Global Benchmark Report
Global Benchmark Report
2009
Table of Contents
Benchmarks Overview 9
Benchmarks Tables
North America 12
Europe 13
APAC 15
Latin America 16
United States 17
Canada 18
United Kingdom 19
Spain 20
France 21
Germany 22
Australia 23
Japan 24
Taiwan 25
Verticals Definitions 26
Metrics Definitions 27
Page 2
Executive Summary
Welcome to this new and improved edition of the Eyeblaster benchmarks. In this new format, we will provide more than
just the numbers. In addition to a comprehensive set of benchmarks tables, summarizing the industry’s performance in
five markets and nine countries, this edition of the Global Benchmark Report offers an in-depth dive into the numbers.
The Benchmarks Insights and Analysis section is divided into two parts: The first provides an analysis of the benchmark’s
numbers in the context of a prevalent industry question. The second provides a general overview of the benchmarks and
a review of changes from the last edition.
This edition of the Global Benchmark Report is dedicated to unit sizes. Results were surprising—while the CTR
performance of Standard Banners tends to improve as unit size increases, in Rich Media, size is a poor predictor of
performance. In Rich Media, size is only one component of banner visibility on the site.
The analysis indicates that better predictors of performance in Rich Media are creative features such as video, ad format,
flash features and expansions. Therefore, to improve performance, advertisers should focus on enhancing ads with video
and other features, rather than increasing unit size.
Size is now in the heat of an industry debate, as publishers are offering larger and larger sizes in the hope of increasing the
marketing effectiveness of ads on their site. Advertisers, on the other hand, are faced with options to allocate their budget
to either bigger size or other rich media features. In this research, we look at which sizes are most frequently used, and
then test whether size is a good predictor for performance in Standard and Rich Media.
This study is one of the most comprehensive of its kind, comprising of more than a quarter of a million ads worldwide.
Ads participating in this analysis are from all sizes, formats and markets, and were served between Q3 ’08 and Q2 ’09.
Background
A study by the Interactive Advertising Bureau (IAB) found that larger ad units are 25% more effective in lifting key brand
metrics such as brand awareness and message association, even after one exposure. The research also shows that
additional exposures significantly increased persuasion metrics such as purchase intent1.
1
Interactive Advertising Bureau (IAB).
Available at: http://www.iab.net/about_the_iab/recent_press_releases/press_release_archive/press_release/4426
Source: IAB
Based on the assumption that ad performance increases with size, publishers are pushing for larger ad sizes. The IAB has
added a new giant ad—720x300—the largest so far. The Online Publishers Association (OPA) has also introduced new
giant sizes: The 336x700 (Fixed Panel), 468x648 (The XXL Box) and 970x418 (The Pushdown).
While 300x250 and 728x90 shares of total ads remain similar across markets, smaller sizes shares do change. 160x600
(Wide Skyscraper) is popular in North America and less common in the rest of the world, while 234x60 (Half Banner) is the
other way around.
100% Ad Size
5% 11% 180x150
11% 6% 14%
13% 300x600
80% 9% 336x280
24%
120x600
Share of Impressions
12% 11%
33%
32% 468x60
60% 6%
22% 234x60
26% 15%
160x600
9%
40% Other
728x90
46% 300x250
20% 39% 38% 42% 41%
0%
APAC AU and Europe Latin North
NZ America America
Worldwide, the popularity of unit sizes does not differ significantly between ad formats. Most unit sizes maintain equal
proportion of Standard Banners and Rich Media. Notable exceptions are 234x60 (Half Banner) and 180x150 (Rectangle),
which lean toward Rich Media and 160x600 (Wide Skyscraper) and 468x60 (Full Banner), which lean toward Standard Banners.
160x600
120x600
180x150
468x60
300x600
336x280
This conclusion is also supported by a research conducted by Dynamic Logic. Analyzing the results from 4,800 campaigns,
the marketing research firm found that the best-performing ad size, in terms of metrics such as brand awareness, recall and
purchase intent, was 180x150 (rectangle)2. Dynamic Logic’s research indicates that ad effectiveness depends less on size
than it does on creative, shape and placement.
The chart above presents the CTR performance of Standard Banners and Rich Media ads by size. The ads are sorted by
total pixel size, i.e. the largest, 300x600 is on the left and the smallest, 234x60 is on the right. While Rich Media ads’ CTR
performance has no visible correlation with total pixel size, Standard Banner ads’ performance does.
For Standard Banners, the link between size and performance is more than visible. A statistical analysis conducted by
Eyeblaster Research indicates that on average, an increase of 50,000 pixels yields an increase of 0.044% in CTR. This is
quite an impressive figure, considering that the overall CTR of Standard Banners is 0.1%. Therefore, in Standard Banners,
where unit size is one of the major factors to attract users’ attention, increasing unit size is likely to boost CTR performance.
While CTR is the only metric to measure Standard Banners’ performance, Rich Media has more robust metrics. Eyeblaster
has developed two proprietary metrics to measure users’ engagement with ads—Dwell Rate and Dwell Time. Dwell Rate
measures how many impressions were intentionally engaged with by users out of total impressions. Dwell Time measures
2
Abbey Klaassen. Why Large Online Ad Formats Aren't Industry's Silver Bullet.
AdvertisingAge, August 20, 2009. Available at: http://adage.com/digital/article?article_id=138554.
Even when taking Dwell Rate, Dwell Time and CTR into account, unit size remains a poor predictor of performance.
Ads which present superior average performance on all three metrics do not seem to have larger unit size as a common attribute.
468x60
Dw ell Time
180x150
45
728x90
Average
160x600
40
120x600 300x250
300x600
35 336x280
30
4% 6% 8% 10% 12% 14%
Dwell Rate
The chart above represents a comprehensive performance analysis of various Rich Media unit sizes. Dwell Rate is
represented on the X-axis and Dwell Time is represented in the Y-axis; the size of the circles represents the unit’s size.
Two average lines divide the chart into four quarters; the upper right quarter offers superior performance on Dwell Time
and Dwell Rate. This analysis indicates that 234x60 (Half Banner), and 180x150 (Rectangle) are the highest performers.
468x60 (Full Banner) has superior engagement in terms of Dwell Time, but is disappointing on Dwell Rate, while 336x280
(Large Rectangle) and 300x600 (Half Page Ad) have it the other way around. 300x250 (Medium Rectangle) has nearly
50% of all impressions and due to its impressions volume, comprises a large portion of the average. Overall, the top
performers are not the largest ads.
The chart above analyzes the performance of Polite Banners by unit size, with and without video. Unit sizes were sorted by
total pixel size from high to low. To simplify the analysis, performance is measured in Total Dwell—Dwell Rate multiplied by
Dwell Time. This metric quantifies the average dwell time spent on the ad by all users who were exposed to it, and provides
a simple measure of Rich Media performance.
The results are that in all but three sizes, video increases performance significantly—overall by 71%. This finding cannot
be underestimated—adding video and Rich Media features to ads augments their performance in nearly all sizes.
End Note
So, does size matter? It definitely does for Standard Banners but not so much for Rich Media. Adding video and other Rich
Media features to ads makes more of a difference than increasing the size only.
As we mentioned at the beginning of this research, there is a tradeoff between unit sizes and other features, i.e. assuming
constant budget, larger unit sizes means less video and rich content. The main points of this analysis is that increasing unit
size should be considered for Standard Banners, while for Rich Media, advertisers should focus on adding video, richer
features and elaborate creative.
Overview
More than 100 billion impressions, over a quarter of a million ads and tens of thousands of campaigns were analyzed to
create this edition of the benchmarks. These benchmarks outline industry performance averages and allow advertisers and
agencies to compare their performance to that of their peers. With five markets and nine countries, this edition represents
the most comprehensive international coverage in the industry.
To evaluate the overall performance of Rich Media, Eyeblaster has developed two proprietary metrics—Dwell Rate and
Dwell Time. Dwell Rate measures the ratio of impressions in which viewers engaged with an ad out of total impressions,
while Dwell Time measures the average duration of these engagements. Overall Dwell Rate is 7%, which means that one
of 14 impressions was dwelled upon by a user for more than one second. These users spent an average of 43 seconds of
Dwell Time with these impressions—nearly one and a half times as much as the average TV spot.
Impressions with Any Interaction Rate measures how many impressions had interactions out of total served impressions;
worldwide, these average to 3%. When measuring total interactions (Including multiple Interactions) out of served
impression—Interaction Rate—this figure increases to 6%.
AU and NZ
Europe
Latin America
North America
Rich Media metrics provide a more comprehensive measurement of advertising effectiveness compared with CTR.
However, CTR is the only metric to measure the performance of Standard Banners.
Worldwide, Standard Banners achieved Click Through Rate (CTR) of 0.1%, compared to slightly more than 0.3% for
Rich Media—more than a three times increase. CTR for Rich Media is higher than CTR for Standard Banners across all
markets.
Europe
Latin America
North
America
Markets differentiate, naturally, in users’ propensity to interact with ads. While Latin Americans are enthusiastically
engaging and clicking on ads, other users in North America and APAC tend to be more selective.
One explanation for the differences is market maturity. In less mature markets, where users are novice in online
advertising, they receive ads with more open arms. When the market matures, the users’ fatigue together with increased
competition for their attention reduces engagement significantly. The users’ propensity to engage with ads should be taken
into account when comparing benchmarks between geographies and verticals.
What’s New?
Whenever a new edition of the benchmarks is published, advertisers, agencies and Eyeblaster’s internal team suggest
new ideas for improvement. This edition offers many innovations as a part of our constant effort to create the most
comprehensive set of benchmarks in the industry.
• More markets and countries covered: In addition to adding benchmarks for Latin America, we added more individual
countries with their own benchmarks. The country level benchmarks can assist advertisers by comparing their
performance to that of a more relevant list of peers, while market level benchmarks provide a larger sample size.
• Performance by unit size: After constant feedback to the Research team that size does matter, Eyeblaster’s
benchmarks will now include the three most popular sizes for each ad format. Our analysis shows that the three most
popular sizes cover more than 70% of ads.
• Insights and Analysis: Eyeblaster Research team can contribute more than a summary of the numbers. In each
edition of the Benchmarks Report, we will provide actionable analysis that goes beyond the numbers. With the most
sophisticated data visualization capabilities and up to date statistical computer models, these insights will provide
rigorous answers to industry questions.
• Comprehensive metrics definitions: Thorough understanding of the metrics is key for improving campaign
performance. This new glossary provides in-depth definitions of the metrics and explains how to use them.
• Verticals definitions: When comparing a campaign to its peers, it is important to select the right set of campaigns for
comparison. To avoid confusion, we included a precise definition of each vertical.
Page 12
Page 13
Page 14
Page 15
Page 16
Page 17
Page 18
Page 19
Page 20
Page 21
Page 22
Page 23
Page 24
Page 25
Page 26
Page 27