-
Interactive Advertising Bureau / Dynamic Logic Ad Unit
Effectiveness Study
Research conducted by Dynamic Logic on behalf of The IAB
March - June 2001
Contacts: Interactive Advertising Bureau: Robin Webster,
President & CEO Interactive Advertising Bureau [email protected]
(917) 368-7255 Rex Briggs, IAB Research Committee [email protected]
(415) 559-9374 PR Contact: Stu Ginsburg [email protected] (917)
206-3088
Dynamic Logic: Molly Hislop, Director of Research &
Development, Dynamic Logic [email protected] (212) 591-9177
Jeffrey Graham, Director of Client Services Dynamic Logic
[email protected] (212) 591-9177 PR Contact: Michelle
Robertson [email protected] (212) 827-3748
mailto:[email protected]:[email protected]:[email protected]:[email protected]:[email protected]
-
- 2 - IAB/Dynamic Logic Ad Unit Effectiveness Study
Table of Contents
Introduction…………..………………………………………………………………………...3-4
Background…………..…………………………………………………………...……………...5 Study
Design.……….……………………………………………………………………………6
Methodology..………………………………………………….…………………………………7 Executive
Summary………………………………………………………………………….….8 Overall Results: Brand
Awareness and Message Association……………………..………9 Overall Results:
Favorability and Purchase Intent………………………………………10-11 Case Study 1 –
Genuity……………………………………………………………………12-14 Case Study 2 –
Vaniqa……………………………………………………………….……15-17 Case Study 3 –
uBid………………………………………………………………………..18-20
Glossary………………………………………………………….…………………….………..24 Appendix: AdIndex
Methodology…………………………………………..……..………25-27
-
- 3 - IAB/Dynamic Logic Ad Unit Effectiveness Study
Introduction The purpose of advertising, a word that is based on
the Latin term “advertere” meaning “turn forward,” is to orient
consumers towards the purchase of products.1 The premise of
advertising is that changing a consumer’s attitudes is a necessary
prerequisite for stimulating purchase behavior. Even the most
effective advertising does not instantly cause consumers to
purchase a product. Rather, good advertising causes incremental
shifts in a consumer’s thinking, creating awareness, fostering
association between the product and certain messages or qualities,
and finally creating intent to purchase2. Measuring the ability to
affect these attitudes provides us with brand metrics, which are
the basis for evaluating advertising success. Numerous studies have
linked attitudinal metrics such as brand awareness and brand image
with market share. Over a ten-year period, car rental company Avis
found a positive relationship between certain brand metrics and
share of market.3 Similarly, Taster’s Choice Coffee found that its
transformation into a market leader was partly based on the
strength of a popular advertising campaign.4 Simply put, it is
necessary to make consumers aware of a product, and to ensure they
care about it, before they can ever make the decision to purchase
it. This fundamental marketing principle has sometimes been
forgotten in the “click-through”-oriented world of Internet
advertising. The findings of the IAB/Dynamic Logic Ad Unit
Effectiveness Study confirm online advertising’s ability to foster
brand attitudes among consumers. In addition, the data show that
creative innovation, in this case the new, larger sized ad units,
can extend online advertising’s branding power. 1 C.J. Dirkson and
A. Kroeger, Advertising Principles, Problems and Cases, Homewood,
IL: Irwin, 1960. 2 J.R. Rossiter and L. Percy , Advertising
Communications and Promotions Management, Boston: McGraw Hill,
1997. 3 P. Georgiou and S. Miller. “10 Years of Tracking in the
Rent-A-Car Business.” Advertising Research Foundation, 1997. 4 L.
Rickard, “Taster’s Choice Rolls Love Potion Number 9,” Advertising
Age, March 28, 1994, p.70.
-
- 4 - IAB/Dynamic Logic Ad Unit Effectiveness Study
Clearly, large units themselves do not guarantee an effective
online advertising campaign. There are many factors that contribute
to a campaign’s success. Online advertising needs to provide a
relevant message to the right audience, in a manner that is not
only noticeable, but also consistent with good user experience.5 A
well-placed banner in a high-value area of a site might have more
of an impact on consumers than a large unit in an area that gets
little attention. The results of this study are reported primarily
as case studies because of the inherent and unavoidable differences
among online advertising campaigns. The results of the Coca-Cola
campaign, for example, with a very high baseline awareness level,
are not directly comparable to those of uBid, a less established,
dot-com brand. Among other factors, differences in the quality of
ad creative, the degree of consideration for the product category,
the audiences of campaign web sites, placement of creative, and
coincidence with offline marketing efforts, all can affect the
brand metrics of online advertising campaigns. Nonetheless, the
study found that the larger format sizes, which are naturally more
visible and provide for more creative freedom, did prove to be
significantly more effective than smaller, standard banners across
all campaigns. These units might develop even more of an edge in
the future, since their size accommodates more interactivity and
technological flexibility. In the end, the results of the study
should provide solid evidence for the branding power of the
Internet and provide support for continued investment in extending
creative innovation and development.
5 J. Graham, “The Two R’s of Online Advertising,” ClickZ
(www.clickz.com), December 6, 1999.
-
- 5 - IAB/Dynamic Logic Ad Unit Effectiveness Study
Background Since 1996, when the Interactive Advertising Bureau
(IAB) issued voluntary banner size guidelines for the online
advertising industry, the branding effectiveness of online
advertising had been repeatedly proven. Studies published by
independent research companies have established banner
advertising‘s ability to improve consumers’ perceptions of brands
and increase their intent to purchase products. On February 26,
2001, the IAB issued a new wave of voluntary guidelines to add to
its roster of existing banner sizes. The new units, which include
two vertical and five rectangle sizes, are designed to expand
interactivity and flexibility for online marketers. With the
branding effectiveness of standard banners firmly established, in
the spring of 2001 the IAB commissioned Dynamic Logic, an
independent research company, to explore the branding impact of the
new larger units. Dynamic Logic designed a study to compare the
relative branding effectiveness of the standard banner, Large
Rectangle, and vertical “Skyscraper” banners units. Working in
consultation with the IAB Research Committee, Dynamic Logic
partnered with four web site publishers and four advertisers to
take part in the study. The study participants were as follows:
Publisher /Advertiser
CNET Networks, Inc. / Genuity iWon / Vaniqa (Bristol-Myers
Squibb) MSN / uBid Snowball / Coke
A total of 8,750 respondents were recruited from April 26, 2001
until June 11, 2001. This report provides the findings of the
study. Please note: Coca-Cola's specific results are not released
but are in the aggregate findings.
-
- 6 - IAB/Dynamic Logic Ad Unit Effectiveness Study
Study Design Since the purpose of the study was to isolate the
relative impact of three different ad unit sizes, the three ad
units were tested across four different campaigns running on four
different web sites. Additionally, each ad unit was tested against
a campaign control group to provide comparison of each ad format
against a baseline. Other controls include limiting the frequency
of the exposure to one and the duration of time between exposure
and survey (recency) to less than one hour. The following schematic
illustrates the design of the study:
Since the ad format size guidelines are new, some flexibility
was given to publisher partners as to the size of the ads used for
the study. Skyscraper sizes included 120X600 and 160X600 pixels,
and Large Rectangle units included 240x400, 336x280, 360x300 and
300X250 pixels. Banners were all 468x60. NOTE: Each campaign is
reported in this document as a case study. It should be noted that
since each campaign had different branding objectives, and each
brand had different branding baselines (as well as numerous other
variables). This study should not be interpreted as a comparison of
the study’s participating web sites.
CNET Genuity
Control
468X60 Banner
iWon Vaniqa
MSN uBid
Snowball Coke
Skyscraper
Large Rectangle
Control Control Control
-
- 7 - IAB/Dynamic Logic Ad Unit Effectiveness Study
Methodology The Dynamic Logic AdIndex™ methodology is based on a
classic test versus control in-market design. Dynamic Logic's
patent-pending approach enables the testing of actual ads in a live
environment, as opposed to a simulated environment or "forced
exposure." In order to measure the effectiveness of online
advertising on how people think or feel, the AdIndex methodology
combines ad exposure data with online survey data. By using
proprietary code, Dynamic Logic can randomly recruit respondents
from web sites and, based on actual exposure history, assign them
to the appropriate control/test cell for comparison. The data is
gathered through an online survey that is customized to the brand
advertised. The metrics are gathered in a "blind" manner - that is
respondents are asked about their perception of the advertiser’s
brand along with several other brands. This reduces any potential
bias in the results. The metrics included in the research are:
o Brand Awareness o Brand Favorability o Message Association o
Purchase Intent o Ad Recognition / Recall
Respondents are offered the incentive of a 1 in 100 chance of
winning $100. Statistical significance is determined through a
one-tailed hypothesis test for the difference between two
population proportions. The confidence level assigned for
statistical significance for this study is 90%. NOTE: In order to
isolate the impact of ad unit size from other variables such as
exposure frequency (seeing one ad unit more than once) and unit
blending (seeing different ad units in combination), it was
necessary to isolate the exposure to only one ad unit format per
respondent in the case studies. It should be noted that Dynamic
Logic has found that multiple exposures to campaign creative does
increase branding effectiveness. In future studies, Dynamic Logic
will look at the relative value of frequency and different ad unit
formats.
-
- 8 - IAB/Dynamic Logic Ad Unit Effectiveness Study
Executive Summary (1) The IAB / Dynamic Logic Ad Unit
Effectiveness Study found:
• The study confirms the branding effectiveness of banners. The
468X60
banner produced significant increases in certain brand metrics
for all four campaigns.
• The new unit sizes (Skyscraper and Large Rectangle) were more
3-6 times
as effective than standard banners in increasing Brand Awareness
and Message Association.
• The campaigns were most effective in increasing Brand
Awareness and
Message Association (ability to link a message to a brand).
Purchase Intent and Brand Favorability were not generally increased
by one ad exposure although there were significant increases in
Purchase Intent in two campaigns.
• While not the focus of the study, the study found that higher
exposure levels
(more than one) did significantly increase Purchase Intent and
Brand Favorability in the aggregated data.
• All three unit sizes were effective in increasing Brand
Awareness in at
least one campaign, with statistical significance. The
Skyscraper unit was most effective, followed by the Large Rectangle
and then the standard banner.
• The larger formats were effective in significantly increasing
Message
Association in multiple campaigns. The Large Rectangle was most
effective, followed by the Skyscraper.
Individual case studies are reported on pages 10-20.
-
- 9 - IAB/Dynamic Logic Ad Unit Effectiveness Study
Executive Summary (2): Overall Results Brand Awareness and
Message Association Brand Awareness Across all four advertising
campaigns studied, the Skyscraper format was most effective in
increasing Brand Awareness, with a statistically significant
increase of seven percentage points. On average, Large Rectangles
increased Brand Awareness by five percentage points. Overall,
Skyscrapers were more than three times more effective than standard
banners at increasing Brand Awareness. Message Association Overall,
Large Rectangle units were most effective at increasing Message
Association. Large Rectangles increased Message Association by 10
percentage points, on average. Across all campaigns, Skyscrapers
increased Message Association by eight percentage points. Overall,
there was no significant increase due to exposure to the standard
banner.
55%57%
55%
60%
55%
62%
40%
50%
60%
Banner Large Rectangle Skyscraper
Fig. 1 - All Results: Brand Awareness by Format
Control Exposed
Source: Dynam ic Logic, June 2001 * Statistically Significant at
90%
+2*+5*
+7*
14% 15% 14%
24%
14%
22%
0%
10%
20%
30%
Banner Large Rectangle Skyscraper
Fig. 2 - All Results: Message Association by Format
Control Exposed
+1
+10* +8*
Source: Dynam ic Logic, June 2001 * Statistically Significant at
90%
-
- 10 - IAB/Dynamic Logic Ad Unit Effectiveness Study
Executive Summary (3): Overall Results Purchase Intent and Brand
Favorability Dynamic Logic has conducted numerous studies that have
consistently shown that Purchase Intent and Brand Favorability are
two of the most difficult metrics to increase. Since the present
study controlled for frequency by limiting each respondent to only
one exposure, the ability of the campaigns to create significant
increases in Brand Favorability and Purchase Intent was diminished.
As Dynamic Logic's AdIndex methodology captures actual exposure
data (as opposed to forced exposure data), there was an additional
data set comprised of respondents who had multiple exposure to the
ad units (either to one ad unit or different ad units). This
enables additional data points for examination. The data indicates
a direct relationship between frequency of exposure and levels of
both Purchase Intent and Brand Favorability. The data show that the
tested ad units, when analyzed in the aggregate by discrete
frequency of exposure, increase persuasion metrics but require more
exposure than Brand Awareness or Message Association to increase
significantly. The chart at right illustrates the relationship
between Purchase Intent and exposure frequency. Among respondents
exposed three and four times, the increase is significant. With
three exposures, the Purchase Intent increases six percentage
points. With four exposures, the Purchase Intent increases 10
percentage points.
44%43%
46%
50%
54%
30%
40%
50%
60%
Control Exp 1 Exp 2 Exp 3 Exp 4
Fig. 3 - Purchase Intent by Exposure Frequency
Source: Dynam ic Logic, June 2001 * Statistically Significant at
90%
-1+2
+6*
+10*
-
- 11 - IAB/Dynamic Logic Ad Unit Effectiveness Study
Overall Results: Purchase Intent and Brand Favorability (2) As
with Purchase Intent, the results for Brand Favorability also show
a direct relationship between the frequency of ad exposure and
level of Brand Favorability. Among respondents exposed more than
twice, the impact in Brand Favorability was statistically
significant. The chart at right illustrates that Brand Favorability
increases with the number of times a respondent was exposed to the
creative. Among those exposed 3 or more times, Brand Favorability
increases five percentage points. With 4 exposures, Brand
Favorability increases eight percentage points. NOTE: Due to the
need to compare ad units with an exposure level of one for the case
study results, significant quantities of multiple-exposure
respondents were not recruited for each ad unit. As a result, many
of the respondents exposed multiple times actually saw different
formats of the same campaign. There was insufficient data for those
exposed to a single ad unit multiple times. There was also
insufficient data for those exposed to any ad unit five or more
times. Consequently, the Purchase Intent and Brand Favorability
results by frequency can be looked at in aggregate and across
multiple ad units, but cannot be broken out by specific ad
units.
57% 57%
59%
62%
65%
50%
55%
60%
65%
Control Exp 1 Exp 2 Exp 3 Exp 4
Fig. 4 - Brand Favorability by Exposure Frequency
Source: Dynam ic Logic, June 2001 * Statistically Significant at
90%
+0
+2
+5*
+8*
-
- 12 - IAB/Dynamic Logic Ad Unit Effectiveness Study
Case Study 1 - Genuity Publisher: CNET Networks, Inc.
Advertiser: Genuity Ad Units: Skyscraper (160X600), Large Rectangle
(360X300), Banner (468X60)
Overview CNET Networks, Inc. is the premier source for
technology news on the Web. The client was Genuity, a
business-to-business network services provider that was formerly
known as GTE Internetworking. Message
- “Online Since ‘69”
Summary • All three ad format sizes increased one or more
branding metrics. • The Large Rectangle was successful in raising
Brand Awareness and Message
Association. Exposure to the Skyscraper format significantly
increased Message Association.
• Exposure to two standard banners led to a significant increase
of five percentage
points in Purchase Intent. NOTE: CNET Networks, Inc. does not
serve banners individually, but instead two are shown to visitors
at a time, one at the top of the page and another at the bottom.
While only respondents with one exposure were allowed in the
Skyscraper and Large Rectangle exposed cells, the frequency maximum
was two for the banner format.
-
- 13 - IAB/Dynamic Logic Ad Unit Effectiveness Study
Case Study 1 – Genuity In the CNET Networks, Inc./Genuity
campaign, both the Skyscraper and Large Rectangle formats showed
significant increases in Brand Awareness. Taking into account
different control baselines, the Large Rectangle unit had almost
twice the increase in Brand Awareness over the Skyscraper format.
The increase in Brand Awareness due to exposure to two standard
banners was not significant. There were statistically significant
increases in Message Association as a result of exposure to both
the Skyscraper and Large Rectangle format. The increase in Message
Association for the Skyscraper unit was 13 percentage points; for
the Large Rectangle, 10 percentage points; there was no significant
increase in the banner category. NOTE: Since the three format sizes
ran on areas of CNET Networks, Inc. that had different audiences,
separate control (non-exposed) groups were recruited into the study
in the event that the baseline attitudes towards the Genuity brand
varied in these areas.
13% 14% 14%
24%
16%
29%
10%
20%
30%
Banner Large Rectange Skyscraper
Fig. 6 - Genuity: Message Association by Format
Control Exposed
+1
+13*
+10*
Source: Dynam ic Logic, June 2001 * Statistically Significant at
90%
47%
51%
45%
55% 55%
61%
40%
50%
60%
Banner Large Rectange Skyscraper
Fig. 5 - Genuity: Brand Awareness by Format
Control Exposed
+4
+6*
+10*
Source: Dynam ic Logic, June 2001 * Statistically Significant at
90%
-
- 14 - IAB/Dynamic Logic Ad Unit Effectiveness Study
Case Study 1 – Genuity Data OVERALL
Brand Awareness Message Assoc. Brand Favor. Purchase Intent
+ 4 + 1 + 2 (IMP) + 5* (IMP)
47%
51% +4 8% 13% 14% +1 8% 51% 53% +2 4% 36% 41% +5 12%
+ 7* (TT) B. + 4* (TT) 9% 15% +6 76% 12% 16% +4 36%
Banner (468X60)
Cntrl n390
Exps n340
Diff Lift Cntrl n390
Exps n340
Diff Lift
Cntrl n390
Exps n340
Diff Lift Cntrl n390
Exps n340
Diff Lift
Brand Awareness Message Assoc. Brand Favor. Purchase Intent
+ 10* + 10* + 3 (IMP) + 2 (IMP)
45%
55% +10 24% 14% 24% +10 70% 51% 54% +3 5% 37% 39% +2 7%
+ 4* (TT) + 3* (TT) 10% 15% +5 47% 11% 14% +3 26%
Rectangle (360X300)
Cntrl n817
Exps n657
Diff Lift Cntrl n817
Exps n657
Diff Lift
Cntrl n817
Exps n657
Diff Lift Cntrl n817
Exps n657
Diff Lift
Brand Awareness Message Assoc. Brand Favor. Purchase Intent
+ 6* + 13* + 2 (IMP) + 1 (IMP)
55%
61% +6 10% 16% 29% +13 79% 52% 54% +2 4% 40% 41% +1 2%
+ 4 (TT) + 1 (TT) 12% 16% +4 31% 15% 17% +2 11%
Skyscraper (160X600)
Cntrl n537
Exps n196
Diff Lift Cntrl n537
Exps n196
Diff Lift
Cntrl n537
Exps n196
Diff Lift Cntrl n537
Exps n196
Diff Lift
IMP: Indexed Mean Percentage TT: Top Two Box
* Statistically significant at 90% Please note that computed
fields (Difference and Lift) are based on actual numbers, and that
the rounded numbers are shown in this report.
-
- 15 - IAB/Dynamic Logic Ad Unit Effectiveness Study
Case Study 2 - Vaniqa Publisher: iWon Advertiser: Vaniqa –
Bristol-Myers Squibb Ad Units: Skyscraper (120X600), Large
Rectangle (336X280), Banner (468X60)
Overview iWon (www.iwon.com) is a leading web portal. The brand
advertised was Vaniqa, a pharmaceutical product that inhibits the
growth of facial hair on women. Message
- “Now Up Close is Up to You”
Summary
• Exposure to the Skyscraper, Large Rectangle and banner all
created significant lifts in Brand Awareness and Message
Association.
• The Large Rectangle was the best performing unit, followed by
the Skyscraper
and standard banner. Exposure to the Large Rectangle unit
created the largest significant lifts in both Brand Awareness (+10
pts.) and Message Association (+14 pts.).
• The increase in Brand Awareness due to exposure to the Large
Rectangle was
double that of the banner.
• In terms of Message Association, the Large Rectangle performed
best, increasing this metric by 14 percentage points; the standard
banner increased it by eight percentage points; the Skyscraper, six
percentage points.
http://www.iwon.com/
-
- 16 - IAB/Dynamic Logic Ad Unit Effectiveness Study
Case Study 2 - Vaniqa For the Vaniqa campaign running on iWon,
exposure to all three format sizes significantly increased
awareness of the brand. The Large Rectangle increased Brand
Awareness by 10 percentage points, double the increase due to
exposure to the standard banner. Exposure to the Skyscraper lifted
Brand Awareness by seven percentage points. Exposure to all three
format sizes significantly increased Message Association. The Large
Rectangle performed best, increasing Message Association by 14
percentage points; the standard banner increased it by eight
percentage points; the Skyscraper, six percentage points.
14%
22%
14%
28%
14%
20%
10%
20%
30%
Banner Large Rectangle Skyscraper
Fig. 8 - Vaniqa: Message Association by Format
Control Exposed
Source: Dynam ic Logic, June 2001 * Statistically Significant at
90%
+8*+6*
+14*
23%
28%
23%
33%
23%
30%
10%
20%
30%
40%
Banner Large Rectangle Skyscraper
Fig. 7 - Vaniqa: Brand Awareness by Format
Control Exposed
+6* +7*
+10*
Source: Dynam ic Logic, June 2001 * Statistically Significant at
90%
-
- 17 - IAB/Dynamic Logic Ad Unit Effectiveness Study
Case Study 2 – Vaniqa Data OVERALL
Brand Awareness Message Assoc. Brand Favor. Purchase Intent
+ 6* + 8* - 2 (IMP) - 4 (IMP)
23%
28% +5 26% 14% 22% +8 57% 52% 50% -2 - 34% 30% -4 -
- 2 (TT) + 1 (TT) 9% 7% -2 - 8% 9% +1 11%
Banner (468X60)
Cntrl n471
Exps n314
Diff Lift Cntrl n471
Exps n314
Diff Lift
Cntrl n471
Exps n314
Diff Lift Cntrl n471
Exps n314
Diff Lift
Brand Awareness Message Assoc. Brand Favor. Purchase Intent
+ 10* + 14* - 1 (IMP) - 5 (IMP)
23%
33% +10 46% 14% 28% +14 98% 52% 51% -1 - 34% 29% -5 -
- 1 (TT) + 0.4 (TT) 9% 8% -1 - 8% 9% +1 4%
Rectangle (336X280)
Cntrl n471
Exps n718
Diff Lift Cntrl n471
Exps n718
Diff Lift
Cntrl n471
Exps n718
Diff Lift Cntrl n471
Exps n718
Diff Lift
Brand Awareness Message Assoc. Brand Favor. Purchase Intent
+ 7* + 6* 0 (IMP) + 2 (IMP)
23%
30% +7 32% 14% 20% +6 43% 52% 52% 0 - 34% 36% +2 6%
+ 2 (TT) + 4* (TT) 9% 11% +2 18% 8% 12% +4 48%
Skyscraper (120X600)
Cntrl n471
Exps n408
Diff Lift Cntrl n471
Exps n408
Diff Lift
Cntrl n471
Exps n408
Diff Lift Cntrl n471
Exps n408
Diff Lift
* Statistically significant at 90% Please note that computed
fields (Difference and Lift) are based on actual numbers, and that
the rounded numbers are shown in this report.
IMP: Indexed Mean Percentage TT: Top Two Box
-
- 18 - IAB/Dynamic Logic Ad Unit Effectiveness Study
Case Study 3 - uBid Publisher: MSN Advertiser: uBid Ad Units:
Skyscraper (160X600), Large Rectangle (240X400), Banner
(468X60)
Overview MSN (www.msn.com) is a leading web portal. The campaign
advertiser, uBid (www.ubid.com), is an eCommerce web site.
Messages
- “Protect Yourself. Brand Name Warranties.” - “Discover Winning
at Paying Less”
Summary
• The larger ad formats outperformed the standard banner across
all of the branding metrics. Exposure to the Skyscraper unit
created the largest, significant increases in both Brand Awareness
(+7 pts.) and Message Association (+6 pts.).
• The Skyscraper unit showed significant lifts in both Brand
Awareness and
Message Association. Exposure to the Large Rectangle unit
increased Message Association only.
• There were no significant lifts among respondents exposed to
the standard
banner format.
http://www.msn.com/http://www.ubid.com/
-
- 19 - IAB/Dynamic Logic Ad Unit Effectiveness Study
Case Study 3 - Ubid
For the uBid campaign on MSN, only the Skyscraper significantly
increased Brand Awareness. The increase was seven percentage
points. There were no significant increases in Brand Awareness as a
result of exposure to the Large Rectangle or banner. Both the
Skyscraper and Large Rectangle significantly increased Message
Association. Exposure to the Skyscraper increased Message
Association by six percentage points, double the increase from the
Large Rectangle. There was no significant lift in Message
Association as a result of exposure to the banner.
10%
12%
10%
13%
10%
16%
0%
10%
20%
Banner Large Rectangle Skyscraper
Fig. 10 - uBid: Message Association by Format
Control Exposed
Source: Dynam ic Logic, June 2001 * Statistically Significant at
90%
+2 +3*
+6*
51% 51% 51% 51% 51%
58%
40%
50%
60%
Banner Large Rectangle Skyscraper
Fig. 9 - uBid: Brand Awareness by Format
Control Exposed
+0.6 +0
+7*
-
- 20 - IAB/Dynamic Logic Ad Unit Effectiveness Study
Case Study 3 – Ubid Data OVERALL
Brand Awareness Message Assoc. Brand Favor. Purchase Intent
+ 0.6 + 2 + 2 (IMP) 0 (IMP)
51%
51% 0 0% 10% 12% +2 20% 52% 53% +1 2% 36% 36% 0 0%
+ 0.5 (TT) 0 (TT) 13% 13% +0.5 5% 16% 16% 0 0%
Banner (468X60)
Cntrl n448
Exps n1381
Diff Lift Cntrl n448
Exps n1381
Diff Lift
Cntrl n448
Exps n1381
Diff Lift Cntrl n448
Exps n1381
Diff Lift
Brand Awareness Message Assoc. Brand Favor. Purchase Intent
0 + 3* + 1 (IMP) + 2 (IMP)
51%
51% 0 0% 10% 13% +3 21% 52% 53% +1 2% 36% 38% +2 6%
+ 2 (TT) + 4* (TT) 13% 15% +2 19% 16% 20% +4 25%
Rectangle (240X400)
Cntrl n448
Exps n623
Diff Lift Cntrl n448
Exps n623
Diff Lift
Cntrl n448
Exps n623
Diff Lift Cntrl n448
Exps n623
Diff Lift
Brand Awareness Message Assoc. Brand Favor. Purchase Intent
+ 7* + 6* 0 (IMP) + 2 (IMP)
51%
58% +7 14% 10% 16% +6 55% 52% 52% 0 0% 36% 38% +2 6%
+ 1 (TT) + 3 (TT) 13% 14% +1 4% 16% 19% +3 18%
Skyscraper (160X600)
Cntrl n448
Exps n195
Diff Lift Cntrl n448
Exps n195
Diff Lift
Cntrl n448
Exps n195
Diff Lift Cntrl n448
Exps n195
Diff Lift
* Statistically significant at 90% Please note that computed
fields (Difference and Lift) are based on actual numbers, and that
the rounded numbers are shown in this report.
IMP: Indexed Mean Percentage TT: Top Two Box
-
- 21 - IAB/Dynamic Logic Ad Unit Effectiveness Study
Glossary Brand Awareness - Measures the level of familiarity
respondents have with the brand (aided). 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 purchase the brand in the future.
Message Association - Measures the extent to which respondents can
match the copy or messages in the creative to the brand.
Statistical Significance - Measures the probability that the
difference between the population proportions in the sample is
representative of the entire population of web visitors targeted in
the campaign. For this study, the minimum significance level to
make projectable conclusions was 90%. Lift - The relative
difference or percent change between the exposed and control
(non-exposed) cells. It is a measure of progress relative to the
baseline. The formula is (exposed % - control %) / control%.
Difference - The absolute difference between the exposed and
control (non-exposed) cells. It is calculated by subtracting the
control from the exposed. In the report, this metric is referred to
as the increase or decrease in the branding metrics. Indexed Mean
Percentage – Method of computing Purchase Intent and Brand
Favorability that gives weights to the 5 possible answers, with
greater weights given to positive points on the scale. e.g. Very
Likely – 100% Somewhat likely – 75% May or may not be likely – 50%
Not very likely – 25% Not at all likely – 0%
-
- 22 - IAB/Dynamic Logic Ad Unit Effectiveness Study
Top Two Box – Method of computing Purchase Intent and Brand
Favorability that includes those who respond via the top two
answers (very and somewhat likely, or very and somewhat
favorable).
-
- 23 - IAB/Dynamic Logic Ad Unit Effectiveness Study
Appendix A: ADINDEX™ Methodology Dynamic Logic measures consumer
perceptions and attitudes toward an advertiser's brand by capturing
consumer opinions through an online survey. AdIndex gathers this
data while the online campaign is running. To isolate the impact of
advertising exposure on consumer attitudes (i.e. Purchase Intent,
Message Association, etc.), two groups of online consumers are
sampled at the same time and from the same Web sites on which the
campaign is running. As the only difference between the groups is
the presence of the advertising, any attitudinal differences
between the two groups can be attributed to the exposure to the
specific online ad campaign.
Chart: AdIndex Methodology Sampling for AdIndex is bifurcated,
in that respondents are grouped into control and exposed cells for
analysis. The control/exposed methodology is dependent on (1)
reliable tracking technology to recognize a consumer’s ad exposure
level, (2) an online questionnaire development tool, and (3) a
sound recruitment process to ensure a statistically valid
comparison between exposed and control groups.
Tracking Technology
AdIndex uses a proprietary technology named AdScout™ to track
exposure to online advertising. When survey respondents enter an
AdIndex survey, the system recognizes whether or not they have seen
any ads in a given campaign. In addition, the system identifies the
version of creative, frequency of exposure and time elapsed since
last exposure; the later two defined as frequency and recency.
2 categories of ads
Campaign ads
Control ads
Client Campaign
Every nth exposure sampled
Every nth exposure sampled
AdIndex™ Survey Statistically
similar samples recruited from the same sites over the same
period of time
Recognizes exposure levels including frequency
-
- 24 - IAB/Dynamic Logic Ad Unit Effectiveness Study
The tracking technology does not require a specific third party
ad server, nor does it rely on the expensive and time-consuming
practice of querying third party ad serving databases.
Survey Development The standard AdIndex survey system measures
the following branding metrics:
- Brand Awareness - Purchase Intent - Brand Favorability -
Message Association - Advertising Recall
Questions are created through an automated system that enables
an online survey to be developed quickly. While the survey is
standardized, clients can customize questions that are intended to
drive segmentation of the results. These customized questions
include two category usage questions and one qualitative open-ended
question. All questions are asked in a "blind" manner where the
advertiser is presented in a list along with several competitors.
This prevents the respondent from discerning who the client is,
which may bias the survey results. As mentioned previously, the
survey asks all branding questions in a standardized manner to
enable cross campaign comparison and comparison to a normative
database.
Recruitment Process Both samples of respondents are recruited
via an online survey invitation; typically a pop-up window or
banner. For those in the exposed group, the pop-up is tied to the
ads in the campaign and launched for every nth visitor leaving the
page where the ad is served. To gather a control group, dummy
banners, usually Dynamic Logic banners, running in the same
targeting footprint as the actual campaign, launch pop-up
invitations in a similar manner. Since both the control and exposed
groups are recruited in the same fashion, at the same point in
time, and from the same sites, it statistically can be and is
assumed, that they have the same likelihood of exposure to
alternate channel marketing efforts, such as television, print, and
radio. Known as the assumption of initial equivalence, the AdIndex
methodology controls for all other influences outside of the online
advertising campaign and therefore allows any differences in
attitudes between the two groups to be attributed specifically to
the online advertising campaign in question.
-
- 25 - IAB/Dynamic Logic Ad Unit Effectiveness Study
Data Calculation
All data is calculated as proportions (or percentages) of the
sample that feel a certain way or indicate a specific quality.
- Message Association: = (count of those who match message with
correct brand)/ total in cell
- Brand Awareness (measured on a weighted scale):
= (count of who are "definitely aware" + 0.5 * count of those
who "may be aware") /total in cell
- Indexed Mean Percentage: Purchase Intent (measured on a
weighted scale):
= (count of those "very likely" + 0.75 * count of those
"somewhat likely" + 0.5* count of those "may or may not be likely"
+ 0.25 * count of those "not very likely")/ total in cell
- Indexed Mean Percentage: Brand Favorability (measured on a
weighted scale): = (count of those "very favorable" + 0.75 * count
of those "somewhat favorable" + 0.5 * count of those "neutral" +
0.25 * count of those "not very favorable")/ total in cell
- Top Two Box: Purchase Intent: = (count of those "very likely"
+ count of those "somewhat likely") / total in cell
- Top Two Box: Brand Favorability: = (count of those "very
favorable" + count of those "somewhat favorable") / total in
cell
Interactive Advertising Bureau / Dynamic LogicAd Unit
Effectiveness StudyResearch conducted by Dynamic Logic on behalf of
The IAB
Table of ContentsIntroductionBackgroundStudy
DesignMethodologyExecutive Summary (1)Executive Summary (2):
Overall ResultsExecutive Summary (3): Overall ResultsPurchase
Intent and Brand FavorabilityOverall Results:Purchase Intent and
Brand Favorability (2)Case Study 1 - GenuityOverviewCNET Networks,
Inc. is the premier source for technology news on the Web. The
client was Genuity, a business-to-business network services
provider that was formerly known as GTE
Internetworking.MessageSummary
Case Study 1 – GenuityCase Study 1 – Genuity Data OVERALL+ 2
(IMP)OverviewiWon (www.iwon.com) is a leading web portal. The brand
advertised was Vaniqa, a pharmaceutical product that inhibits the
growth of facial hair on women.
Case Study 2 - VaniqaCase Study 2 – Vaniqa Data OVERALL- 2
(IMP)OverviewMessagesSummary
Case Study 3 - UbidCase Study 3 – Ubid Data
OVERALLGlossaryAppendix A: ADINDEX™ MethodologyChart: AdIndex
Methodology
Tracking TechnologySurvey DevelopmentRecruitment Process