United We Stand

 

Brand Advertising; Online Advertising

 

Marketers have been learning online brand advertising on the run. In the early years we
didn’t really know if it even worked. Advertisers committed funds either on faith or simply
as an experiment in a promising, and highly visible, new medium. The ads themselves
reflected this experimental character, as advertisers tested the limits of technology and
communication tools. Although rules of thumb for good advertising online were sometimes
promulgated (e.g., see Carlon et al., 2000; DoubleClick, 2001), they were easily subjugated to
the enticements of newer formats and approaches.

The industry has now reached the next stage of maturity. There is compelling evidence that
online advertising is an effective branding tool (e.g., see Dreze and Hussherr, 1999; Briggs,
2001, Wakeling and Murphy, 2002). It is an accepted, if perhaps under-appreciated, part of
advertising communication. It seems time to strengthen the guidelines on how to use online
effectively.

This paper explores the effectiveness of online brand advertising. What sorts of Internet
advertising work best, for what purposes, for what brands and categories? And how can
advertisers maximize the effectiveness of their online advertising? Based on our analysis, we
offer some lessons to guide advertisers as they develop their campaigns.
Does online advertising work?
It is worth stating upfront that for online advertising, as for traditional advertising, there is
no simple formula for what works. Good advertising requires creativity, the right frequency,
targeting, etc.. Does online advertising work? It can, but not always. If you are unsure how
best to use online adverting, you are not alone. Many have probably realized that success
does not come just from using the latest formats. So, lacking confidence, some hold back.
This paper is offered as a way to understand where effectiveness can be found and we hope
the MarketNorms database will continue to be a resource to help the industry answer these
questions.

 

Background and Approach
Our analysis is based on the Dynamic Logic MarketNorms Database. As part of its market
research activities over the last few years, Dynamic Logic has gathered detailed information
on hundreds of online marketing campaigns, with over 5,000 classified creative executions,
and close to half a million respondents within those campaigns. The data span a wide range
of advertiser industries, brands, ad formats and includes advertising on thousands of web
sites, including 10 of the top 10 publishers.
This proprietary database contains information on many metrics, by creative execution and
campaign overall, but version 1.0 of MarketNorms focuses on four widely recognized
measures of branding effectiveness:

 

 

To learn more about these metrics and how they fit into the traditional hierarchy of
advertising effects, please see the Appendix.
These measures are calculated through a rigorous methodology that compares people
exposed to a campaign with a control group. The AdIndex methodology allows campaigns
to be measured as they run live on websites. This unique approach, which Dynamic Logic
helped pioneer, creates a real-world test environment of marketing in action - tested in its
actual environment as opposed to a research facility.
Essentially, the research compares the attitudes of those exposed to creative units to those
who have not been exposed. The purpose is to isolate the impact of the online advertising
on people's impressions of the brand, while controlling for any effects from offline
advertising. For more on the AdIndex methodology and detailed definitions, please see the
Appendix.

Dynamic Logic has gone to great lengths to code additional data points onto each campaign.
Additional data points include coding on the features of the brand advertised, the
product/service industry, the frequency of exposure, and various characteristics of the
creative format and execution. The MarketNorms database thus provides a unique,
comprehensive and consistent body of online advertising campaigns for our study.

Analysis Technique
We analyzed the data using multiple regression analysis. Many earlier studies have relied
primarily on bivariate analyses, essentially correlating a measure of effectiveness with each
possible explanatory variable in turn. While this bivariate approach is useful in identifying
key factors, it is possible that any one factor could be picking up the effect of another. The
multiple regressions helped us isolate the effects of each variable while controlling for all the
others. Since many of the explanatory variables tend to be correlated with one another, the
regression approach, in many respects, provides a stricter statistical test. A more detailed list
of the variables and regression results are listed in the Appendix.
We focused the analysis at the level of the campaign, rather than each creative execution
separately. This allowed us to see the effects of the overall online advertising approach.
Campaigns were flagged if they included a specific creative execution and a median
frequency level was computed. The database does enable analysis by execution and
frequency level, which will be useful for some tactical questions, and we will be exploring
this in further work.
Overall Findings
At the campaign level, the data offer a resounding reaffirmation of the effectiveness of
Internet advertising. The evidence shows statistically significant improvements in all four
key branding measures: aided Brand Awareness, Message Association, Brand Favorability,
and Purchase Intent.

 

 

 

This table shows how across hundreds of campaigns, there are statistically significant
increases in each of the four key metrics. The delta, or absolute difference, is the percentage
point increase that each metric increases on average as a result of exposure to online
advertising. For example, Message Association increases by 5.61 points on average whereas
Purchase Intent increases 1.44 points on average. This makes intuitive sense since
"awareness metrics" such as Brand Awareness and Message Association are traditionally
more sensitive to advertising impact than "persuasion metrics" such as Brand Favorability
and Purchase Intent.
So, online advertising is, indeed, effective in building brands. Our main goal, however, is to
explain what factors best explain this improvement. Our findings are summarized below.
We considered four dependent variables, each reflecting one of our four different measures
of effectiveness. The variables indicate the difference for a campaign between the control
and exposed groups. These dependent variables were then regressed on a set of possible
explanatory variables. Note: we chose a few variables for this paper based on personal
interests. Of course, other variables should and will be explored in future work.

The table below provides an overview of the regression results, indicating the effect of each
variable, holding the other variables constant. Overall, we were able to explain between 4%
and 27% of the variation in campaign effectiveness, depending on the effectiveness measure
used. While the data does not explain 100% of campaign effectiveness variance, the amount
that it does explain is significant and compares favorably to the capabilities of complex
behavioral research (Jaccard et al, 1997, see Appendix for more).

 

 

 

Industry factors and ad format are very influential determinants of advertising
effectiveness.

o Ad format plays an important role in online advertising effectiveness. This
may not surprise many people but it plays a pivotal role in how formats can
be utilized and what they are worth to both the advertiser and the publisher.
Perhaps more surprising, however, is that the industry is also a key
determinant. Certain industries, overall, are either better or worse on moving
three out of the four key metrics. This is explored further in Lesson Two on
page ten of this paper. The fact that both ad format and industry are MORE
indicative of advertising effectiveness than brand tenure, historical timing,
and frequency suggests that the multiple regression has teased out something
new - which is that different industries should utilize online advertising in
different ways and for different goals.

Over time, online advertisers have become better at building Brand Awareness.

o The fact that time (as in the historical date of the campaign) is an indicator of
effectiveness means that the multiple regression is showing that online
advertisers have improved their abilities to raise Brand Awareness. Now
there may be a certain amount of self-selection here given that brand
advertisers tend to use Dynamic Logic research more often than non-brand
advertisers. Still, if that were somewhat constant over time, than the brand
advertisers using the Internet over the last few years, as measured by
Dynamic Logic, have improved in their ability to raise awareness through
online vehicles. This is especially significant given that the types of clients
participating in Dynamic Logic research has grown populous with traditional
brand advertisers - the types of brands that usually have higher levels of
baseline awareness. This means that even big traditional brands are figuring
out ways to build brands online over and above their higher baselines. This
is discussed more in Lesson One on page nine. This is good news in that the
medium is proving to be an awareness raiser like television, but still, online
advertisers overall have not yet made statistically significant strides in the
other areas: Message Association, Brand Favorability and Purchase Intent.
Frequency of exposure is statistically significant only in explaining Brand Awareness.
For other measures of effectiveness, it has no significant explanatory value.

o The reason may simply lay the fact that showing a bad ad to someone more
often does not make the ad more effective. So while it will enhance a good
ad, it does not help a bad ad become more effective. Therefore, it is not a
determinant by itself. They may be limitations to this analysis in that it relied
on median frequency levels for a campaign. Further analysis at the

 

respondent or creative unit level may provide further insight into this
variable.
Message Association differences are explained by the tenure of the brand.
o Message Association may be strongly correlated with brand tenure because
more mature advertisers (with older brands) are often more focused in their
advertising creative on providing a very simple clear message. This is partly
due to the fact that they often have high baseline levels of aided Brand
Awareness. Therefore, this marketer is often looking to build Message
Association and further shape the brand in the consumer's mind by linking it
to a specific message, thus moving the consumer along in the hierarchy of
advertising effects (see Appendix). Traditional advertisers also seem to
recognize the value of simple, strong and consistent messaging; their
campaigns and ad creative often reflect that.
Brand Favorability and Purchase Intent differences are explained by differences in
the brand industry and the format of the ad.
o As mentioned earlier, Persuasion metrics such as these are difficult to move
in general. As explored in Lesson Two, certain industries have shown that
they may be better able to move these metrics. It is hard to determine
whether this is because that has been their aim or because online advertising
is a more effective vehicle for these specific industries and brands. Either
way, it shows up in the analysis as statistically significant and worth
exploring.

 

Analysis

Our findings can be grouped into three main lessons:
? Lesson One: Maturity Matters
All new tools take time to learn how to use effectively, and Internet advertising seems to fit
that condition. The chart below shows that ad effectiveness has increased over time for all
four of our measures. In all four cases, there is a positive correlation between the timing of
the campaign (measured in months from the start of the database) and its effectiveness.

 

 

What explains this improvement? Our analysis provides only a partial answer. In the
multivariate analysis, i.e., the regressions of effectiveness on our various explanatory
variables, the timing of the campaign is a significant explanatory variable only in explaining
Brand Awareness. For the other effectiveness measures, timing is not statistically significant
once we control for other features of the campaign.
The answer may lie in the way advertisers incorporate their learning over time into
advertising. At the simplest level, there is likely to be some advertiser self-selection, as
bad ads and bad advertisers drop out. Certainly, in the mid to late ‘90s, there were many toein-
the-water experiments that were misguided and/or poorly executed. It is also likely that
the learning shows up in more effective selection of formats. As noted earlier, there have
been many studies suggesting ad format guidelines. So good advertisers have had steadily
improving guidance. In addition, as online advertising has increasingly shifted from
communication by dot-coms to communication from more traditional brands, advertising is
being guided by a more practiced group of marketers.

Overall, the improved effectiveness reflects a maturing medium, where the fervor for
technological novelty is giving way to the essentials of brand marketing. In this more mature
environment, marketers’ creative skills once again come to the fore. Lesson Three on page
12 addresses this issue further.
? Lesson Two: Brands Differ – and So Does Their Online Advertising Effectiveness
From the beginning, advertisers have recognized that the medium’s special features were
likely to be more suitable for some types of brands and categories. The medium’s ability to
engage, to provide information, or simply to interact all means that it may work better in
some circumstances than in others. Our results confirm that effectiveness varies
considerably by brand and industry (see below).
Another significant finding is the relatively lower effectiveness of advertising at increasing
Brand Awareness for high-tenure, i.e., well-established brands. Of course by definition, such
brands have less opportunity for improvement in aided awareness. It is difficult for a Coca-
Cola type off advertiser to move Aided Brand Awareness higher from its high levels. Yet, it
is interesting that medium-tenure brands score higher than either low or high-tenure brands.
Perhaps the Internet is especially effective in building on modest awareness, reinforcing the
initial offline recognition. The medium-tenure brands may well be the “sweet spot” for
Internet advertising.
Effectiveness also varies considerably across categories. Industry classifications are
statistically significant in most of our regressions. However, the significance and the relative
impact of each industry vary depending on which measure of effectiveness is considered.
This is consistent with recent results presented by Hislop (2002) through ESOMAR.

 

 

 

As the table shows, certain industries tend to be more effective at moving certain metrics.
Automotive is very effective in the Awareness metrics. The financial services advertisers
have been more successful than other industries at moving Purchase Intent, which is a
difficult metric to move. They are all about equal in terms of ability to move Brand
Favorability - an area where they are all equally weak on average. It seems no industry has
yet mastered the ability to use online advertising to significantly move Brand Favorability
beyond the 1.17 points norm.
The standout performer in our analysis is the packaged goods (CPG) industry. Campaigns
from this industry perform significantly better than other industries in three out of four of
the regressions. This may be due to the industry’s ability, building on a long marketing
heritage, to home in on the medium’s communication possibilities. It may also be that CPG
brands are susceptible to such campaigns. In any case, the results are remarkable.
In addition to the savvy of the CPG marketing departments, the category may be easier to
sell online. The products are straightforward to explain (detergent is not as complex as
financial services or technology products). Additionally, the low-consideration level of many
of the products may make it easier to persuade consumers to take financial risks. Purchasing
a snack food that turns out not be as expected does not carry the same level of consequence
as committing one’s company to a telecommunications provider that turns out to be
unsatisfactory.

The highly considered brands analyzed have had a more difficult time increasing brand
perceptions through online advertising. This may be in part due to the high price points of
the products and services, in addition to the complexity of the offerings themselves.
Campaigns targeting business audiences (“B2B”) are rarely successful at increasing
Persuasion metrics. Additionally, there are consumer-targeted brands that also have
challenges increasing Persuasion metrics, such as automotive brands. Vehicles are very
expensive purchases and are not made very often, thus making it hard to significantly
increase Purchase Intent.
These results reflect a number of features of Internet advertising to date. Undoubtedly,
some brands and some categories have simply done a better job of adapting to the medium
and there are always winning and losing campaigns in each industry category. But, beyond
that, it is also evident that the medium does some things better than it does others, and
works better for some sorts of messaging. The brand’s stage of development, the industry,
and the advertising goals all matter, and campaign planning will need to carefully mesh these
all together.

? Lesson Three: Creativity Rules, but the Rules Are Complicated
An important paper two years ago offered five “Golden Rules” for online branding (Carlon
et al., 2000). The study reported findings and recommendations on four creative attributes:
ad unit clutter, logo size (relative to size of the unit), creative unit size and presence of a
human face. The fifth rule looked at frequency of exposure. The analysis was based on a
series of correlations between each of these five attributes and each of the six measures of
effectiveness. The current Dynamic Logic MarketNorms database provides a more
extensive set of campaigns and executions to explore these attributes. And the multivariate
approach allows us to look at the creative and campaign attributes together.
Our results do reaffirm some of the findings of the earlier Golden Rules study. They show a
positive effect of frequency, although it is only statistically significant in explaining
improvements in Brand Awareness. Larger ads (in our study, large rectangle ads) also have
a positive effect, on both Brand Awareness and Purchase Intent. Additionally, the presence
of a life form has a positive effect on both Brand Favorability and Purchase Intent. (In our
study we considered both humans and animals life forms.)
These results are reassuring and reinforce much of the common sense of the Golden Rules.
Frequency should matter, at least up to a point. Others have corroborated this result
(Briggs, 2001; Hislop, 2002). Bigger ads should also be expected to grab attention. And
the role of a life form suggests that the Internet can indeed engender an emotional appeal.
Still, looked at more broadly, our results indicate that there are very few hard and fast rules
for effective Internet brand advertising. Ads communicate successfully in a variety of ways.
There are no universal rules of content or format. Some formats, e.g., large rectangle ads, do
show up as consistently positive in impact. But other formats, even some out of favor, also

work in some instances. For example, banners seem good at increasing Message Association
perhaps because they are smaller and usually can only contain a logo and a message.
Along the same lines, we saw mixed effects of two “Golden Rule” factors: logo size and
clutter. Each of these variables flirted with significance in our analysis, and each was, on its
own, correlated with effectiveness, but neither stood out as significant once other factors
were accounted for. Keep in mind that our measure of clutter is a “clutter ratio.” It is our
absolute clutter score of words and images divided by the pixel area of the ad unit. This was
to control for larger sized ad units containing more words and images (see Appendix). It
should also be noted that when the original "Golden Rules" paper was prepared in 2000,
most of the data was from the predominant ad format, which was then the banner (468x60)
ad unit.
Overall, we looked at campaigns that contained more than a dozen formats, and numerous
features of the brand and the industry. We examined the variables on their own and
interacting among themselves. Still, most of the variation in effectiveness could not be
explained by these measured variables. This is important in that it means creative is still a
wildcard, in online advertising, just like offline advertising.
Additionally, for some measures, notably Message Association, our variables explained very
little of the variation in effectiveness. Clearly, there are harder to measure factors at work,
notably the creativity that goes into the brand communication. For example, it seems likely
that skyscrapers could be effective; but, as a new format, the creative input has often been of
very poor quality (for example, the logos are often placed at the bottom instead of the top
where they are more easily viewed, so although the ad is larger, consumers are not
connecting the creative with a particular brand).
More generally, these findings suggest the development of a rich and complex medium.
They reflect the maturation of online advertising. This is not simply a technological
medium; good, creative advertising matters online at least as much as it always has offline.
Jaffe has written recently that we should think of online advertising as an art that eschews
rigid rules (Jaffe, 2002). We agree. Successful campaigns won’t come simply by following a
rulebook, but will require a complex interaction of creative attributes and placements.

Where Next
Growth in online advertising as a medium for brand development depends on at least two
conditions: evidence that online advertising works and an understanding of how it works.
Considerable progress has been made in demonstrating its effectiveness; we now know it can
work. Yet many traditional marketers hold back because of a lack of confidence in the how:
just how will a specific campaign accomplish the brand’s specific goals. Research of the sort
we presented here is a step toward developing a more systematic and reliable approach to
campaign development.
Our three lessons do not in themselves offer a prescription for optimizing the effectiveness
of online brand advertising. Indeed, they suggest that effective online advertising requires
both careful analysis of your brand’s attributes and rigorous, high quality campaign
development.
It seems that many research papers end with a call for more research. Yet, due to the fact
that advertising effectiveness can vary widely by brand and by marketing vehicle, more
research is inevitable as each marketer, agency and publisher learns what works with their
respective target audience. Case studies are helpful, but what is effective for one advertiser
may not be true for another. Only specific repetitive research will provide data stability and
insights that make marketers smarter and more effective in this complex but compelling
medium.

Appendix
Authors and Contributors
Dr. Tony Romeo is CEO of Strategic Dynamics, which provides strategy, marketing and new
business support to companies facing the challenges of a changing marketplace. He is a
former senior executive at Unilever, where he founded and chaired their Interactive Brand
Center. He has been a Professor of Economics at London Business School and the
University of Connecticut.
Nick Nyhan is the President of Dynamic Logic, an independent research company that
specializes in measuring marketing effectiveness. His background includes traditional
qualitative and quantitative research in his work at Bozell and online research while working
at Modem Media Poppe Tyson. He has been a regular lecturer at New York University's
Stern School of Business.
Molly Hislop is Vice President of Research Services at Dynamic Logic. Andrew Latzman is
a Research Supervisor at Dynamic Logic. Amanda Salomon is a Research Manager at
Dynamic Logic. Kaii Tu is a Research Intern at Dynamic Logic and is currently a junior at
Harvard College.
References
-- “IAB Online Advertising Effectiveness Study 1997.” 1997.
http://www.mbinteractive.com/resources/reports/iab_97.html
-- DoubleClick online brand-building study. July 2001. http://iab.digit.
net/main/doubleclickmedstudy.pdf
Dreze, Xavier and Hussherr, Francois. "Internet Advertising: Is Anybody Watching?"
Working White Paper, USC Marshall School of Business, Ecole National Superieure
de Telecommunication. Aug. 1999. http://www.xdreze.org/Publications/list.html
Briggs, Rex. “The Role of Creative Execution in Online Advertising Success.” Measuring
Success: an Advertising Effectiveness Series from the IAB, (1, 4). October 2001.
http://www.iab.net/measuringsuccess/img/Creative.pdf
Wakeling, Patti and Murphy, Brian. "CPG Ad Commandments." 2002. Unilever and IRI
presentation at ARF Spring 2002 conference.
http://www.iab.net/measuringsuccess/
Carlon, Michael, Marc Ryan, and Risa Weledniger. “The Five Golden Rules of Online
Branding.” October 2000. http://www.dynamiclogic.com/DL_5gold_rules.pdf
Hislop, Molly. “Lessons Learned from the AdIndex Normative Database.” Presentation,
June 2002. http://www.dynamiclogic.com/press_release_060502.php
Jaffe, Joseph, “Calder to Action,” iMediaConnection, July 25, 2002.

Jones, John Philip, “The Mismanagement of Advertising.” Harvard Business Review, Jan/Feb
2000.
Jaccard, James and Michael Becker (1997). Statistics for the Behavioral Sciences. Third
Edition. Pacific Grove: Brooks/Cole Publishing Company, p. 140.
"In behavioral science research, where complex behaviors are studied, correlations of
.20 to .30 (and -.20 to -.30) are often considered important." Correlations of 0.20
and 0.30 are the equivalent of r-squares of 0.04 and 0.09, respectively.

 

Appendix
Hierarchy of Advertising Effects
When consumers are introduced to a brand, there are a series of sequential and cognitive
steps that often happen in the process of converting prospects into customers. To start,
consumers must become familiar with the name of the brand, and if appropriate, recognize a
logo as a symbol of that brand. Next, the brand must explain its value proposition to the
target audience. This part of the process often relies on advertising messaging to articulate
what the product or service does and how it can be of use.

One a consumer is familiar with a brand and what
it stands for, they are considered “aware”.
Building awareness is a straightforward process
relying on the communications of factual
information – names, symbols and messages. The
next stage is a greater challenge to marketer for
they must persuade the educated prospects that
the brand can be of value to them.
There are fundamentally two types of “value” that
influence purchase and usage motives (Rossiter &
Percy, 1997): Informational and Transformational.
Informational products / services take a consumer
from negative state and return them to a state of
equilibrium. Examples of this category include
pain relievers, food (satiating hunger) and
household products. Marketers selling
transformational goods must persuade consumers
that their brand solves specific problems.

 

Transformational products take a consumer from a state of equilibrium to a higher and more
positive state, and can be often labeled as rewards. Luxury cars, entertainment, vacations
and ice cream are examples of products / services that are intended to make people feel
happier, and are usually not considered things that consumers need. Marketers in this
category must entice consumers to indulge and reward themselves or their families.
There are many brands that can fall into both categories and consequently have a greater
challenge in persuading consumers. Some clothing brands, for example, are necessary items
in the minds of some consumers, and at the same time, luxury items to others. Certain
automotive brands also fall into both categories.
The first part of the persuasion process involves convincing a consumer that the brand is a
good choice. No one is going to consider purchasing a product or service if they do not
have a positive feeling towards it. Brand Favorability is a measure of how strongly
consumers feel about the brand – positively or negatively.
Once consumers have formulated an opinion about a brand, the next logical step is whether
the will engage in a desired behavior. In most cases, the behavior that the marketer is
looking to drive is a purchase or financial transaction. There are many instances, however,

 

Measuring the Attributes of Successful Branding Banners

Loyalty

where the advertising is aiming to drive a non-purchase behavior. Examples include
campaigns striving to increase visitation of a web site, test-driving of a vehicle, consultation
with a physician or watching of a network television show.
The ability of a marketer to increase intent to engage in the desired behavior often depends
on the perceived risk to the consumer. Very expensive products / services require more
consideration before a financial commitment is made because the consequence of making a
poor choice can be great. Conversely, less thought or consideration is put forth for an
inexpensive – sometimes called an “impulse-buy” – product / service as there is less risk
associated with the purchase.

Statistical Appendix
The tables below summarize the regression output. Note that:
- Although there are 350 campaigns in the database, there were missing data points for
some explanatory variables. So the regressions were run with varying numbers of
observations.
- The regressions presented include only those with variables where the null
hypothesis of no effect could be rejected at the 10% level or better.
- All four regressions are statistically significant.

 

 

Variable Definitions:
Dependent Variables:
The dependent variables were the difference between the exposed and control groups for
four variables:
Aided Brand Awareness - Measures the level of familiarity respondents have with the
brand listed.
Message / Sponsorship Association - Measures the extent to which respondents can
match the copy or messages in the creative to the brand.
Brand Favorability - Measures the extent to which respondents have a positive or
favorable opinion of the brand.

Purchase Intent - Measures the likelihood of respondents to take purchase action on the
brand in the future.
Independent variables tested:
Note that to enable the multivariate regression analysis, many of the variables below are
“dummy” variables, equal to 1 if a characteristic holds and 0 otherwise. For example “life
form” equals 1 if a person or animal is present and 0 otherwise. Within a set of multiple
mutually exclusive features, e.g., product categories, it is necessary to create multiple
dummies, but for statistical reasons, the number of dummies for any set of n features must
be n-1.
Demographic variables:
Average age: the average of ages for exposed respondents
Months from 4/20/00: number of months from 4/20/00 to campaign start date
Creative execution variables:
Product shot: 1 = product shot in campaign creative, 0 = no product shots in campaign
creatives
Percent logo average: the average of the number of frames in which logo appears divided by
total number of frames
Clutter ratio average: the average of the ratio of clutter (number of words and images in an
ad) over the pixel area of the ads for a campaign
Interactivity: 1 = campaign uses interactive ad(s), 0 = campaign does not use interactive ad
Lifeform presence: 1 = campaign uses one or more ads that has picture of human or animal,
0 = campaign does not use ads that have picture of human or animal
Creative formats:
Banner – format: 1 = campaign used banner(s), 0 = campaign used no banners (see
http://www.iab.net/iab_banner_standards/bannersource.html for description of a banner)
Button – format: 1 = campaign used button(s), 0 = campaign used no buttons (see
http://www.iab.net/iab_banner_standards/bannersource.html for description of a button)

Full page – format: 1 = campaign used full page ad(s), 0 = campaign used no full page ads (a
creative execution that consumes the space of an entire page)
Large rectangle/square – format: 1 = campaign used large rectangle(s)/square(s), 0 =
campaign used no large rectangles/squares (see
http://www.iab.net/iab_banner_standards/bannersource.html for description of a large
rectangle and square)
Skyscraper – format: 1 = campaign used skyscraper(s), 0 = campaign used skyscrapers (see
http://www.iab.net/iab_banner_standards/bannersource.html for description of a
skyscraper)
Number of format categories: the number of different types of formats used in a campaign
Technology:
gif/jpg: 1 = campaign uses a gif or jpg, 0 = campaign does not use a gif or jpg
html/flash: 1 = campaign uses an html/flash creative execution, 0 = campaign does not use
an html/flash creative execution
Rich media: 1 = campaign uses a rich media creative execution (Supersitial, bridge/minisite,
Next Generation Ad, eyeblaster, dhtml, pointroll, hybrid), 0 = campaign does not use a rich
media creative execution
Brand attribute variables:
Consumer Type B2C: 1 = B2C brand, 0 = B2B brand
Consideration High: 1 = high consideration brand, 0 = low consideration brand
New product: 1 = new product, 0 = not a new product
Tenure High: 1 = baseline awareness of the brand in 67-100 percent range, 0 otherwise
Tenure Medium: 1 = baseline awareness of the brand in the 34-66 percent range, 0 otherwise
(Tenure Low)
Top 100 Advertisers: 1 = brand is top 100 advertiser in traditional media spending according
to AdAge, 0 = brand is not top 100
Top 200 Advertisers: 1 = brand is top101-200 advertiser in traditional media spending
according to AdAge, 0 = brand is not top 101-200
(Neither Top 100 nor Top 200 Advertiser)

Traditional: 1 = traditional brand, 0 = online brand
Sector Service: 1 = brand is in the service sector, 0 = brand is in the product sector
Industry:
Industry Alcohol: 1 = brand is an alcohol, 0 = brand is another industry
Industry Automotive: 1 = brand is automotive, 0 = brand is another industry
Industry Packaged Good: 1 = brand is a packaged good, 0 = brand is another industry
Industry Entertainment: 1 = brand is entertainment, 0 = brand is another industry
Industry Financial Services: 1 = brand is financial services, 0 = brand is another industry
Industry Pharmaceutical: 1 = brand is a pharmaceutical, 0 = brand is another industry
Industry Restaurant: 1 = brand is a restaurant, 0 = brand is another industry
Industry Retail: 1 = brand is retail, 0 = brand is another industry
Industry Technology: 1 = brand is technology-related, 0 = brand is another industry
Industry Utility/Telecom: 1 = brand is a utility or telecom, 0 = brand is another industry
(Industry Travel)
Other variable:
Frequency median: the median number of times individual respondents were exposed to ads
from a given campaign

Research by Dynamic Logic

 

 

 

 

	Contact Us
	© 2002-2004 EmarketingMag is emarketing, branding, Search Engines, Information Retrieval, PageRank optimization, and web site design resources		Optimized by emarketingmag.com