Excelente white paper, lo que más me gusto es que muestran las tendencias del CTR por zona mundial, México & Latam entran en " North America "
Gracias Lazaro por el archivo
2. Background Summary of findings
In a time of recession, • Click-thru Rate vs. Conversion Rate:
Worldwide (WW) ’08 CTR is 0.14%
accountability of online while WW’08 Conversion Rate stands at
advertising becomes more 0.32% - more than double.
critical than ever. Whether you • Correlation between Format Type and
Conversion Rate: Rich media ad formats
are pursuing a direct response increase the tendency of users to convert
by 2.5x.
initiative or looking to achieve
• Impact of Data Capture on Conversion
measurable branding objectives, Rate: Lowering the number of phases the
the focus is on the process of user needs to go through by bringing the
conversion mechanism into the banner.
`conversion’: that moment when You are 5x more likely to convert in a
an individual exposed to an ad banner than you are on a site.
performs a desired action.1 • CTR vs. Data Capture: Comparisons
across unique data reveals users are
This issue of the Analytics nearly 8x more likely to convert in a
banner than click on it and about 34% of
Bulletin focuses on the role unique users who interact will fill in form
details within a banner.
of conversions. Given the
• Automatic Optimization and its impact
accountability and measurability on Conversion Rate: Conversion rate
of online media, what measures of campaigns using an optimization
algorithm is significantly higher (1.25x)
can we take in order to maximize than the average conversion rate of
conversions at zero or minimum rich media and practically double the
global average conversion rate for all ad
cost? Eyeblaster conversion formats.
tags give the ability to track • Search as a conversion metric: One can
measure an increase in search activity
user’s activity on the advertiser’s as a result of display advertising that is
site after viewing or clicking on sustained for a sizeable amount of time
after the display campaign has finished.
an ad, but this is only one of a
• Behavioral sequencing: Consumers
myriad of methods that can be exposed to targeted messages in
deployed to ensure a better yield sequence are more likely to move
through a conversion sequence.
of success.
Pg 193 – Marketing Sherpa’s 2008
1 Pg 193 - Marketing Sherpa's 2008
3. Table of contents
Research Methodology ............................................................................. 4
Results in Detail ........................................................................................ 4
Post-Click vs. Post-Impression ................................................................. 5
Improving the conversion process with Data Capture ................................ 6
Automatic Optimization mechanism and its impact on conversion rate ..... 10
Search within the conversion process .......................................................11
Benchmarks
Appendix 1: north America .................................................................... 15
Appendix 2: Europe ................................................................................ 16
Appendix 3: Spain .................................................................................. 17
Appendix 4: APAC .................................................................................. 18
Appendix 5: Taiwan ................................................................................. 19
3 AnALYTICS BuLLETIn ISSuE 3 | MARCh 2009
4. Research Methodology Results in detail
This study is based on data of more than 2,000 Click-thru Rate (CTR) vs. Conversion Rate
campaigns representing 29 billion impressions collected
CTR is a valuable metric to measure the amount of users
between Jan ’08 and Dec ’08 covering all sectors.
who wish to visit the advertiser’s site, but this in itself is not
These campaigns represent advertisers using Eyeblaster
a true conversion. The common perception is that CTR
conversion tags. Conversion rates were based on
indicates campaign success; the following analysis shows
total conversions with no differentiation of the type.
that low or high CTR does not necessarily correlate with
All calculations were performed on daily delivery data
the campaign’s overall objectives. Conversion tags placed
aggregated to the ad level.
on the advertiser’s web site on the other hand, track actual
user activity that the advertiser wishes to track.
Thinking about conversions… In observing the CTR metric alone, we expect all
consumers to react immediately and overlook consumers
who converted after being exposed to – and possibly
interacting with – the ad.
In order to see the full picture we should sum both types
It could be argued that of conversions.
a true conversion would
be where a consumer CTR vs. Conversion Rate
has made a monetary 0.35%
transaction directly with the 0.32%
advertiser, i.e. purchased 0.30%
something. The focus here 0.25%
is on conversions where
.5
x2
0.20%
the conversion means 0.14%
‘showing intent’ – such 0.15%
as through data capture 0.10%
– indicating a consumer’s
0.05%
desire to purchase.
0.00%
It is worth noting that a total conversion CTR Total Conversion Rate
could happen directly on an advertiser’s site
Eyeblaster Jan.2009
but possibly more so on a reseller’s site or in
some other place, such as offline or in-store.
The true effects of digital advertising against Figure 1 Source: Eyeblaster Research 2009
this overall conversion process (included
within or in isolation from other media) are
In looking at the overall lift in the conversion rate, it
not considered here due to current tracking
indicates that a higher percentage of consumers take in
methods. For the purpose of this document,
the advertising message, then require time to digest the
we shall limit conversions to those indicating
information, possibly clarify it from other sources, before
intent within the ad itself or direct on an
finally being in a position to actually convert.
advertiser’s site, appreciating this in itself is
only showing a slither of the actual overall
Conversion rate provides more accurate and precise
conversion rate and ROI.
measurement of the campaign’s success and global
data reveals that by focusing on CTR alone, we are
missing 2.5x the actual data an advertiser requires to
correctly analyze the overall consumer journey.
4 AnALYTICS BuLLETIn ISSuE 3 | MARCh 2009
5. Post-Click vs. Post-Impression Standard advertising plays an important role in enhancing
reach and frequency of the entire media mix (Rich,
Standard, Search) so while not having as strong a recall
Correlation between Format Type and
and conversion rate, it acts as a reminder for those already
Conversion Rate with awareness, and as a reach extender. We shall look at
ways to enhance the standard banner effectiveness later.
With pressure mounting on trying to ascertain the most
cost-effective online advertising, it would seem that high
Another way to look at this conversion process is to try to
volume standard display advertising would naturally be a
understand when the conversion ultimately takes place,
first choice. however, when we look at the true conversion
i.e. is the advertising message creating an immediate
rates for the 3 most popular ad format types online, we
or a time delayed conversion? Previously we have seen
clearly distinguish a higher rate for rich media advertising.
that rich media is creating stand-out to ensure brand
recall later, but is this a trend we should expect when we
Conversion Rate by Formats consider all display advertising?
Let’s compare the following two scenarios:
• See-Click-Convert (post click)
0.60%
• See-Delay-Convert (post impression)
0.50% 0.47%
0.50%
Conversion Rate
0.40%
0.16%
0.30%
0.14%
0.20%
0.14%
0.20%
0.12%
0.10%
5
.7
0.10% x1
0.00%
0.08%
STD Banner Polite Banner Expandable 0.08%
Banner
0.06%
Eyeblaster Jan.2009
Figure 2 Source: Eyeblaster Research 2009
0.04%
0.02%
Both popular rich media
formats are showing 2.5x
0.00%
Post - Click Post - Impression
uplift in the ability to generate
conversions over standard Figure 3 Source: Eyeblaster Research 2009
display inventory.
The results indeed show that there is a sizeable uplift in
the number of consumers who do not necessarily respond
The results reveal that rich media is doing two things
and convert immediately to an advertising message,
here. First, it is attracting attention of the viewer against
and this factor needs to be explored further in trying to
site content, and thereby ensuring greater stand out
understand the consumer’s mindset.
from site editorial. Second, by allowing a consumer to
tangibly touch and interact with the advertising message,
the consumer experience is enhanced through physical
1.75x conversions happen
involvement, with a resulting positive effect for the
advertiser, shown in actual conversions.
after a time delay
5 AnALYTICS BuLLETIn ISSuE 3 | MARCh 2009
6. Improving the conversion process with Data
Capture
Whether you go for standard
We all know that in the “viewer to converter” funnel there
is a drop off at every step. Therefore, we should aim to
or rich display advertising,
minimize the funnel steps. In general, data capture is the
ability to collect information submitted by the user. By
not including data capture
placing data capture functionality into the advert itself;
we are addressing a type of conversion that occurs on
within any DR campaign is at
the web site where the advert is placed, instead of the
advertiser’s web site. This feature can be implemented in
least halving your overall ROI
various ways: one of the most common is allowing users
to request additional advertiser information from within the
effectiveness.
ad. It can also be used for polling users’ opinion, collecting
email addresses, cell phone numbers, or any other details,
including full transactional details.
“Viewer to Converter” Funnel
On-Site conversion in-Banner Conversion Rich
Regular banner flow (Rich Media banner flow
or Standard)
Viewing and/or interacting
Viewing and/or interacting
with the ad
with the ad
Clicking on the advert
Arriving at the Details Submission
web page
Sign in-start
page Details
Confirmation
Sign in -
Confirmation
page
Figure 4 Source: Eyeblaster Research 2009
6 AnALYTICS BuLLETIn ISSuE 3 | MARCh 2009
7. You are 5x more likely to
Conversion Rate (on-Site vs. in-Banner)
convert in a banner than
convert on a site as a result of
1.20%
1.06%
clicking on a standard banner
1.00%
x5
0.80%
Case Study
Impact drives Conversions
0.60%
health and Beauty campaign
0.51%
0.40%
Let’s try to clarify the power of data capture by
0.20% referring to the following case study.
0.20%
In a health and Beauty campaign we monitored
the amount of email addresses submitted on
0.00%
a daily basis during the campaign’s lifetime, 22
Standard Rich Banner Data Capture
days. (Over 3 million impressions were served on
Banner
Yahoo, nine.MSN and news.com.)
Figure 5 Source: Eyeblaster Research 2009
in-Banner vs. on-Site Conversions
When comparing the data capture within a banner
60
(1.06%) with that of on-site (0.20% and 0.51%) we
Conversion or Form Submission
see that by removing a number of required steps, the 50
conversion rate is 5x that of standard display and
40
double that of rich media as a result!
30
By simplifying the process that the consumer is required
to follow, the advertiser achieves a more desirable result. 20
however, it is important to note that the level of interactivity 10
within Rich Media is nOT a bearing on the conversion
0
result here; it is purely about making things easier for the 1234 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
consumer to connect and convert in general. users want Days from campaign’s start date
to connect with a brand where they are. Form Submissions Confirmation Conversions
When we consider that standard advertising does not Eyeblaster Jan.2009
have the ability to measure anything beyond a click,
Figure 6 Source: Eyeblaster Research 2009
the need for inclusion of a data capture facility across
networks is startlingly obvious. not only is this challenging
the notions of cost-effectiveness of standard media buys
The amount of emails submitted
at driving conversions, but also what we have come to
from the banner itself (data
define as Direct Response advertising online. It would
serve all standard advertising well to move towards the
capture) was significantly higher
inclusion of simple response based functionality.
throughout the campaign’s lifetime
in comparison to emails submitted
from the advertiser’s site.
7 AnALYTICS BuLLETIn ISSuE 3 | MARCh 2009
8. Further challenging the notion of CTR as an Personalizing the “Viewer to Converter”
effective conversion metric Funnel
With CTR being the de-facto measurement of online DR Impressions and CTR are non-personal attempts at trying
campaigns, how does this need to be addressed given the to understand what consumers actually do online. In
wealth of data on consumer behaviour discovered over the breaking down the “viewer to converter” funnel for those
last 10 years about rich media advertising? who convert in banner, we need to try and understand how
many times does a consumer get exposed to an advert,
how many touch the advert, for how long do they interact
Unique Clicking Users vs. Form Submission with it and what is the resulting conversion pattern?
Therefore we see a funnel of total impressions ➔ unique
impressions ➔ unique interacting users ➔ number of
1.20%
conversions (which we will determine as form submission
1.06%
and assuming that person will not submit more than one
1.00% form and thus can be related to as a unique number.)
• See-Interact-Convert (in-Banner)
0.80%
0.60%
77
x7.
0.40%
0.20%
0.14%
0.00%
Click Through Rate In - Banner
Conversion Rate
Figure 7 Source: Eyeblaster Research 2009
• unique viewers’ rate – unique out of total impressions
using global data, we can see that the inclusion of data
capture within a banner is just shy of 8x more effective at • unique Interacting users - unique interacting users out
addressing conversions than CTR. It begs the question, of unique impressions
why don’t all DR campaigns have basic data capture
• Average Dwell Time – number of seconds a viewer
functionality built-in as standard?
actively is engaged with an advert
Consumers are nearly 8x more • Converting users – form submissions out of unique
interacting users
likely to fill in a form in a banner
34% of unique users who
then click on it
interact will fill in form details
within a banner
What we see from the data is the viewer is being exposed
to the ad at least twice (42.79%).
7.40% of these then go on to interact with the ad, and do
so for well over a minute (78.08 secs).
33.56% of those who interact go on to convert within the
banner.
8 AnALYTICS BuLLETIn ISSuE 3 | MARCh 2009
9. Communication and conversion
Case Study People love to talk. Whether it is product feedback or
recommending to a friend, there are a variety of methods
e-Commerce transactional to start to penetrate the consumer conversation to ensure
higher conversion rates. Widgets within social networks
banner are one such method. But what is more interesting is
that due to the poised typing position people have within
How far can we push conversions communication channels, such as email and Messenger,
there is often more of a willingness to fill in data capture
in-Banner?
forms within these channels.
A leading European mobile operator wanted to
attract new subscribers to their network. The Health Campaign
concept of data capture was taken to the next CTR vs. Data Capture
stage as they wanted to get much more personal
0.70%
attributes from the interested user as part of
0.60%
entering into a contract with the operator. In 0.60%
this case not just address details, but additional
0.50%
qualifiers such as passport or ID card number
as well as bank account details. The incentive 0.40%
to do so was an offer for a free mobile handset.
.06
x4
0.30%
The placement of a creative on a leading
portal seemed to offer assurance to the 0.20%
0.10%
interested consumer and combined with the
0.10%
incentive was enough to inspire confidence
of 500 consumers who went on to give full 0.00%
personal details within the advert! This level Click Through Rate In - Banner
of success was well-beyond the expectations of Submission
the advertiser, but it goes to prove the willingness
of modern consumers – and the need to not Figure 8 Source: Eyeblaster Research 2009
overlook data capture within creative executions.
Taking standard display advertising to the
next level:
A health and Beauty Campaign that utilized a standard
display advert within a communication channel found that
the CTR was under performing. however, the inclusion of
a simple data-capture form that opened in an expansion
panel upon clicking on a button within the creative saw
a quadruple rise in in-banner submissions over the CTR,
incentivized by the chance to win a pair of sunglasses.
Inclusion of a simple data-
capture form… saw a
quadruple rise in in-banner
submissions over the CTR
This is a trend that is seen in Automotive, Travel and
health who often utilize data capture as part of their
campaign objectives.
9 AnALYTICS BuLLETIn ISSuE 3 | MARCh 2009
10. Automatic Optimization mechanism and its
Case Study impact on conversion rate
not only can you minimize the number of steps in the
Data Capture as the “viewer to converter” funnel, there is an ability to monitor in
desired call to action real-time the effects of one creative against another during
the campaign.
Eyeblaster’s Automatic Optimization mechanism uses
Masterfoods hoped to compel interested
an algorithm to find the ads with the highest success
consumers to sample a new type of chocolate
rate based on the metric defined, which can often vary
without the need to click through to a mini-site. upon placement of media by section or publisher. The
They created a contest to drive interest and case study below demonstrates the results of such
engagement and a series of ads was developed a methodology deployed and the overall campaign
performance based on conversions.
around the idea of “Mix It up”. The prize was the
chance to win a 3 month’s supply of chocolates
Comparisons against similar
in exchange for personal contact details. The
executions were adapted to suit the publisher
campaigns - one with
environment; e.g. high impact on the homepage,
with the ability to play a game prior to entry
optimization and one without,
into the competition – but data-capture alone
reveals a considerable change
within communication channels. All creatives
were designed to remove steps to conversion
in the ability to control a
by immediately presenting the competition entry
form on roll-over which was the focus, instead
positive shift in ROI.
of muddying the call to action with click-thru.
It worked. In a single day 35,000 people
entered details indicating they wanted to From data collected across one Automotive advertiser
taste Mars chocolates! An incredible indication over a six month period, the conversion rate of campaigns
to a brand as to audience favorability and a true using an optimization algorithm (0.63%) is significantly
higher (1.25x) than the average conversion rate of rich
demonstration of online DR at work.
media ads (0.51%). In fact, it is practically double the
average conversion rate for WW ’08 – 0.32%.
The inclusion of an incentive, To understand this more fully, let’s delve inside a campaign
that implemented real-time optimization.
such as a giveaway or access to
exclusive content, will positively 0.70%
impact the data capture rates.
0.60%
Automotive
Advertiser (0.63%)
0.50%
x1.25
Rich Media
0.40%
(0.51%)
x1.97
0.30%
0.20%
WW ’08
(0.32%)
0.10%
0.00%
Conversion Rate
Figure 9 Source: Eyeblaster Research 2009
10 AnALYTICS BuLLETIn ISSuE 3 | MARCh 2009
11. Search within the conversion process
Case Study As we have seen, not all display advertising messages can
be tracked either by an immediate response within the
A campaign using the banner, but neither can they be determined by a post-view
arrival at an advertiser’s site.
Automatic Optimization
mechanism Influence of Online Display on Clicks
of Search Campaigns
Creative A is tested against Creative B in the
same automotive campaign. All factors such as
placement, ad format and size remain consistent, 2000 600
just the creative messaging was different – there 1800
were 5 messages per placement. Looking at the 500
1600
final conversion rate as the ultimate point to weigh
against – irrespective of click-thru or interaction
Impressions (in 1000s)
1400
400
rate – we are able to control a significant uplift in 1200
conversions.
Clicks
1000 300
800
Conversion Rate with Automatic 200
Optimization 600
400
100
200
25.00%
Weeks 0 0
1 2 3 4 5 6 7 8
20.00%
Impressions of display campaigns Clicks of search campaigns
15.00%
Figure 11 Source: Eyeblaster Research 2009
10.00%
In some scenarios, after a consumer is exposed to a
message, they wish to perform an independent evaluation,
5.00%
which ultimately may even lead to a purchase offline or at
a reseller’s site. Think about your recent purchase behavior
– did you review information on a comparison site first?
0.00%
And was that purchase ultimately made directly on the
1 3 5 7 9 11 13 15 17 19 21 23 25 27
advertisers site – if online at all? Therefore we need to look
Conv Rate of a Regular f light Conv Rate of an Optimized f light
at cross-channel activity in order to ascertain a deeper
understanding of overall conversion effectiveness beyond
Figure 10 Source: Eyeblaster Research 2009
in-banner or on-site conversions.
In this example, we find that initially search activity is fairly
The non-optimized flight was performing better dormant. Once the display activity has begun, there is a
initially, but once the technology automatically followed lift in the search activity. What is important to note
adjusted the weighting of the creative in the is that often there is a delay between the display exposure
optimized flight, you can see the uplift in the trend. and the performing of the search, as a consumer finds
their own time to do their research.
When looking at the final data, we find that the
optimized creative over time lifts the overall
Optimizing creatives lifts
conversion rate by 0.82% for the non-optimized
message to create impressive spikes of 23.42% –
the overall conversion rate
and totally smashes the average global conversion
rate of 0.32%.
11 AnALYTICS BuLLETIn ISSuE 3 | MARCh 2009
12. 4500
4500
4000
4000
Thomas McIlheran, Senior Media Manager
3500
3500
says: “The Eyeblaster Channel Connect
Impressions (in 1000s)
Reporting system is the most progressive
3000
3000
reporting tool for monitoring the joint impact of
2500
2500
Clicks
search and display advertising. CC4S allows
2000
2000
advertisers and marketers to put numbers
around exactly how well these online tactics
1500
1500
work together as opposed to how they work
1000
1000
against or in spite of one another.”
500
500
0
0
03/10/ 17/10/ 31/10/ 14/11/ 28/11/ 12/12/ 12/12/ 09/01/
“Using Channel Connect for Search and
2008 2008 2008 2008 2008 2008 2008 2009
Display, Mindshare was able to pinpoint crucial
Impressions of display campaigns Clicks of search campaigns
campaign data and draw important insights
Figure 12 Source: Eyeblaster Research 2009
about the interaction of our search and display
ads. In the end, it provided us with a more
comprehensive overview of user behavior”,
The tailing-off effect of the search activity is maintained
at a much slower decline than the display advertising, as Harry Case, Director of Media Analytics
the emotional connections a consumer has made as a and Technology, Mindshare.
result of the ad exposure continue to play on their minds
for perhaps the next couple of weeks. It could also be fair
to say that the multiple exposures to display messaging
during this time are having a positive impact in the desire
to search, and this higher frequency ensures a greatly
drawn-out time span of searches – long after the effects Targeted messaging with behavioral
can be seen within traditional display advertising metrics,
sequencing
i.e. clicks have declined, but searches are maintained.
This inherently challenges traditional notions of how The ability to adaptively sequence creative messaging can
to optimize display advertising frequency, based be done using various tracking methodologies utilizing
upon display advert CTR. cookies and Flash Shared Objects. This has distinct uses
to ensure relevance to the viewer.
There is a delay between the First, you can ascertain if a user has previously been to the
advertiser’s site, and begin to delve where on the site and
display exposure and the the profile of the person based upon established criteria.
This enables knowledge of the person exposed to the
performance of the search - creative to be ascertained prior to being shown a creative
and is known in Eyeblaster as DéjàVu.
potentially by as much as 3 Second, there is the ability to learn what creatives have
weeks! been previously shown and/or interacted with and adapt
the creative messaging accordingly. This is known as
behavioral sequencing.
From campaigns we have looked at already, we have
Finally, there is also the ability to read in host-site
found a positive shift towards brand/product name
content of where the ad is sitting and adapt the creative
searches as a result of the display activity, proving recall
messaging to the surrounding page content. Called
effect of display advertising. We have also seen uplift in
Mash-up technology, this is another way to increase
search activity as a result of display impressions. We shall
advertising relevance.
look forward to sharing such results in future bulletins.
12 AnALYTICS BuLLETIn ISSuE 3 | MARCh 2009
13. Case Study
Sony Ericsson k550i campaign
Sony Ericsson ran a campaign across CEEMEA, which adapted differently depending on how a user
had interacted – and where – on site or within a banner. The creative was also changed each week to
try and minimize ad fatigue through over exposure.
http://creativezone.eyeblaster.com/#ItemName=SE%20K550
Total
Data Impressions Clicks CTR Stage 1 Stage 2 Upload Send
Custom
Week 1 20,827,665 38,825 0.19% 1,704 249 25 1,978
Week 2 33,928,483 32,264 0.10% 12,416 85 1,058 200 13,759
Week 3 35,917,203 32,233 0.09% 20,245 423 1,600 280 22,548
Week 4 36,036,434 33,369 0.09% 17,978 111 1,827 357 20,273
Total 126,709,785 136,691 0.12% 52,343 619 4,734 862 58,558
Figure 13 Source: Eyeblaster Research 2009
Those exposed to a more
The results demonstrated that as the creative
fatigue happened with each week the campaign
targeted message… were
ran, even with changing the creative per week,
more likely to move through
CTR steadily declined due to over-saturation.
the conversion sequence
Stages 1 and 2 represent a targeted message
showing a personal message for this week’s
competition (stage 1) or if the user had already
participated, a message advertising the coming
week’s competition (stage 2) up to the final week
4. The data revealed that those exposed to a
more targeted message through prior involvement,
either on site or within the banner, were more likely
to move through the conversion sequence; in this
case uploading photographs alongside personal
data and as a result, actively recommend the
campaign to their friends with a forward-to-friend
mechanism built into the ad unit.
This is one of the most exciting findings of how
activity on-site and in-banner can work together
to engage consumers and lead them through
a conversion cycle, as well as engage brand
advocacy to their friends.
13 AnALYTICS BuLLETIn ISSuE 3 | MARCh 2009
14. Conclusion
The potential to deliver better
The advent of being able to link data across-channels is a
revolutionary breakthrough for the advertising community
ROI for clients is easily within
and heralds the start of a completely new way of being
able to effectively and accurately strategize media
grasp for agencies and their
and reduce overall marketing dollar wastage. We look
forward, in future Analytics Bulletins, to show how this is
clients
progressing.
We have shown how there are multiple ways to utilize
conversions and gain higher ROI for advertisers. As with
most software, between 5-10% of possible functionality is
currently being utilized within most campaigns and thereby
the potential to develop more cost-effective advertising is
being lost through lack of knowledge of what is possible
with modern day ad serving. As this functionality is
generally included within the ad serving costs, coupled
with a little creative strategy, the potential to deliver better
ROI for clients is easily within the grasp for most agencies
and their clients.
Ask your Eyeblaster representative about how
you can harness the potential of Eyeblaster’s Ad
Campaign Manager to deliver greater ROI for your
next client objective.
14 AnALYTICS BuLLETIn ISSuE 3 | MARCh 2009
15. Appendix 1
Benchmark for North America Q4 ’07 - Q3 ’08
Performance Metrics (Format and Verticals)
Basic Metrics Video Metrics Expandable Metrics
Interaction Impressions Avg.
Avg. Video 50% Fully Total
Average Started with any Panel Expansion
IR CTR Duration Played Played Expansion
Duration Rate Expansion Duration
(Seconds) Rate Rate Rate
(Seconds) Rate (Seconds)
Standard Banner — — 0.09% — — — — — — —
Rich Media
Commercial Break 13% 2.73 2.61% 5.46 72% 10% 3% — — —
Expandable Banner 11% 3.87 0.30% 23.62 15% 79% 69% 11% 7% 51.46
Floating Ad 3% 2.78 3.15% 6.64 48% 50% 34% — — —
Formats
Floating Ad With
5% 3.00 2.80% — — — — — — —
Reminder
Floating Expandable 10% 3.26 1.52% 5.26 62% 83% 70% 72% 5% 23.70
In Game 13% — 4.67% 16.60 94% 71% 47% — — —
Polite Banner 5% 3.36 0.17% 19.76 53% 68% 52% — — —
Push Down Banner 4% 2.29 0.28% — — — — 35% 3% 18.80
Video Strip 17% 4.02 0.17% 17.26 48% 59% 31% 46% 6% 70.23
Interaction Impressions Avg.
CTR CTR Avg. Video 50% Fully Total
Average Started with any Panel Expansion
IR (Rich (Standard Duration Played Played Expansion
Duration Rate Expansion Duration
Media) Media) (Seconds) Rate Rate Rate
(Seconds) Rate (Seconds)
Apparel 12% 3.33 0.30% 0.24% 15.04 47% 74% 36% 20% 7% 30.41
Auto 9% 3.35 0.23% 0.11% 19.63 16% 76% 64% 17% 9% 31.20
B2B 6% 3.11 0.48% 0.17% — — — — 12% 9% 65.23
Careers 10% 3.45 0.30% 0.09% — — — — 24% 16% 53.43
Consumer Packaged
7% 3.63 0.35% 0.11% 24.90 25% 54% 33% 9% 6% 48.21
Goods
Corporate 1% 3.09 0.09% 0.05% 22.05 7% 47% 31% 12% 8% 78.68
Electronics 8% 3.71 0.24% 0.12% 19.63 20% 66% 71% 15% 11% 42.02
Entertainment 11% 3.60 0.26% 0.13% 15.39 13% 61% 46% 9% 5% 49.49
Financial 5% 3.35 0.15% 0.07% 13.08 56% 76% 60% 6% 4% 46.69
Verticals
Gaming 7% 4.35 0.63% 0.08% 59.46 17% 49% 35% 22% 8% 61.49
Government/utilities 5% 3.18 0.17% 0.07% 24.48 62% 78% 66% 16% 4% 26.07
health/Beauty 13% 3.29 0.24% 0.17% 30.80 8% 63% 41% 19% 13% 61.72
Medical 6% 3.07 0.14% 0.10% 21.88 38% 72% 58% 7% 5% 36.49
news/Media 10% 2.64 0.27% 0.07% — — — — 14% 9% 22.58
Restaurant 1% 2.94 0.13% 0.09% 16.04 50% 39% 27% 3% 2% 33.43
Retail 11% 3.09 0.29% 0.07% 23.48 11% 40% 25% 16% 9% 28.08
Services 8% 3.08 0.38% 0.05% 23.23 56% 77% 63% 33% 21% 70.24
Sports 7% 2.98 0.28% 0.08% 11.37 55% 68% 52% 7% 4% 44.50
Tech/Internet 4% 3.59 0.17% 0.14% 14.82 51% 77% 65% 6% 4% 88.84
Telecom 7% 3.28 0.18% 0.09% 23.55 34% 81% 62% 15% 10% 48.35
Travel 4% 3.80 0.21% 0.08% 40.20 59% 78% 61% 13% 5% 31.28
north America: including uS and Canada. The benchmark is for a rolling year, last updated Q4 2008.
Interaction Rate (IR) user Initiated Interactions' divided by 'Served Impressions'
Click-through Rate (CTR) Frequency of Click-throughs as a percentage of served impressions. 'Clicks' divided by 'Impressions'.
Avg. Video Duration The average duration the video was played, including user and auto initiated videos (in seconds).
Started Rate The number of times the video started out of video's served Impressions
50% Played Rate The number of videos that played over 50% of their total video length divided by 'Video Started'.
Fully Played Rate The number of videos that were fully played divided by video that started
Total Expansion Rate Total expansions divided by served impressions. Including Auto-initiated expansions
Impressions with any Panel Expansion Rate The number of impressions with at least one panel expansion.
Avg. Expansion Duration The average time a panel was expanded, including user and auto initiated expansions
Rich Media Ads All Eyeblaster's format excluding standard banner, wallpaper and window ads
— not Available - either this measurement is not applicable to the format or there was not enough data to be statistically relevant
15 AnALYTICS BuLLETIn ISSuE 3 | MARCh 2009
16. Appendix 2
Benchmark for Europe Q4 ’07 - Q3 ’08
Performance Metrics (Format and Verticals)
Basic Metrics Video Metrics Expandable Metrics
Interaction Impressions Avg.
Avg. Video 50% Fully Total
Average Started with any Panel Expansion
IR CTR Duration Played Played Expansion
Duration Rate Expansion Duration
(Seconds) Rate Rate Rate
(Seconds) Rate (Seconds)
Standard Banner — — 0.14% — — — — — —
Rich Media —
Commercial Break 4% 2.30 3.96% 3.97 72% 48% 29% — — —
Expandable Banner 35% 3.27 0.66% 44.12 25% 62% 48% 41% 17% 52.60
Floating Ad 13% 2.68 4.16% 8.90 65% 51% 30% — — —
Formats
Floating Ad With
6% 2.38 4.66% — — — — — — —
Reminder
Floating Expandable 20% 2.58 0.48% — — — — 56% 14% 41.95
In Game 19% — 5.80% 17.17 99% 93% 80% — — —
Polite Banner 7% 2.95 0.21% 44.42 59% 67% 51% — — —
Push Down Banner 16% 3.46 0.15% 44.72 55% 84% 78% 22% 12% 57.37
Video Strip 23% 4.87 0.18% 63.20 10% 74% 60% 43% 8% 72.39
Interaction Impressions Avg.
CTR CTR Avg. Video 50% Fully Total
Average Started with any Panel Expansion
IR (Rich (Standard Duration Played Played Expansion
Duration Rate Expansion Duration
Media) Media) (Seconds) Rate Rate Rate
(Seconds) Rate (Seconds)
Apparel 15% 3.05 0.38% 0.17% 51.92 35% 57% 44% 47% 16% 47.41
Auto 14% 2.85 0.41% 0.15% 49.68 51% 68% 51% 42% 16% 55.11
B2B 9% 3.01 0.44% 0.34% — — — — 12% 6% 46.55
Careers 11% 3.00 0.54% 0.15% — — — — 25% 11% 51.31
Consumer
25% 3.58 0.57% 0.12% 39.68 31% 56% 42% 47% 18% 63.97
Packaged Goods
Corporate 18% 3.11 0.32% 0.08% — — — — 59% 26% 74.74
Electronics 21% 2.90 0.35% 0.13% 46.15 36% 72% 59% 42% 17% 48.40
Entertainment 28% 3.35 0.57% 0.18% 54.69 30% 62% 46% 40% 18% 39.44
Financial 17% 3.03 0.30% 0.14% 33.11 45% 62% 43% 36% 17% 64.06
Verticals
Gaming 19% 3.48 0.68% 0.21% 74.16 39% 61% 41% 41% 17% 51.12
Government/utilities 20% 3.33 0.47% 0.08% 49.45 41% 64% 42% 41% 16% 51.84
health/Beauty 18% 2.95 0.42% 0.10% 40.92 38% 62% 45% 30% 15% 47.21
Medical 12% 2.92 0.34% 0.28% 52.04 32% 50% 32% 54% 18% 57.12
news/Media 8% 3.04 0.35% 0.15% 93.30 45% 60% 42% 42% 14% 30.76
Restaurant 23% 3.13 0.38% 0.07% 25.33 53% 46% 37% 44% 15% 57.49
Retail 21% 2.83 0.47% 0.15% 42.97 57% 63% 44% 54% 22% 57.30
Services 21% 3.13 0.44% 0.08% 56.89 29% 62% 42% 42% 18% 55.09
Sports 8% 3.05 0.31% 0.10% 32.58 42% 75% 53% 38% 24% 28.36
Tech/Internet 15% 3.05 0.38% 0.14% 32.89 43% 82% 72% 37% 14% 46.39
Telecom 22% 2.98 0.43% 0.13% 49.42 48% 56% 41% 42% 17% 51.14
Travel 22% 3.15 0.49% 0.10% 42.38 59% 69% 57% 40% 14% 46.39
Europe: including Austria, Belgium, Denmark, France, Germany, Greece, Ireland, Italy, netherlands, Poland, Portugal, Spain, Sweden, Switzerland, uK.
The benchmark is for a rolling year, last updated Q4 2008.
Interaction Rate (IR) user Initiated Interactions' divided by 'Served Impressions'
Click-through Rate (CTR) Frequency of Click-throughs as a percentage of served impressions. 'Clicks' divided by 'Impressions'.
Avg. Video Duration The average duration the video was played, including user and auto initiated videos (in seconds).
Started Rate The number of times the video started out of video's served Impressions
50% Played Rate The number of videos that played over 50% of their total video length divided by 'Video Started'.
Fully Played Rate The number of videos that were fully played divided by video that started
Total Expansion Rate Total expansions divided by served impressions. Including Auto-initiated expansions
Impressions with any Panel Expansion Rate The number of impressions with at least one panel expansion.
Avg. Expansion Duration The average time a panel was expanded, including user and auto initiated expansions
Rich Media Ads All Eyeblaster's format excluding standard banner, wallpaper and window ads
— not Available - either this measurement is not applicable to the format or there was not enough data to be statistically relevant
16 AnALYTICS BuLLETIn ISSuE 3 | MARCh 2009
17. Appendix 3
Benchmark for Spain Q4 ’07 - Q3 ’08
Performance Metrics (Format and Verticals)
Basic Metrics Video Metrics Expandable Metrics
Avg.
Avg. Video 50% Fully Total Impressions
Started Expansion
IR CTR Duration Played Played Expansion with any Panel
Rate Duration
(Seconds) Rate Rate Rate Expansion Rate
(Seconds)
Standard Banner — 0.18% — — — — — — —
Rich Media
Commercial Break 4% 3.87% — — — — — — —
Formats
Expandable Banner 53% 1.02% 73.23 20% 56% 44% 64% 23% 67.38
Floating Ad 8% 2.08% — — — — — — —
Polite Banner 6% 0.25% 60.29 47% 65% 50% — — —
Push Down Banner 12% 0.21% — — — — — — —
Avg.
CTR CTR Avg. Video 50% Fully Total Impressions
Started Expansion
IR (Rich (Standard Duration Played Played Expansion with any Panel
Rate Duration
Media) Media) (Seconds) Rate Rate Rate Expansion Rate
(Seconds)
Apparel 21% 0.62% 0.45% 26.86 41% 55% 40% — — —
Auto 19% 0.44% 0.18% 21.57 54% 54% 40% 76% 30% 58.70
Consumer Packaged
42% 0.95% 0.15% 31.62 25% 62% 38% — — —
Goods
Electronics 17% 0.45% 0.16% 40.81 61% 74% 60% 47% 22% 57.02
Entertainment 51% 1.06% 0.22% 35.28 44% 66% 51% 53% 10% 51.19
Financial 14% 0.34% 0.12% 87.28 60% 84% 52% — — —
Verticals
Gaming 40% 0.94% 0.58% 38.05 44% 62% 45% — — —
Government/utilities 13% 0.30% 0.08% — — — — 46% 7% 73.58
health/Beauty 38% 0.81% 0.22% 54.38 35% 62% 46% 43% 13% 51.90
Retail 18% 0.65% —
Services 41% 0.54% 0.14% — — — — 79% 23% 63.60
Tech/Internet 9% 0.48% 0.14% 44.00 31% 69% 40% 68% 24% 57.03
Telecom 34% 0.50% 0.39% — — — — — — —
Travel 22% 0.68% — 35.64 60% 43% 29% 50% 8% 38.69
The benchmark is for a rolling year, last updated Q4 2008.
Interaction Rate (IR) user Initiated Interactions' divided by 'Served Impressions'
Click-through Rate (CTR) Frequency of Click-throughs as a percentage of served impressions. 'Clicks' divided by 'Impressions'.
Avg. Video Duration The average duration the video was played, including user and auto initiated videos (in seconds).
Started Rate The number of times the video started out of video's served Impressions
50% Played Rate The number of videos that played over 50% of their total video length divided by 'Video Started'.
Fully Played Rate The number of videos that were fully played divided by video that started
Total Expansion Rate Total expansions divided by served impressions. Including Auto-initiated expansions
Impressions with any Panel Expansion Rate The number of impressions with at least one panel expansion.
Avg. Expansion Duration The average time a panel was expanded, including user and auto initiated expansions
Rich Media Ads All Eyeblaster's format excluding standard banner, wallpaper and window ads
— not Available - either this measurement is not applicable to the format or there was not enough data to be statistically relevant
17 AnALYTICS BuLLETIn ISSuE 3 | MARCh 2009
18. Appendix 4
Benchmark for APAC Q4 ’07 - Q3 ’08
Performance Metrics (Format and Verticals)
Basic Metrics Video Metrics Expandable Metrics
Interaction Impressions Avg.
Avg. Video 50% Fully Total
Average Started with any Panel Expansion
IR CTR Duration Played Played Expansion
Duration Rate Expansion Duration
(Seconds) Rate Rate Rate
(Seconds) Rate (Seconds)
Standard Banner — — 0.11% — — — — — — —
Rich Media
Commercial
3% 2.70 3.14% — — — — — — —
Break
Expandable
21% 3.32 0.34% 22.96 6% 47% 36% 24% 11% 57.25
Banner
Formats
Floating Ad 8% 3.00 2.28% 50.89 76% 84% 78% — — —
Floating Ad With
5% 2.64 1.60% 11.04 63% 64% 21% — — —
Reminder
Floating
32% 2.93 0.43% — — — — 78% 32% 151.76
Expandable
Polite Banner 2% 2.82 0.11% 40.54 40% 66% 50% — — —
Push Down
26% 2.89 0.32% 17.32 23% 63% 47% 45% 11% 24.57
Banner
Interaction Impressions Avg.
CTR Avg. Video 50% Fully Total
Average CTR (Rich Started with any Panel Expansion
IR (Standard Duration Played Played Expansion
Duration Media) Rate Expansion Duration
Media) (Seconds) Rate Rate Rate
(Seconds) Rate (Seconds)
Apparel 10% 3.11 0.36% 0.18% 17.65 17% 49% 34% 26% 12% 67.31
Auto 7% 2.92 0.23% 0.22% 46.30 27% 78% 69% 17% 10% 52.91
Careers 24% 3.45 0.46% 0.04% — — — — 62% 22% 36.10
Consumer
12% 3.27 0.33% 0.12% 24.46 25% 70% 56% 26% 13% 50.61
Packaged Goods
Corporate 5% 2.55 0.12% 0.05% — — — — 10% 5% 50.69
Electronics 16% 2.88 0.32% 0.11% 34.20 34% 64% 50% 29% 15% 56.57
Entertainment 11% 3.09 0.35% 0.16% 29.23 16% 68% 55% 21% 10% 56.96
Financial 8% 2.76 0.15% 0.05% 19.87 22% 64% 52% 30% 8% 74.25
Gaming 6% 4.60 0.30% 0.10% 33.03 20% 54% 41% 6% 3% 46.98
Verticals
Government/
18% 3.77 0.31% 0.18% 69.27 20% 38% 24% 25% 13% 55.22
utilities
health/Beauty 12% 2.86 0.45% 0.11% 26.73 18% 47% 34% 23% 13% 62.54
Medical 15% 2.75 0.41% 0.09% — — — — 32% 16% 55.26
news/Media 11% 3.51 0.21% 0.18% — — — — 19% 11% 68.97
Restaurant 27% 4.16 0.67% 0.17% — — — — 37% 17% 32.42
Retail 10% 3.70 0.22% 0.11% 35.47 21% 50% 37% 21% 11% 36.07
Services 9% 2.77 0.14% 0.06% 50.42 13% 51% 35% 15% 8% 53.70
Sports 3% 2.36 0.40% 0.06% — — — — — — —
Tech/Internet 4% 2.61 0.12% 0.10% 27.48 27% 58% 38% 23% 11% 67.82
Telecom 13% 3.23 0.31% 0.09% 32.26 11% 52% 40% 34% 15% 49.93
Travel 15% 2.95 0.32% 0.15% 12.99 17% 63% 51% 25% 10% 48.51
APAC: including Australia, China, hong Kong, India, Japan, Korea, Malaysia, new Zealand, Pakistan, Singapore. The benchmark is for a rolling year, last updated Q4 2008.
Interaction Rate (IR) user Initiated Interactions' divided by 'Served Impressions'
Click-through Rate (CTR) Frequency of Click-throughs as a percentage of served impressions. 'Clicks' divided by 'Impressions'.
Avg. Video Duration The average duration the video was played, including user and auto initiated videos (in seconds).
Started Rate The number of times the video started out of video's served Impressions
50% Played Rate The number of videos that played over 50% of their total video length divided by 'Video Started'.
Fully Played Rate The number of videos that were fully played divided by video that started
Total Expansion Rate Total expansions divided by served impressions. Including Auto-initiated expansions
Impressions with any Panel Expansion Rate The number of impressions with at least one panel expansion.
Avg. Expansion Duration The average time a panel was expanded, including user and auto initiated expansions
Rich Media Ads All Eyeblaster's format excluding standard banner, wallpaper and window ads
— not Available - either this measurement is not applicable to the format or there was not enough data to be statistically relevant
18 AnALYTICS BuLLETIn ISSuE 3 | MARCh 2009