cited in Bellizzi & Milner, 1991, p.
73).
Given this brief discussion, it seems reasonable to expect that:
H2: Females will be more accepting of the three web sites and will rate them as
higher than males in terms of likability and intent to revisit.
Web sites that target females have been differentiated from male sites, at
least anecdotally, on a number of variables. In some cases, women's sites are
differentiated from men's as having "frilly" content (Kuchinskas, 1998). Much of
that content is said to have emotional overtones and focuses on building
relationships (Tracy, 1997). Such a focus creates the general perception that
the site is solely for females.
Male sites, on the other hand, deal with much of the same content as female
sites (e.g., fitness, business, news, relationships, etc.), but the content is
presented in a more straight forward, less emotional context (Peltz, 1998). Past
research clearly indicates that people--regardless of gender--will respond most
favorably to advertising that captures their interests. Thus, if women truly
"want a room of their own in cyberspace," as proprietors have pointed out, then
it seems likely that they will find these same "rooms" more likable, and will
return to them more often than will men. The same may be said of men.
This point has been illustrated by a study undertaken to determine whether
males would prefer brands that were pitched as "masculine," and whether females
would prefer "feminine" brands (Alreck, Settle & Belch, 1982). Findings indicate
that men were significantly more likely to use the masculine brand and women
were more likely to use the feminine brand. In short, "each gender tended to
prefer 'their own' gender of product to that of the opposite gender" (p. 29).
Indeed, this position is supported by ownness bias, which purports that people
regard their own attitudes, interests, cognitions, behaviors and belongings as
more original and highly valued than those held by others (Perloff & Brock,
1980).
Thus, visiting a web site that is not considered "their own," should be an
unpleasant experience for that gender. Therefore,
H3: Sites targeted to a specific gender will be rated highest by that gender in
terms of likability and intent to visit.
Based on this research, we also expect an interaction between gender and the
targeted sites. Thus,
H4: Both genders will rate the neutral site highest in likability, and will be
more likely to express an intent to revist the neutral site.
To summarize, we expect to find the following:
H1: Neutral sites will be rated higher than male and female sites for
likability and intent to visit.
H2: Females will be more accepting of the three web sites and will rate them as
higher than males in terms of likability and intent to revisit.
H3: Sites targeted to a specific gender will be rated highest by that gender in
terms of likability and intent to visit.
H4: Both genders will rate the neutral site highest in likability, and will be
more likely to express an intent to revist the neutral site.
Method
Participants
One hundred and seven students (65 women and 42 men) from a large midwestern
university volunteered to participate in the study. Undergraduate and graduate
students were recruited in advertising and journalism classes, and received
extra credit for their participation.
Design
A 2 x 3 between-subjects design was employed. The first factor was gender (male
vs. female). The second factor was the gender targeted by the web site (male vs.
female vs. neutral). A balanced design was used to randomly assign web sites to
reduce the likelihood of an ordering effect.
Stimulus Materials
Three fitness magazine sites were selected for the study--a male-targeted,
female-targeted, and neutral (no gender-targeted) site. Fitness magazines were
selected because they lend themselves to being easily "gendered," and allowed us
to control product category across sites.
A lengthy two-part process was undertaken to select online fitness sites. First,
a list was compiled of possible sites from trade journals, such as Advertising
Age, and search engines such as Yahoo!. In addition, a group of 16 students were
asked to find fitness sites that they believed targeted a specific gender or no
gender at all. The final list included about 50 sites.
Each site was then visited by at least one of the authors, who, based on
anecdotal evidence in trade journals, rated the sites as male, female or neutral
(see Appendix A for list). From this list, about 40 of the sites were weeded
out, or were deemed inappropriate for the study. For example, several sites
required a user identification number, which would not work well in an
experiment for obvious reasons. Other sites were virtual storefronts that sold
health products, and had little to no content.
Second, a separate group of 17 journalism students were asked to visit the 10
sites and rate whether the site would be most appropriate for males, females or
both. Students were also asked to tell why they chose the targeted-category they
did, and provide specific examples supporting that choice. After students rated
the sites for gender specification, Scott's pi index (1955) was computed for
each of the 10 web sites. Five of the sites yielded a reliability of .82 or
higher. All five sites were revisited by the authors and the three sites chosen
were the ones that had the most in common in terms of quality content, design
and navigation (see Appendix B for URLs and examples of web sites).
Student comments about what made the site male, female or neutral were content
analyzed by one of the authors. Findings indicated that student's comments
paralleled what authors have already pointed out in trade journals (see Appendix
A). In general, the site's design, content, overall tone and advertisements were
used as cues in determining which gender would find the site more appealing.
Thus, this exercise gave us additional confidence in, not only the sites chosen
for the study, but the reasons why those sites were perceived as appropriate for
a man, woman or both.
Procedure
The study was held in a computer lab that seated about 22 people. Two computers
were always left unoccupied in case of technical difficulties. The proctor and a
computer technician arrived two hours before the experiment to standardize each
computer's settings. The screen resolution was set to 1024/768 pixels, 256
colors was selected on the color palette and refresh rate of 60 hertz was
selected. Font size was also set at a standard "large" size for every computer.
Every computer was preset with the homepage of the experiment appearing on the
screen. At the top of the homepage were the words, "Fitness Magazine Study",
with "please wait for additional instructions" in the middle, and a button at
the bottom that read "click here." The second page contained a set of
instructions and three buttons at the bottom, one button for each magazine.
The instructions on this second web page told the participants that they were to
visit the assigned magazine site much the same way they would a site of their
choosing. This was done to promote a more realistic web experience for each
participant. Participants were also told that they would be timed for seven
minutes, they were to keep their eyes on their own screen and stay in the
assigned site.
As participants arrived, they were told to sit at a computer until given
additional instructions. Once all of the participants were seated, a survey was
placed at each computer terminal. Participants were told to take note of the
button at the bottom of the survey. This represented the magazine they had been
randomly assigned to. The participants were then instructed to press "click
here" on their computer screens, they were to read the instructions carefully,
and then click on the magazine (button) they were assigned to. A proctor walked
around the room to be sure each participant understood his/her task. In only a
few cases, the proctor had to show the participant which button to click to
begin the study.
The proctor also walked around during the experiment to be sure that
participants did not link out of the site. Fifteen, half-hour sessions were held
throughout the day, and a computer crashed during three of the sessions. These
participants were moved to a new computer, and the time that expired was
approximated and a timer was reset for these individuals.
After seven minutes had expired, participants were then asked to fill out a
qusetionnaire. Two questions tested whether participants remembered the main
theme of the web site and one content item found on the front page of the web
site. Another section of the survey asked questions about the design and
navigation of the web site--the guise of the study. The questionnaire also asked
participants to rate the web site and the content on several items. Last,
participants were queried on how offensive (or appealing) they found the web
site, and their intentions to revisit the site.
Participants received a debriefing after filling out the questionnaire.
Dependent Measures
The questionnaire was written with two main criteria in mind: attitude formation
and behavioral intentions. These were measured in the following ways. Attitude
for the web sites was measured using semantic differential scales and Likert
scales. A "liking" index for attitude was calculated and consisted of how:
interesting, enjoyable, favorable, appealing and "drawn to" the site the
participant was (Alpha=.88). Scores ranged from 0 to 25, with the high value
indicating agreement with pro-site statements. The mean is 13.90, indicating a
moderately high degree of liking for the sites.
Participants' intention to revisit the site was measured using a Likert scale,
which asked the liklihood of visiting the site again. Scores ranged from 1 to 5,
with the high value indicating a high intent to revisit the site. The mean is
2.30, indicating a moderate intent to revisit the sites.
Results
Manipulation Checks
Participants rated the site for design, quality of content and ease of
navigation. No statistically significant differences were found for any of these
varibles across the three sites, meaning the sites were viewed equally in terms
of design, navigation and content.
When asked which gender would prefer the site, in almost all cases, the gender
that had been pretested for the site was chosen as the "appropriate" target
gender. In addition, 98% of the participants had never heard of any of the three
sites. Using unfamiliar sites was part of our objective so that preconceived
attitudes toward the site could be somewhat controlled.
Descriptives
The central focus of this study concerned whether gender-targeted web sites
offend members within and outside the target group, and whether such efforts
alienate either groups. To address this question, we first ran descriptive
statistics (see Table 1). As mentioned earlier, 65 females and 42 males
participated in this study. Males gave higher ratings of the web sites (M=14.05)
than did females (M=13.80). And, females had slightly higher intent to revisit
scores (M=2.31) than did males (M=2.29).
The neutral web site had the highest liking scores overall (M=16.17), followed
by the male site (M=13.37) and the female site (M=12.19). The neutral site also
elicited the highest intent to revisit scores (M=2.89), followed by the male
site (M=2.14) and the female site (M=1.89).
Hypothesis Testing
Hypothesis 1 proposed that neutral sites will be rated higher than male or
female sites for likability. This hypothesis was supported. The neutral site was
the most liked of the three sites, followed by the male site and female site (F
(2, 100) = 6.27, p <. 01). Planned Tukey's post hoc contrasts using the neutral
condition as the reference category indicated differences between the
female/neutral (p < .01) and male/neutral sites (p < .05).
Hypothesis 1 also proposed that neutral sites will be more likely to motivate
repeat visits than both the male and female sites, which was also supported. A
mean of 2.89 was found for repeat visits to the neutral site, followed by a mean
of 2.14 for the male site, and 1.89 for the female site (F (2, 104) = 6.81, p <.
01). Tukey's post hoc comparison revealed significance at the .01 level for the
female/neutral site and significance at the .05 level for the male/neutral site.
Hypothesis 2 predicted that because females are more accepting of
opposite-gender products and brands, they will rate the three sites higher than
males. This hypothesis was not supported. In fact, gender did not significantly
affect the liking of the site nor the intent to revisit the site.
Hypothesis 3 and 4 concerned the influence of the magazine site and gender on
liking attitudes and visit intent. Both hypotheses were supported. That is, each
gender rated its own site as more likable than the opposite gender, but rated
the neutral site as highest in likability overall (F (2, 100) = 4.09, p < .05).
Tukey's planned post hoc comparison test revealed significance at the .01 level
for the female/neutral site and significance at the .05 level for the
male/neutral site.
Likewise, both genders were more likely to revisit their "own" web sites, but
rated the neutral site highest in terms of revisit intent (F (1, 101) = 5.07, p
< .001). Planned post hoc comparisons revealed that females had a lower intent
to revisit the male over the neutral site (p < .01). In the same way, males were
not as motivated to return to the female site as they were their own (p <. 05).
Males were also more likely to return to the neutral over the female site (p <
.05).
Discussion and Conclusions
This study provides a picture of how gender segmenting on the web affects
consumer response and visit intent. No support was found for the notion that
females will prefer their own web sites and males, theirs. Even though men
tended to rate their own sites higher than female sites in terms of likability,
these differences were not significant. The same was true of women.
Significant support was found for the notion that neutral sites will outperform
male- and female-targeted sites. Not only did both genders like the neutral site
more, they also indicated a higher intent to return to the site.
Perhaps the most important finding was the fact that the male site was not
found to appeal to males nor did the female site appeal to females. This finding
supports the idea that our society is changing. More women are running their own
businesses and are making important economic and political decisions (Bridge,
1997). Thus, consumer lifestyles, interests, needs and wants are also changing.
To build effective web sites, advertisers will need to understand these changes
so that a stronger web presence can be made. Because web consumers tend to be
more educated and affluent, they are more likely to demand high quality content
that benefits, but does not demean, them. Most importantly,
consumers--especially women--use web sites that benefit their lives (see
Masiclat & Cropp, 1999). We suspect that web pages that overlook this point will
soon go by the wayside.
As with any study, ours has limitations. One of the limitations of the study is
that a random sample of participants was not used. Therefore, the results should
not be projected to the population as a whole--even the population of college
students. A second limitation is that only one product, fitness magazines, was
used in the study. A fitness magazine did provide an appropriate product
category that could be targeted just to females, just to males, or to both males
and females. Few product categories allow such a comparison. In addition,
selecting fitness magazine web sites allowed us to hold the product category
constant so that an apples-to-apples comparison was possible, at least in terms
of the product.
A third limitation concerns the design and content of the sites themselves.
Because we wanted the participants to have a more naturalistic experience, we
did not create our own web sites, as many experimental studies do. We went to
great lengths, however, to choose sites that were, not only equal across design,
navigation and content, but that met the objectives of the study. High
reliabilities and our manipulation check speak to this accomplishment.
The last limitation concerns the generalizability of the web sites themselves.
Even though we went to great lengths to find web sites that were perceived as
"male", "female" or "neutral," finding sites that were representative of
fitness magazines on the web as a whole was not accomplished. Then again, the
purpose of the study was to examine the relationship between genders and
attitudes toward the gender-targeted sites, and not the attitudes toward fitness
magazine sites as a whole. Therefore, a discrete sample seemed appropriate.
In conclusion, our study supports the idea that traditional targeting principles
do not work on the web--at least when the demographic, gender, is used. Other
demographics, and other segmenting approaches, such as psychographics, will need
to be tested in this regard.
References
Abels, E. G., White, M. D. & Hahn, K. (1997). Identifying user-based criteria
for web pages. Internet Research: Electronic Networking Applications and Policy,
7(4), 252-262.
Afzali, C. (1999). iVillage makes stunning debut. Internetnews.com,
http://www.internetnews. com/fina-news/article0,1087,5_83261,00.htm.
Aldridge, A., Forcht, K. & Pierson, J. (1997). Get linked or getlost: Marketing
strategy for the Internet. Internet Research: Electronic Networking Applications
and Policy, 7(3), 161-169.
Allen, D. (1997). Women are creating their own communications systems. Media
Report to Women, 25(1), 9.
Alreck, P. L., Settle, R. B. & Belch, M. A. (1982). Who responds to "gendered"
ads, and how? Journal of Advertising Research, 22(2), 25-34.
Barry, T. E., Gilly, M. C. & Doran, L. E. (1985). Advertising to women with
different career orientations. Journal of Advertising Research, 25(2), 26-36.
Bellizzi, J. A. & Milner, L. (1991, June/July). Gender positioning of a
traditionally male-dominant product. Journal of Advertising Research, 31(3),
72-79.
Berthon, P., Pitt, L. F. & Pendergast, G. (1997). Visits, hits, caching and
counting on the World Wide Web: Old wine in new bottles? Internet Research:
Electronic Networking Applications and Policy, 7(1), 5-8.
Berthon, P., Pitt, L. F. & Watson, R. T. (1996). The World Wide Web as an
advertising medium: Toward an understanding of conversion efficiency. Journal of
Advertising Research, 36(1), 43-54.
Berthon, P., Pitt, L. & Watson, R. T. (1996). Re-surfing W3: Research
perspectives on marketing communication and buyer behaviour on the Worldwide
Web. International Journal of Advertising, 15, 287-301.
Boettiger, A. (1998, September). Response to an online discussion list.
http://www.internetadvertising.org/.
Boettiger, A. (1999, March). A telephone interview between Adam Boettiger, vice
president of Eyescream, an online advertising agency, and Shelly Rodgers.
Bridge, M. Junior (1997). Slipping from the scene: News coverage of females
drops. In S. Biagi and M. Kern-Foxworth (Eds.), Facing difference: Race, gender,
and mass media (pp. 102-112). Thousand Oaks, CA: Pine Forge Press.
Bucy, E. P., Lang, A., Potter, R. F. & Grabe, M. E. (1998). Structural features
of cyberspace: A content analysis of the World Wide Web. Paper presented to the
Association for Education in Journalism and Mass Communication Conferences,
Communication Theory & Methodology Division, Baltimore, MD.
Bush, A J., Bush, V. & Harris, S. (1998). Advertiser perceptions of the Internet
as a marketing communications tool. Journal of Advertising Research, 38(2),
17-27.
Carey, A. R. & Lynn G. (1999, March). USA Snapshots: Changing the face of
Internet. A survey by Princeton Survey Research Associates for Pew Center For
the People & The Press. USA Today, March 2, 1.
Cook, W. A. (1997). Editorial: World Wide wake-up call. Journal of Advertising
Research, 37(2), 5.
Daniel, J. (1996, January). Not playing around. Texas Monthly, 24(1), 25.
Day, A. (1997). A model for monitoring web site effectiveness. Internet
Research: Electronic Networking Applications and Policy, 7(2), 109-115.
Debevec, K. & Iyer, E. (1986). The influence of spokespersons in altering a
product's gender image: Implication for advertising effectiveness. Journal of
Advertising, 15, 4, 12-20.
Dickens, J. & Chappel (1977, April). A study of the sexual polarization of food
products. Journal of Marketing Research Society, 19 (2), 76-92.
Edwards, S. & LaFerle, C. (1998). Cross media promotion of Internet sites in
television advertisements. Paper presented at the American Academy of
Advertising Conference, Lexington, KY, March 27-30.
Eng, P. N. (1997, May 5). A coffee klatch for moms and dads. Business Week, 76.
Garrubo, G. (1999). What makes women click? Internet Advertising Bureau Online,
http://www.iab.net, March.
Gentry, J. W., Doering, M. & O'Brien, T. V. (1978). Masculinity and femininity
factors in product perception and self-image. In Advances in Consumer Research,
K. Hunt (Ed.). Ann Arbor, MI: Association for Consumer Research.
Hodges, J. (1996, September 23). Robert Levitan. Advertising Age, 67(39), 42.
Hollis, N. S. (1995). Like it or not, liking is not enough. Journal of
Advertising Research, 35(5), 7-16.
King, J. (1997, February). Who's in the online pool? Computerworld, 31(6), 24.
Kottler, P (1980). Marketing Management Analysis, Planning, and Control, 4th
ed., Upper Saddle River, NJ: Prentice Hall,
Kuchinskas, S. (1998, September). It's a woman's web. Brandweek, 39(33), 46-48.
Langer, J. (1987, August). A woman is still a woman, but a man now has his
doubts. Marketing News, August 28, 1987.
Lee, T. (1999, January). Targeting customers on the Internet. The Internet
Advertising Discussion List, http://www.internetadvertising.org/, January 15,
1999.
Leong, E. K. F., Huang, X. & Stanners, P. J. (1998, October). Comparing the
effectiveness of the web site with traditional media. Journal of Advertising
Research, 38(5), 44-51.
Machlis, S. (1998, August). Women gaining notice of online retailers.
Computerworld, 32(32), 6.
Mand, A. (1998, June 8). Where the surfers are. Brandweek, 39(23), 58-60.
Masiclat, S., & Cropp, F. (1999). Designing online advertising: The search for
useful alternatives to the banner ad. Paper presented at the Annual Conference
of the American Academy of Advertising, Albuquerque, NM.
McCarthy, J. (1998, October). It takes iVillage. Shape, 18(2), 20.
Morris, G. P. & Cundiff, E. W. (1971, August). Acceptance by males of feminine
products. Journal of Marketing Research, 8 (3), 372-374.
Morton, L. P. (1998). Segmenting publics: An introduction. Public Relations
Quarterly, 43(3), 33-34.
Neff, J. (1997, November). P&G reach out to women at work. Advertising Age,
November 10, 1997, s4.
Peltz, J. (1998, December). Women join the Net. PC World Online, December 2,
1998, http://www.pcworld.com/pcwtoday/article.
Perloff, R. M. & Brock, T. C. (1980). ...And thinking makes it so: Cognitive
responses to persuasion. In M. Roloff & G. Miller (Eds.), Persuasion: New
directions in theory and research. Beverly Hills, CA: Sage.
Prakash, V. (1992). Sex roles and advertising preferences. Journal of
Advertising Research, 32(3), 43-52.
Raman, N. V. (1997). Wandering on the web. In M. M. Carole (Ed.), The
proceedings of the 1997 conference of the American Academy of Advertising.
University of Cincinnati: Cincinnati, Ohio, 1997, pp. 224-228.
Raman, N. V. & White, C. (1998). Managing the new media: What web site
decision-makers plan, know, and believe. In D. D. Muehling (Ed.), The
proceedings of the 1998 conference of the American Academy of Advertising,
Washington State University: Pullman, Washington, 1998, pp. 79, abstract only.
Raphel, M. (1998, January). Women on the Internet: How Swede it is! Direct
Marketing, 60(9), 26.
Rodgers, S. & Sheldon, K. (1999). The web motivation inventory: Reasons for
using the web and their correlates. Paper presented at the 1999 annual
conference of the American Academy of Advertising, Albuquerque, NM, March.
Sautter, S. J. & Linquist, J. D. (1998). Exploring print ads and web sites
target marketing and approach match-ups: Interactive video games and facial
cosmetics. Developments in Marketing Science, 21, 174-178.
Scott, W. (1955). Reliability and content analysis: The case of nominal scale
coding. Public Opinion Quarterly, 17(1), 133-39.
Snyder, B. (1997, September 15). New networks, site upgrades target women.
Advertising Age, 52.
Solomon, M. B. (1996). Targeting trendsetters: Cyberstudies capture the
perceptions and insights of early adopters. Marketing Research: A Magazine of
Management & Applications, 8(2), 9-11.
Spataro, M. (1998, August). Net relations: A fusion of direct marketing and
public relations. Direct Marketing, 61(4), 16-19.
Tracy, B. (1997, September). Survey says women want web sites that build
relationships. Advertising Age, September 22, 1997, 32.
Wheatley, A. (1998, April). Web watch: Ladies, start your (search) engines.
Working Woman, 23(4), 22.
Whipple, T. W. & Courtney, A. E. (1980). How to portray women in TV commercials.
Journal of Advertising Research, 20(2), 53-59.
Whitkowski, T. H. (1975). An experimental comparison of women's self and
advertising image. In R. C. Curham (Ed.), New marketing for social and economic
progress. Chicago: American Marketing Association.
Appendix A
Variables used to determine whether a site was male- or female-targeted
y Design Elements--colors (light for females, dark for males) and types of
graphics,
y Pictures - more emotional, relationship-type pics/graphics for females, fewer
for males).
y Type of Content/hotlinks, for the female sites:female: recipes/cooking,
relaxation, exercise/health,clothes, hair, career, mom-stuff, child/childcare,
household business,family time, balancing career with family, shopping, romance,
specialoffers, coupons, freebies, etc.
y Product offerings - stereotypical "female" products like perfume, skin care
products, etc.)
y Overall "tone" or "feel" of the site--emotional: females; content-heavy: males
Appendix B
Female site: www.physical.com/home/janski (reliability = 1.0)
Neutral site: www.runnersworld.com (reliability = .82)
Male site: www.boxing-monthly.co.uk/ (reliability = .96)
Table 1
Mean Scores for Likability and Intent to Visit
Independent Variables Likability SD Visit Intent SD
Gender
Female 13.80 5.04 2.31 1.21
Male 14.05 5.23 2.29 1.31
Web Site
Female 12.19 4.31 1.89 .99
Male 13.37 5.03 2.14 1.24
Neutral 16.17** 5.18 2.89** 1.30
**p < .01
|
